1 What: /sys/devices/system/cpu/ 1 What: /sys/devices/system/cpu/
2 Date: pre-git history 2 Date: pre-git history
3 Contact: Linux kernel mailing list <linu 3 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
4 Description: 4 Description:
5 A collection of both global an 5 A collection of both global and individual CPU attributes
6 6
7 Individual CPU attributes are 7 Individual CPU attributes are contained in subdirectories
8 named by the kernel's logical 8 named by the kernel's logical CPU number, e.g.:
9 9
10 /sys/devices/system/cpu/cpuX/ !! 10 /sys/devices/system/cpu/cpu#/
11 11
12 What: /sys/devices/system/cpu/kernel 12 What: /sys/devices/system/cpu/kernel_max
13 /sys/devices/system/cpu/offlin 13 /sys/devices/system/cpu/offline
14 /sys/devices/system/cpu/online 14 /sys/devices/system/cpu/online
15 /sys/devices/system/cpu/possib 15 /sys/devices/system/cpu/possible
16 /sys/devices/system/cpu/presen 16 /sys/devices/system/cpu/present
17 Date: December 2008 17 Date: December 2008
18 Contact: Linux kernel mailing list <linu 18 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
19 Description: CPU topology files that descri 19 Description: CPU topology files that describe kernel limits related to
20 hotplug. Briefly: 20 hotplug. Briefly:
21 21
22 kernel_max: the maximum cpu in 22 kernel_max: the maximum cpu index allowed by the kernel
23 configuration. 23 configuration.
24 24
25 offline: cpus that are not onl 25 offline: cpus that are not online because they have been
26 HOTPLUGGED off or exceed the l 26 HOTPLUGGED off or exceed the limit of cpus allowed by the
27 kernel configuration (kernel_m 27 kernel configuration (kernel_max above).
28 28
29 online: cpus that are online a 29 online: cpus that are online and being scheduled.
30 30
31 possible: cpus that have been 31 possible: cpus that have been allocated resources and can be
32 brought online if they are pre 32 brought online if they are present.
33 33
34 present: cpus that have been i 34 present: cpus that have been identified as being present in
35 the system. 35 the system.
36 36
37 See Documentation/admin-guide/ 37 See Documentation/admin-guide/cputopology.rst for more information.
38 38
39 39
40 What: /sys/devices/system/cpu/probe 40 What: /sys/devices/system/cpu/probe
41 /sys/devices/system/cpu/releas 41 /sys/devices/system/cpu/release
42 Date: November 2009 42 Date: November 2009
43 Contact: Linux kernel mailing list <linu 43 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
44 Description: Dynamic addition and removal o 44 Description: Dynamic addition and removal of CPU's. This is not hotplug
45 removal, this is meant complet 45 removal, this is meant complete removal/addition of the CPU
46 from the system. 46 from the system.
47 47
48 probe: writes to this file wil 48 probe: writes to this file will dynamically add a CPU to the
49 system. Information written t 49 system. Information written to the file to add CPU's is
50 architecture specific. 50 architecture specific.
51 51
52 release: writes to this file d 52 release: writes to this file dynamically remove a CPU from
53 the system. Information writt !! 53 the system. Information writtento the file to remove CPU's
54 is architecture specific. 54 is architecture specific.
55 55
56 What: /sys/devices/system/cpu/cpuX/n !! 56 What: /sys/devices/system/cpu/cpu#/node
57 Date: October 2009 57 Date: October 2009
58 Contact: Linux memory management mailin< 58 Contact: Linux memory management mailing list <linux-mm@kvack.org>
59 Description: Discover NUMA node a CPU belon 59 Description: Discover NUMA node a CPU belongs to
60 60
61 When CONFIG_NUMA is enabled, a 61 When CONFIG_NUMA is enabled, a symbolic link that points
62 to the corresponding NUMA node 62 to the corresponding NUMA node directory.
63 63
64 For example, the following sym 64 For example, the following symlink is created for cpu42
65 in NUMA node 2: 65 in NUMA node 2:
66 66
67 /sys/devices/system/cpu/cpu42/ 67 /sys/devices/system/cpu/cpu42/node2 -> ../../node/node2
68 68
69 69
70 What: /sys/devices/system/cpu/cpuX/t !! 70 What: /sys/devices/system/cpu/cpu#/topology/core_id
71 /sys/devices/system/cpu/cpuX/t !! 71 /sys/devices/system/cpu/cpu#/topology/core_siblings
72 /sys/devices/system/cpu/cpuX/t !! 72 /sys/devices/system/cpu/cpu#/topology/core_siblings_list
73 /sys/devices/system/cpu/cpuX/t !! 73 /sys/devices/system/cpu/cpu#/topology/physical_package_id
74 /sys/devices/system/cpu/cpuX/t !! 74 /sys/devices/system/cpu/cpu#/topology/thread_siblings
75 /sys/devices/system/cpu/cpuX/t !! 75 /sys/devices/system/cpu/cpu#/topology/thread_siblings_list
76 Date: December 2008 76 Date: December 2008
77 Contact: Linux kernel mailing list <linu 77 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
78 Description: CPU topology files that descri 78 Description: CPU topology files that describe a logical CPU's relationship
79 to other cores and threads in 79 to other cores and threads in the same physical package.
80 80
81 One cpuX directory is created !! 81 One cpu# directory is created per logical CPU in the system,
82 e.g. /sys/devices/system/cpu/c 82 e.g. /sys/devices/system/cpu/cpu42/.
