Alex Chiang | 2ceb3fb | 2009-10-21 21:45:20 -0600 | [diff] [blame] | 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 | |
Alex Chiang | d93fc86 | 2009-10-21 21:45:25 -0600 | [diff] [blame] | 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 | |
Nathan Fontenot | 12633e8 | 2009-11-25 17:23:25 +0000 | [diff] [blame] | 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. |
Alex Chiang | 657348a | 2009-10-21 22:15:30 -0600 | [diff] [blame] | 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 | |
Alex Chiang | 663fb2f | 2009-10-21 21:45:31 -0600 | [diff] [blame] | 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 | |
Alex Chiang | c1fb5c4 | 2009-10-21 21:45:41 -0600 | [diff] [blame] | 109 | What: /sys/devices/system/cpu/cpuidle/current_driver |
| 110 | /sys/devices/system/cpu/cpuidle/current_governer_ro |
| 111 | Date: September 2007 |
| 112 | Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org> |
| 113 | Description: Discover cpuidle policy and mechanism |
| 114 | |
| 115 | Various CPUs today support multiple idle levels that are |
| 116 | differentiated by varying exit latencies and power |
| 117 | consumption during idle. |
| 118 | |
| 119 | Idle policy (governor) is differentiated from idle mechanism |
| 120 | (driver) |
| 121 | |
| 122 | current_driver: displays current idle mechanism |
| 123 | |
| 124 | current_governor_ro: displays current idle policy |
| 125 | |
| 126 | See files in Documentation/cpuidle/ for more information. |
| 127 | |
| 128 | |
Alex Chiang | 0cda8b9 | 2009-10-21 21:45:46 -0600 | [diff] [blame] | 129 | What: /sys/devices/system/cpu/cpu#/cpufreq/* |
| 130 | Date: pre-git history |
Viresh Kumar | dec102a | 2014-04-22 10:42:05 +0530 | [diff] [blame] | 131 | Contact: linux-pm@vger.kernel.org |
Alex Chiang | 0cda8b9 | 2009-10-21 21:45:46 -0600 | [diff] [blame] | 132 | Description: Discover and change clock speed of CPUs |
| 133 | |
| 134 | Clock scaling allows you to change the clock speed of the |
| 135 | CPUs on the fly. This is a nice method to save battery |
| 136 | power, because the lower the clock speed, the less power |
| 137 | the CPU consumes. |
| 138 | |
| 139 | There are many knobs to tweak in this directory. |
| 140 | |
| 141 | See files in Documentation/cpu-freq/ for more information. |
| 142 | |
| 143 | In particular, read Documentation/cpu-freq/user-guide.txt |
| 144 | to learn how to control the knobs. |
| 145 | |
| 146 | |
Lan Tianyu | f4fd379 | 2013-06-27 15:08:54 +0800 | [diff] [blame] | 147 | What: /sys/devices/system/cpu/cpu#/cpufreq/freqdomain_cpus |
| 148 | Date: June 2013 |
Viresh Kumar | dec102a | 2014-04-22 10:42:05 +0530 | [diff] [blame] | 149 | Contact: linux-pm@vger.kernel.org |
Lan Tianyu | f4fd379 | 2013-06-27 15:08:54 +0800 | [diff] [blame] | 150 | Description: Discover CPUs in the same CPU frequency coordination domain |
| 151 | |
| 152 | freqdomain_cpus is the list of CPUs (online+offline) that share |
| 153 | the same clock/freq domain (possibly at the hardware level). |
| 154 | That information may be hidden from the cpufreq core and the |
| 155 | value of related_cpus may be different from freqdomain_cpus. This |
| 156 | attribute is useful for user space DVFS controllers to get better |
| 157 | power/performance results for platforms using acpi-cpufreq. |
| 158 | |
| 159 | This file is only present if the acpi-cpufreq driver is in use. |
| 160 | |