83 83
84 Briefly, the files above are: 84 Briefly, the files above are:
85 85
86 core_siblings: internal kernel !! 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
87 within the same physical_packa 91 within the same physical_package_id.
88 92
89 core_siblings_list: human-read 93 core_siblings_list: human-readable list of the logical CPU
90 numbers within the same physic !! 94 numbers within the same physical_package_id as cpu#.
91 95
92 physical_package_id: physical !! 96 physical_package_id: physical package id of cpu#. Typically
93 corresponds to a physical sock 97 corresponds to a physical socket number, but the actual value
94 is architecture and platform d 98 is architecture and platform dependent.
95 99
96 thread_siblings: internal kern !! 100 thread_siblings: internel kernel map of cpu#'s hardware
97 threads within the same core a !! 101 threads within the same core as cpu#
98 <<
99 thread_siblings_list: human-re <<
100 threads within the same core a <<
101 102
102 ppin: human-readable Protected !! 103 thread_siblings_list: human-readable list of cpu#'s hardware
103 Number of the socket the cpu# !! 104 threads within the same core as cpu#
104 one per physical_package_id. F <<
105 admin. <<
106 105
107 See Documentation/admin-guide/ 106 See Documentation/admin-guide/cputopology.rst for more information.
108 107
109 108
110 What: /sys/devices/system/cpu/cpuidl !! 109 What: /sys/devices/system/cpu/cpuidle/current_driver
111 /sys/devices/system/cpu/cpuidl <<
112 /sys/devices/system/cpu/cpuidl <<
113 /sys/devices/system/cpu/cpuidl 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
114 Date: September 2007 113 Date: September 2007
115 Contact: Linux kernel mailing list <linu 114 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
116 Description: Discover cpuidle policy and me 115 Description: Discover cpuidle policy and mechanism
117 116
118 Various CPUs today support mul 117 Various CPUs today support multiple idle levels that are
119 differentiated by varying exit 118 differentiated by varying exit latencies and power
120 consumption during idle. 119 consumption during idle.
121 120
122 Idle policy (governor) is diff 121 Idle policy (governor) is differentiated from idle mechanism
123 (driver). !! 122 (driver)
124 123
125 available_governors: (RO) disp !! 124 current_driver: (RO) displays current idle mechanism
126 available governors. !! 125
>> 126 current_governor_ro: (RO) displays current idle policy
127 127
128 current_driver: (RO) displays !! 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
129 136
130 current_governor: (RW) display 137 current_governor: (RW) displays current idle policy. Users can
131 switch the governor at runtime 138 switch the governor at runtime by writing to this file.
132 139
133 current_governor_ro: (RO) disp <<
134 <<
135 See Documentation/admin-guide/ 140 See Documentation/admin-guide/pm/cpuidle.rst and
136 Documentation/driver-api/pm/cp 141 Documentation/driver-api/pm/cpuidle.rst for more information.
137 142
138 143
139 What: /sys/devices/system/cpu/cpuX/c !! 144 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/name
140 /sys/devices/system/cpu/cpuX/c 145 /sys/devices/system/cpu/cpuX/cpuidle/stateN/latency
141 /sys/devices/system/cpu/cpuX/c 146 /sys/devices/system/cpu/cpuX/cpuidle/stateN/power
142 /sys/devices/system/cpu/cpuX/c 147 /sys/devices/system/cpu/cpuX/cpuidle/stateN/time
143 /sys/devices/system/cpu/cpuX/c 148 /sys/devices/system/cpu/cpuX/cpuidle/stateN/usage
144 /sys/devices/system/cpu/cpuX/c 149 /sys/devices/system/cpu/cpuX/cpuidle/stateN/above
145 /sys/devices/system/cpu/cpuX/c 150 /sys/devices/system/cpu/cpuX/cpuidle/stateN/below
146 Date: September 2007 151 Date: September 2007
147 KernelVersion: v2.6.24 152 KernelVersion: v2.6.24
148 Contact: Linux power management list <li 153 Contact: Linux power management list <linux-pm@vger.kernel.org>
149 Description: 154 Description:
150 The directory /sys/devices/sys 155 The directory /sys/devices/system/cpu/cpuX/cpuidle contains per
151 logical CPU specific cpuidle i 156 logical CPU specific cpuidle information for each online cpu X.
152 The processor idle states whic 157 The processor idle states which are available for use have the
153 following attributes: 158 following attributes:
154 159
155 ======== ==== ================ !! 160 name: (RO) Name of the idle state (string).
156 name: (RO) Name of the idle <<
157 161
158 latency: (RO) The latency to e 162 latency: (RO) The latency to exit out of this idle state (in
159 microseconds). !! 163 microseconds).
160 164
161 power: (RO) The power consum !! 165 power: (RO) The power consumed while in this idle state (in
162 milliwatts). !! 166 milliwatts).
163 167
164 time: (RO) The total time s !! 168 time: (RO) The total time spent in this idle state (in microseconds).
165 (in microseconds <<
166 169
167 usage: (RO) Number of times !! 170 usage: (RO) Number of times this state was entered (a count).
168 171
169 above: (RO) Number of times !! 172 above: (RO) Number of times this state was entered, but the
170 observed CPU idl !! 173 observed CPU idle duration was too short for it (a count).
171 (a count). <<
172 174
173 below: (RO) Number of times !! 175 below: (RO) Number of times this state was entered, but the
174 observed CPU idl !! 176 observed CPU idle duration was too long for it (a count).
175 (a count). <<
176 ======== ==== ================ <<
177 177
178 What: /sys/devices/system/cpu/cpuX/c !! 178 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/desc
179 Date: February 2008 179 Date: February 2008
180 KernelVersion: v2.6.25 180 KernelVersion: v2.6.25
181 Contact: Linux power management list <li 181 Contact: Linux power management list <linux-pm@vger.kernel.org>
182 Description: 182 Description:
183 (RO) A small description about 183 (RO) A small description about the idle state (string).
184 184
185 185
186 What: /sys/devices/system/cpu/cpuX/c !! 186 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/disable
187 Date: March 2012 187 Date: March 2012
188 KernelVersion: v3.10 188 KernelVersion: v3.10
189 Contact: Linux power management list <li 189 Contact: Linux power management list <linux-pm@vger.kernel.org>
190 Description: 190 Description:
191 (RW) Option to disable this id 191 (RW) Option to disable this idle state (bool). The behavior and
192 the effect of the disable vari 192 the effect of the disable variable depends on the implementation
193 of a particular governor. In t 193 of a particular governor. In the ladder governor, for example,
194 it is not coherent, i.e. if on 194 it is not coherent, i.e. if one is disabling a light state, then
195 all deeper states are disabled 195 all deeper states are disabled as well, but the disable variable
196 does not reflect it. Likewise, 196 does not reflect it. Likewise, if one enables a deep state but a
197 lighter state still is disable 197 lighter state still is disabled, then this has no effect.
198 198
199 What: /sys/devices/system/cpu/cpuX/c !! 199 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/default_status
200 Date: December 2019 200 Date: December 2019
201 KernelVersion: v5.6 201 KernelVersion: v5.6
202 Contact: Linux power management list <li 202 Contact: Linux power management list <linux-pm@vger.kernel.org>
203 Description: 203 Description:
204 (RO) The default status of thi 204 (RO) The default status of this state, "enabled" or "disabled".
205 205
206 What: /sys/devices/system/cpu/cpuX/c !! 206 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/residency
207 Date: March 2014 207 Date: March 2014
208 KernelVersion: v3.15 208 KernelVersion: v3.15
209 Contact: Linux power management list <li 209 Contact: Linux power management list <linux-pm@vger.kernel.org>
210 Description: 210 Description:
211 (RO) Display the target reside 211 (RO) Display the target residency i.e. the minimum amount of
212 time (in microseconds) this cp 212 time (in microseconds) this cpu should spend in this idle state
213 to make the transition worth t 213 to make the transition worth the effort.
214 214
215 What: /sys/devices/system/cpu/cpuX/c !! 215 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/
216 Date: March 2018 216 Date: March 2018
217 KernelVersion: v4.17 217 KernelVersion: v4.17
218 Contact: Linux power management list <li 218 Contact: Linux power management list <linux-pm@vger.kernel.org>
219 Description: 219 Description:
220 Idle state usage statistics re 220 Idle state usage statistics related to suspend-to-idle.
221 221
222 This attribute group is only p 222 This attribute group is only present for states that can be
223 used in suspend-to-idle with s 223 used in suspend-to-idle with suspended timekeeping.
224 224
225 What: /sys/devices/system/cpu/cpuX/c !! 225 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/time
226 Date: March 2018 226 Date: March 2018
227 KernelVersion: v4.17 227 KernelVersion: v4.17
228 Contact: Linux power management list <li 228 Contact: Linux power management list <linux-pm@vger.kernel.org>
229 Description: 229 Description:
230 Total time spent by the CPU in 230 Total time spent by the CPU in suspend-to-idle (with scheduler
231 tick suspended) after requesti 231 tick suspended) after requesting this state.
232 232
233 What: /sys/devices/system/cpu/cpuX/c !! 233 What: /sys/devices/system/cpu/cpuX/cpuidle/stateN/s2idle/usage
234 Date: March 2018 234 Date: March 2018
235 KernelVersion: v4.17 235 KernelVersion: v4.17
236 Contact: Linux power management list <li 236 Contact: Linux power management list <linux-pm@vger.kernel.org>
237 Description: 237 Description:
238 Total number of times this sta 238 Total number of times this state has been requested by the CPU
239 while entering suspend-to-idle 239 while entering suspend-to-idle.
240 240
241 What: /sys/devices/system/cpu/cpuX/c !! 241 What: /sys/devices/system/cpu/cpu#/cpufreq/*
242 Date: pre-git history 242 Date: pre-git history
243 Contact: linux-pm@vger.kernel.org 243 Contact: linux-pm@vger.kernel.org
244 Description: Discover and change clock spee 244 Description: Discover and change clock speed of CPUs
245 245
246 Clock scaling allows you to ch 246 Clock scaling allows you to change the clock speed of the
247 CPUs on the fly. This is a nic 247 CPUs on the fly. This is a nice method to save battery
248 power, because the lower the c 248 power, because the lower the clock speed, the less power
249 the CPU consumes. 249 the CPU consumes.
250 250
251 There are many knobs to tweak 251 There are many knobs to tweak in this directory.
252 252
253 See files in Documentation/cpu 253 See files in Documentation/cpu-freq/ for more information.
254 254
255 255
256 What: /sys/devices/system/cpu/cpuX/c !! 256 What: /sys/devices/system/cpu/cpu#/cpufreq/freqdomain_cpus
257 Date: June 2013 257 Date: June 2013
258 Contact: linux-pm@vger.kernel.org 258 Contact: linux-pm@vger.kernel.org
259 Description: Discover CPUs in the same CPU 259 Description: Discover CPUs in the same CPU frequency coordination domain
260 260
261 freqdomain_cpus is the list of 261 freqdomain_cpus is the list of CPUs (online+offline) that share
262 the same clock/freq domain (po 262 the same clock/freq domain (possibly at the hardware level).
263 That information may be hidden 263 That information may be hidden from the cpufreq core and the
264 value of related_cpus may be d 264 value of related_cpus may be different from freqdomain_cpus. This
265 attribute is useful for user s 265 attribute is useful for user space DVFS controllers to get better
266 power/performance results for 266 power/performance results for platforms using acpi-cpufreq.
267 267
268 This file is only present if t !! 268 This file is only present if the acpi-cpufreq driver is in use.
269 drivers are in use. <<
270 269
271 270
272 What: /sys/devices/system/cpu/cpu*/c 271 What: /sys/devices/system/cpu/cpu*/cache/index3/cache_disable_{0,1}
273 Date: August 2008 272 Date: August 2008
274 KernelVersion: 2.6.27 273 KernelVersion: 2.6.27
275 Contact: Linux kernel mailing list <linu 274 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
276 Description: Disable L3 cache indices 275 Description: Disable L3 cache indices
277 276
278 These files exist in every CPU 277 These files exist in every CPU's cache/index3 directory. Each
279 cache_disable_{0,1} file corre 278 cache_disable_{0,1} file corresponds to one disable slot which
280 can be used to disable a cache 279 can be used to disable a cache index. Reading from these files
281 on a processor with this funct 280 on a processor with this functionality will return the currently
282 disabled index for that node. 281 disabled index for that node. There is one L3 structure per
283 node, or per internal node on 282 node, or per internal node on MCM machines. Writing a valid
284 index to one of these files wi !! 283 index to one of these files will cause the specificed cache
285 index to be disabled. 284 index to be disabled.
286 285
287 All AMD processors with L3 cac 286 All AMD processors with L3 caches provide this functionality.
288 For details, see BKDGs at 287 For details, see BKDGs at
289 https://www.amd.com/en/support !! 288 http://developer.amd.com/documentation/guides/Pages/default.aspx
290 289
291 290
292 What: /sys/devices/system/cpu/cpufre 291 What: /sys/devices/system/cpu/cpufreq/boost
293 Date: August 2012 292 Date: August 2012
294 Contact: Linux kernel mailing list <linu 293 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
295 Description: Processor frequency boosting c 294 Description: Processor frequency boosting control
296 295
297 This switch controls the boost 296 This switch controls the boost setting for the whole system.
298 Boosting allows the CPU and th 297 Boosting allows the CPU and the firmware to run at a frequency
299 beyond its nominal limit. !! 298 beyound it's nominal limit.
300 <<
301 More details can be found in 299 More details can be found in
302 Documentation/admin-guide/pm/c 300 Documentation/admin-guide/pm/cpufreq.rst
303 301
304 302
305 What: /sys/devices/system/cpu/cpuX/c !! 303 What: /sys/devices/system/cpu/cpu#/crash_notes
306 /sys/devices/system/cpu/cpuX/c !! 304 /sys/devices/system/cpu/cpu#/crash_notes_size
307 Date: April 2013 305 Date: April 2013
308 Contact: kexec@lists.infradead.org 306 Contact: kexec@lists.infradead.org
309 Description: address and size of the percpu 307 Description: address and size of the percpu note.
310 308
311 crash_notes: the physical addr 309 crash_notes: the physical address of the memory that holds the
312 note of cpuX. !! 310 note of cpu#.
313 311
314 crash_notes_size: size of the !! 312 crash_notes_size: size of the note of cpu#.
315 313
316 314
317 What: /sys/devices/system/cpu/intel_ 315 What: /sys/devices/system/cpu/intel_pstate/max_perf_pct
318 /sys/devices/system/cpu/intel_ 316 /sys/devices/system/cpu/intel_pstate/min_perf_pct
319 /sys/devices/system/cpu/intel_ 317 /sys/devices/system/cpu/intel_pstate/no_turbo
320 Date: February 2013 318 Date: February 2013
321 Contact: linux-pm@vger.kernel.org 319 Contact: linux-pm@vger.kernel.org
322 Description: Parameters for the Intel P-sta 320 Description: Parameters for the Intel P-state driver
323 321
324 Logic for selecting the curren 322 Logic for selecting the current P-state in Intel
325 Sandybridge+ processors. The t 323 Sandybridge+ processors. The three knobs control
326 limits for the P-state that wi 324 limits for the P-state that will be requested by the
327 driver. 325 driver.
328 326
329 max_perf_pct: limits the maxim 327 max_perf_pct: limits the maximum P state that will be requested by
330 the driver stated as a percent 328 the driver stated as a percentage of the available performance.
331 329
332 min_perf_pct: limits the minim 330 min_perf_pct: limits the minimum P state that will be requested by
333 the driver stated as a percent 331 the driver stated as a percentage of the available performance.
334 332
335 no_turbo: limits the driver to 333 no_turbo: limits the driver to selecting P states below the turbo
336 frequency range. 334 frequency range.
337 335
338 More details can be found in 336 More details can be found in
339 Documentation/admin-guide/pm/i 337 Documentation/admin-guide/pm/intel_pstate.rst
340 338
341 What: /sys/devices/system/cpu/cpu*/c 339 What: /sys/devices/system/cpu/cpu*/cache/index*/<set_of_attributes_mentioned_below>
342 Date: July 2014(documented, existed 340 Date: July 2014(documented, existed before August 2008)
343 Contact: Sudeep Holla <sudeep.holla@arm. 341 Contact: Sudeep Holla <sudeep.holla@arm.com>
344 Linux kernel mailing list <linu 342 Linux kernel mailing list <linux-kernel@vger.kernel.org>
345 Description: Parameters for the CPU cache a 343 Description: Parameters for the CPU cache attributes
346 344
347 allocation_policy: 345 allocation_policy:
348 - WriteAllocate: !! 346 - WriteAllocate: allocate a memory location to a cache line
349 alloca !! 347 on a cache miss because of a write
350 on a c !! 348 - ReadAllocate: allocate a memory location to a cache line
351 - ReadAllocate: <<
352 alloca <<
353 on a c 349 on a cache miss because of a read
354 - ReadWriteAllocate: !! 350 - ReadWriteAllocate: both writeallocate and readallocate
355 both w !! 351
>> 352 attributes: LEGACY used only on IA64 and is same as write_policy
356 353
357 coherency_line_size: !! 354 coherency_line_size: the minimum amount of data in bytes that gets
358 the minim <<
359 transferr 355 transferred from memory to cache
360 356
361 level: !! 357 level: the cache hierarchy in the multi-level cache configuration
362 the cache hierarchy in <<
363 358
364 number_of_sets: !! 359 number_of_sets: total number of sets in the cache, a set is a
365 total number o <<
366 collection of 360 collection of cache lines with the same cache index
367 361
368 physical_line_partition: !! 362 physical_line_partition: number of physical cache line per cache tag
369 number of phys <<
370 363
371 shared_cpu_list: !! 364 shared_cpu_list: the list of logical cpus sharing the cache
372 the list of lo <<
373 365
374 shared_cpu_map: !! 366 shared_cpu_map: logical cpu mask containing the list of cpus sharing
375 logical cpu ma <<
376 the cache 367 the cache
377 368
378 size: !! 369 size: the total cache size in kB
379 the total cache size i <<
380 370
381 type: 371 type:
382 - Instruction: cache t 372 - Instruction: cache that only holds instructions
383 - Data: cache that onl 373 - Data: cache that only caches data
384 - Unified: cache that 374 - Unified: cache that holds both data and instructions
385 375
386 ways_of_associativity: !! 376 ways_of_associativity: degree of freedom in placing a particular block
387 degree of freedom in p !! 377 of memory in the cache
388 of memory in the cache <<
389 378
390 write_policy: 379 write_policy:
391 - WriteThrough: !! 380 - WriteThrough: data is written to both the cache line
392 data i <<
393 and to 381 and to the block in the lower-level memory
394 - WriteBack: !! 382 - WriteBack: data is written only to the cache line and
395 data is w <<
396 the modif 383 the modified cache line is written to main
397 memory on 384 memory only when it is replaced
398 385
399 386
400 What: /sys/devices/system/cpu/cpu*/c 387 What: /sys/devices/system/cpu/cpu*/cache/index*/id
401 Date: September 2016 388 Date: September 2016
402 Contact: Linux kernel mailing list <linu 389 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
403 Description: Cache id 390 Description: Cache id
404 391
405 The id provides a unique numbe 392 The id provides a unique number for a specific instance of
406 a cache of a particular type. 393 a cache of a particular type. E.g. there may be a level
407 3 unified cache on each socket 394 3 unified cache on each socket in a server and we may
408 assign them ids 0, 1, 2, ... 395 assign them ids 0, 1, 2, ...
409 396
410 Note that id value can be non- 397 Note that id value can be non-contiguous. E.g. level 1
411 caches typically exist per cor 398 caches typically exist per core, but there may not be a
412 power of two cores on a socket 399 power of two cores on a socket, so these caches may be
413 numbered 0, 1, 2, 3, 4, 5, 8, 400 numbered 0, 1, 2, 3, 4, 5, 8, 9, 10, ...
414 401
415 What: /sys/devices/system/cpu/cpuX/c 402 What: /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats
416 /sys/devices/system/cpu/cpuX/c 403 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/turbo_stat
417 /sys/devices/system/cpu/cpuX/c 404 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/sub_turbo_stat
418 /sys/devices/system/cpu/cpuX/c 405 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/unthrottle
419 /sys/devices/system/cpu/cpuX/c 406 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/powercap
420 /sys/devices/system/cpu/cpuX/c 407 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overtemp
421 /sys/devices/system/cpu/cpuX/c 408 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/supply_fault
422 /sys/devices/system/cpu/cpuX/c 409 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/overcurrent
423 /sys/devices/system/cpu/cpuX/c 410 /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats/occ_reset
424 Date: March 2016 411 Date: March 2016
425 Contact: Linux kernel mailing list <linu 412 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
426 Linux for PowerPC mailing list< !! 413 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
427 Description: POWERNV CPUFreq driver's frequ 414 Description: POWERNV CPUFreq driver's frequency throttle stats directory and
428 attributes 415 attributes
429 416
430 'cpuX/cpufreq/throttle_stats' 417 'cpuX/cpufreq/throttle_stats' directory contains the CPU frequency
431 throttle stat attributes for t 418 throttle stat attributes for the chip. The throttle stats of a cpu
432 is common across all the cpus 419 is common across all the cpus belonging to a chip. Below are the
433 throttle attributes exported i 420 throttle attributes exported in the 'throttle_stats' directory:
434 421
435 - turbo_stat : This file gives 422 - turbo_stat : This file gives the total number of times the max
436 frequency is throttled to lo !! 423 frequency is throttled to lower frequency in turbo (at and above
437 nominal frequency) range of !! 424 nominal frequency) range of frequencies.
438 425
439 - sub_turbo_stat : This file g 426 - sub_turbo_stat : This file gives the total number of times the
440 max frequency is throttled t !! 427 max frequency is throttled to lower frequency in sub-turbo(below
441 nominal frequency) range of !! 428 nominal frequency) range of frequencies.
442 429
443 - unthrottle : This file gives 430 - unthrottle : This file gives the total number of times the max
444 frequency is unthrottled aft !! 431 frequency is unthrottled after being throttled.
445 432
446 - powercap : This file gives t 433 - powercap : This file gives the total number of times the max
447 frequency is throttled due t !! 434 frequency is throttled due to 'Power Capping'.
448 435
449 - overtemp : This file gives t 436 - overtemp : This file gives the total number of times the max
450 frequency is throttled due t !! 437 frequency is throttled due to 'CPU Over Temperature'.
451 438
452 - supply_fault : This file giv 439 - supply_fault : This file gives the total number of times the
453 max frequency is throttled d !! 440 max frequency is throttled due to 'Power Supply Failure'.
454 441
455 - overcurrent : This file give 442 - overcurrent : This file gives the total number of times the
456 max frequency is throttled d !! 443 max frequency is throttled due to 'Overcurrent'.
457 444
458 - occ_reset : This file gives 445 - occ_reset : This file gives the total number of times the max
459 frequency is throttled due t !! 446 frequency is throttled due to 'OCC Reset'.
460 447
461 The sysfs attributes represent 448 The sysfs attributes representing different throttle reasons like
462 powercap, overtemp, supply_fau 449 powercap, overtemp, supply_fault, overcurrent and occ_reset map to
463 the reasons provided by OCC fi 450 the reasons provided by OCC firmware for throttling the frequency.
464 451
465 What: /sys/devices/system/cpu/cpufre 452 What: /sys/devices/system/cpu/cpufreq/policyX/throttle_stats
466 /sys/devices/system/cpu/cpufre 453 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/turbo_stat
467 /sys/devices/system/cpu/cpufre 454 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/sub_turbo_stat
468 /sys/devices/system/cpu/cpufre 455 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/unthrottle
469 /sys/devices/system/cpu/cpufre 456 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/powercap
470 /sys/devices/system/cpu/cpufre 457 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overtemp
471 /sys/devices/system/cpu/cpufre 458 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/supply_fault
472 /sys/devices/system/cpu/cpufre 459 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/overcurrent
473 /sys/devices/system/cpu/cpufre 460 /sys/devices/system/cpu/cpufreq/policyX/throttle_stats/occ_reset
474 Date: March 2016 461 Date: March 2016
475 Contact: Linux kernel mailing list <linu 462 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
476 Linux for PowerPC mailing list< !! 463 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
477 Description: POWERNV CPUFreq driver's frequ 464 Description: POWERNV CPUFreq driver's frequency throttle stats directory and
478 attributes 465 attributes
479 466
480 'policyX/throttle_stats' direc 467 'policyX/throttle_stats' directory and all the attributes are same as
481 the /sys/devices/system/cpu/cp 468 the /sys/devices/system/cpu/cpuX/cpufreq/throttle_stats directory and
482 attributes which give the freq 469 attributes which give the frequency throttle information of the chip.
483 470
484 What: /sys/devices/system/cpu/cpuX/r 471 What: /sys/devices/system/cpu/cpuX/regs/
485 /sys/devices/system/cpu/cpuX/r 472 /sys/devices/system/cpu/cpuX/regs/identification/
486 /sys/devices/system/cpu/cpuX/r 473 /sys/devices/system/cpu/cpuX/regs/identification/midr_el1
487 /sys/devices/system/cpu/cpuX/r 474 /sys/devices/system/cpu/cpuX/regs/identification/revidr_el1
488 /sys/devices/system/cpu/cpuX/r <<
489 Date: June 2016 475 Date: June 2016
490 Contact: Linux ARM Kernel Mailing list < 476 Contact: Linux ARM Kernel Mailing list <linux-arm-kernel@lists.infradead.org>
491 Description: AArch64 CPU registers 477 Description: AArch64 CPU registers
492 <<
493 'identification' directory exp 478 'identification' directory exposes the CPU ID registers for
494 identifying model and revision !! 479 identifying model and revision of the CPU.
495 <<
496 What: /sys/devices/system/cpu/aarch3 <<
497 Date: May 2021 <<
498 Contact: Linux ARM Kernel Mailing list < <<
499 Description: Identifies the subset of CPUs <<
500 AArch32 (32-bit ARM) applicati <<
501 /sys/devices/system/cpu/{offli <<
502 If absent, then all or none of <<
503 applications and execve() will <<
504 480
505 What: /sys/devices/system/cpu/cpuX/c !! 481 What: /sys/devices/system/cpu/cpu#/cpu_capacity
506 Date: December 2016 482 Date: December 2016
507 Contact: Linux kernel mailing list <linu 483 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
508 Description: information about CPUs heterog 484 Description: information about CPUs heterogeneity.
509 485
510 cpu_capacity: capacity of cpuX !! 486 cpu_capacity: capacity of cpu#.
511 487
512 What: /sys/devices/system/cpu/vulner 488 What: /sys/devices/system/cpu/vulnerabilities
513 /sys/devices/system/cpu/vulner <<
514 /sys/devices/system/cpu/vulner <<
515 /sys/devices/system/cpu/vulner <<
516 /sys/devices/system/cpu/vulner <<
517 /sys/devices/system/cpu/vulner 489 /sys/devices/system/cpu/vulnerabilities/meltdown
518 /sys/devices/system/cpu/vulner <<
519 /sys/devices/system/cpu/vulner <<
520 /sys/devices/system/cpu/vulner <<
521 /sys/devices/system/cpu/vulner <<
522 /sys/devices/system/cpu/vulner 490 /sys/devices/system/cpu/vulnerabilities/spectre_v1
523 /sys/devices/system/cpu/vulner 491 /sys/devices/system/cpu/vulnerabilities/spectre_v2
>> 492 /sys/devices/system/cpu/vulnerabilities/spec_store_bypass
>> 493 /sys/devices/system/cpu/vulnerabilities/l1tf
>> 494 /sys/devices/system/cpu/vulnerabilities/mds
524 /sys/devices/system/cpu/vulner 495 /sys/devices/system/cpu/vulnerabilities/srbds
525 /sys/devices/system/cpu/vulner 496 /sys/devices/system/cpu/vulnerabilities/tsx_async_abort
>> 497 /sys/devices/system/cpu/vulnerabilities/itlb_multihit
526 Date: January 2018 498 Date: January 2018
527 Contact: Linux kernel mailing list <linu 499 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
528 Description: Information about CPU vulnerab 500 Description: Information about CPU vulnerabilities
529 501
530 The files are named after the 502 The files are named after the code names of CPU
531 vulnerabilities. The output of 503 vulnerabilities. The output of those files reflects the
532 state of the CPUs in the syste 504 state of the CPUs in the system. Possible output values:
533 505
534 ================ ============ <<
535 "Not affected" CPU is not a 506 "Not affected" CPU is not affected by the vulnerability
536 "Vulnerable" CPU is affec 507 "Vulnerable" CPU is affected and no mitigation in effect
537 "Mitigation: $M" CPU is affec 508 "Mitigation: $M" CPU is affected and mitigation $M is in effect
538 ================ ============ <<
539 509
540 See also: Documentation/admin- 510 See also: Documentation/admin-guide/hw-vuln/index.rst
541 511
542 What: /sys/devices/system/cpu/smt 512 What: /sys/devices/system/cpu/smt
543 /sys/devices/system/cpu/smt/ac 513 /sys/devices/system/cpu/smt/active
544 /sys/devices/system/cpu/smt/co 514 /sys/devices/system/cpu/smt/control
545 Date: June 2018 515 Date: June 2018
546 Contact: Linux kernel mailing list <linu 516 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
547 Description: Control Symmetric Multi Thread !! 517 Description: Control Symetric Multi Threading (SMT)
548 518
549 active: Tells whether SMT is 519 active: Tells whether SMT is active (enabled and siblings online)
550 520
551 control: Read/write interface 521 control: Read/write interface to control SMT. Possible
552 values: 522 values:
553 523
554 ================ ==== <<
555 "on" SMT 524 "on" SMT is enabled
556 "off" SMT 525 "off" SMT is disabled
557 "<N>" SMT <<
558 "forceoff" SMT 526 "forceoff" SMT is force disabled. Cannot be changed.
559 "notsupported" SMT 527 "notsupported" SMT is not supported by the CPU
560 "notimplemented" SMT 528 "notimplemented" SMT runtime toggling is not
561 impl 529 implemented for the architecture
562 ================ ==== <<
563 530
564 If control status is 531 If control status is "forceoff" or "notsupported" writes
565 are rejected. Note th !! 532 are rejected.
566 offline cores. <<
567 533
568 What: /sys/devices/system/cpu/cpuX/p !! 534 What: /sys/devices/system/cpu/cpu#/power/energy_perf_bias
569 Date: March 2019 535 Date: March 2019
570 Contact: linux-pm@vger.kernel.org 536 Contact: linux-pm@vger.kernel.org
571 Description: Intel Energy and Performance B 537 Description: Intel Energy and Performance Bias Hint (EPB)
572 538
573 EPB for the given CPU in a sli 539 EPB for the given CPU in a sliding scale 0 - 15, where a value
574 of 0 corresponds to a hint pre 540 of 0 corresponds to a hint preference for highest performance
575 and a value of 15 corresponds 541 and a value of 15 corresponds to the maximum energy savings.
576 542
577 In order to change the EPB val 543 In order to change the EPB value for the CPU, write either
578 a number in the 0 - 15 sliding 544 a number in the 0 - 15 sliding scale above, or one of the
579 strings: "performance", "balan 545 strings: "performance", "balance-performance", "normal",
580 "balance-power", "power" (that 546 "balance-power", "power" (that represent values reflected by
581 their meaning), to this attrib 547 their meaning), to this attribute.
582 548
583 This attribute is present for 549 This attribute is present for all online CPUs supporting the
584 Intel EPB feature. 550 Intel EPB feature.
585 551
586 What: /sys/devices/system/cpu/umwait 552 What: /sys/devices/system/cpu/umwait_control
587 /sys/devices/system/cpu/umwait 553 /sys/devices/system/cpu/umwait_control/enable_c02
588 /sys/devices/system/cpu/umwait 554 /sys/devices/system/cpu/umwait_control/max_time
589 Date: May 2019 555 Date: May 2019
590 Contact: Linux kernel mailing list <linu 556 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
591 Description: Umwait control 557 Description: Umwait control
592 558
593 enable_c02: Read/write interfa 559 enable_c02: Read/write interface to control umwait C0.2 state
594 Read returns C0.2 stat 560 Read returns C0.2 state status:
595 0: C0.2 is dis 561 0: C0.2 is disabled
596 1: C0.2 is ena 562 1: C0.2 is enabled
597 563
598 Write 'y' or '1' or ' 564 Write 'y' or '1' or 'on' to enable C0.2 state.
599 Write 'n' or '0' or ' 565 Write 'n' or '0' or 'off' to disable C0.2 state.
600 566
601 The interface is case 567 The interface is case insensitive.
602 568
603 max_time: Read/write interface 569 max_time: Read/write interface to control umwait maximum time
604 in TSC-quanta that t 570 in TSC-quanta that the CPU can reside in either C0.1
605 or C0.2 state. The t 571 or C0.2 state. The time is an unsigned 32-bit number.
606 Note that a value of 572 Note that a value of zero means there is no limit.
607 Low order two bits m 573 Low order two bits must be zero.
608 574
609 What: /sys/devices/system/cpu/sev <<
610 /sys/devices/system/cpu/sev/vm <<
611 Date: May 2024 <<
612 Contact: Linux kernel mailing list <linu <<
613 Description: Secure Encrypted Virtualizatio <<
614 <<
615 This directory is only present <<
616 <<
617 vmpl: Reports the Virtual Mach <<
618 the SEV-SNP guest is run <<
619 <<
620 <<
621 What: /sys/devices/system/cpu/svm 575 What: /sys/devices/system/cpu/svm
622 Date: August 2019 576 Date: August 2019
623 Contact: Linux kernel mailing list <linu 577 Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org>
624 Linux for PowerPC mailing list< !! 578 Linux for PowerPC mailing list <linuxppc-dev@ozlabs.org>
625 Description: Secure Virtual Machine 579 Description: Secure Virtual Machine
626 580
627 If 1, it means the system is u 581 If 1, it means the system is using the Protected Execution
628 Facility in POWER9 and newer p 582 Facility in POWER9 and newer processors. i.e., it is a Secure
629 Virtual Machine. 583 Virtual Machine.
630 <<
631 What: /sys/devices/system/cpu/cpuX/p <<
632 Date: Apr 2005 <<
633 Contact: Linux for PowerPC mailing list< <<
634 Description: PURR ticks for this CPU since <<
635 <<
636 The Processor Utilization Reso <<
637 a 64-bit counter which provide <<
638 resources used by the CPU thre <<
639 register increases monotonical <<
640 exposes the number of PURR tic <<
641 <<
642 What: /sys/devices/system/cpu/cpuX/s <<
643 Date: Dec 2006 <<
644 Contact: Linux for PowerPC mailing list< <<
645 Description: SPURR ticks for this CPU since <<
646 <<
647 The Scaled Processor Utilizati <<
648 (SPURR) is a 64-bit counter th <<
649 invariant estimate of the reso <<
650 thread. The contents of this r <<
651 monotonically. This sysfs inte <<
652 of SPURR ticks for cpuX. <<
653 <<
654 What: /sys/devices/system/cpu/cpuX/i <<
655 Date: Apr 2020 <<
656 Contact: Linux for PowerPC mailing list< <<
657 Description: PURR ticks for cpuX when it wa <<
658 <<
659 This sysfs interface exposes t <<
660 for cpuX when it was idle. <<
661 <<
662 What: /sys/devices/system/cpu/cpuX/i <<
663 Date: Apr 2020 <<
664 Contact: Linux for PowerPC mailing list< <<
665 Description: SPURR ticks for cpuX when it w <<
666 <<
667 This sysfs interface exposes t <<
668 for cpuX when it was idle. <<
669 <<
670 What: /sys/devices/system/cpu/cpuX/m <<
671 Date: July 2021 <<
672 Contact: Linux ARM Kernel Mailing list < <<
673 Description: Preferred MTE tag checking mod <<
674 <<
675 When a user program specifies <<
676 mode, this sysfs node is used <<
677 be preferred when scheduling a <<
678 values: <<
679 <<
680 ================ ============ <<
681 "sync" Prefer synch <<
682 "asymm" Prefer asymm <<
683 "async" Prefer async <<
684 ================ ============ <<
685 <<
686 See also: Documentation/arch/a <<
687 <<
688 What: /sys/devices/system/cpu/nohz_f <<
689 Date: Apr 2015 <<
690 Contact: Linux kernel mailing list <linu <<
691 Description: <<
692 (RO) the list of CPUs that are <<
693 These CPUs are set by boot par <<
694 <<
695 What: /sys/devices/system/cpu/isolat <<
696 Date: Apr 2015 <<
697 Contact: Linux kernel mailing list <linu <<
698 Description: <<
699 (RO) the list of CPUs that are <<
700 participate in load balancing. <<
701 boot parameter "isolcpus=". <<
702 <<
703 What: /sys/devices/system/cpu/crash_ <<
704 Date: Aug 2023 <<
705 Contact: Linux kernel mailing list <linu <<
706 Description: <<
707 (RO) indicates whether or not <<
708 segments on memory hot un/plug <<
709 need to reload kdump kernel. <<
710 <<
711 What: /sys/devices/system/cpu/enable <<
712 Date: Nov 2022 <<
713 Contact: Linux kernel mailing list <linu <<
714 Description: <<
715 (RO) the list of CPUs that can <<
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.