| 161 | |
Borislav Petkov | eecaaba | 2011-05-16 15:39:48 +0200 | [diff] [blame] | 162 | What: /sys/devices/system/cpu/cpu*/cache/index3/cache_disable_{0,1} |
| 163 | Date: August 2008 |
Mark Langsdorf | 2fad2d9 | 2009-04-09 15:31:53 +0200 | [diff] [blame] | 164 | KernelVersion: 2.6.27 |
Borislav Petkov | eecaaba | 2011-05-16 15:39:48 +0200 | [diff] [blame] | 165 | Contact: discuss@x86-64.org |
| 166 | Description: Disable L3 cache indices |
Mark Langsdorf | 2fad2d9 | 2009-04-09 15:31:53 +0200 | [diff] [blame] | 167 | |
Borislav Petkov | eecaaba | 2011-05-16 15:39:48 +0200 | [diff] [blame] | 168 | These files exist in every CPU's cache/index3 directory. Each |
| 169 | cache_disable_{0,1} file corresponds to one disable slot which |
| 170 | can be used to disable a cache index. Reading from these files |
| 171 | on a processor with this functionality will return the currently |
| 172 | disabled index for that node. There is one L3 structure per |
| 173 | node, or per internal node on MCM machines. Writing a valid |
| 174 | index to one of these files will cause the specificed cache |
| 175 | index to be disabled. |
| 176 | |
| 177 | All AMD processors with L3 caches provide this functionality. |
| 178 | For details, see BKDGs at |
| 179 | http://developer.amd.com/documentation/guides/Pages/default.aspx |
Andre Przywara | 615b730 | 2012-09-04 08:28:07 +0000 | [diff] [blame] | 180 | |
| 181 | |
| 182 | What: /sys/devices/system/cpu/cpufreq/boost |
| 183 | Date: August 2012 |
| 184 | Contact: Linux kernel mailing list <linux-kernel@vger.kernel.org> |
| 185 | Description: Processor frequency boosting control |
| 186 | |
| 187 | This switch controls the boost setting for the whole system. |
| 188 | Boosting allows the CPU and the firmware to run at a frequency |
| 189 | beyound it's nominal limit. |
| 190 | More details can be found in Documentation/cpu-freq/boost.txt |
Zhang Yanfei | c4fd675 | 2013-03-28 16:16:45 +0800 | [diff] [blame] | 191 | |
| 192 | |
| 193 | What: /sys/devices/system/cpu/cpu#/crash_notes |
| 194 | /sys/devices/system/cpu/cpu#/crash_notes_size |
| 195 | Date: April 2013 |
| 196 | Contact: kexec@lists.infradead.org |
| 197 | Description: address and size of the percpu note. |
| 198 | |
| 199 | crash_notes: the physical address of the memory that holds the |
| 200 | note of cpu#. |
| 201 | |
| 202 | crash_notes_size: size of the note of cpu#. |
Ramkumar Ramachandra | fbe299e | 2014-01-06 18:24:26 +0530 | [diff] [blame] | 203 | |
| 204 | |
| 205 | What: /sys/devices/system/cpu/intel_pstate/max_perf_pct |
| 206 | /sys/devices/system/cpu/intel_pstate/min_perf_pct |
| 207 | /sys/devices/system/cpu/intel_pstate/no_turbo |
| 208 | Date: February 2013 |
| 209 | Contact: linux-pm@vger.kernel.org |
| 210 | Description: Parameters for the Intel P-state driver |
| 211 | |
| 212 | Logic for selecting the current P-state in Intel |
| 213 | Sandybridge+ processors. The three knobs control |
| 214 | limits for the P-state that will be requested by the |
| 215 | driver. |
| 216 | |
| 217 | max_perf_pct: limits the maximum P state that will be requested by |
| 218 | the driver stated as a percentage of the available performance. |
| 219 | |
| 220 | min_perf_pct: limits the minimum P state that will be requested by |
| 221 | the driver stated as a percentage of the available performance. |
| 222 | |
| 223 | no_turbo: limits the driver to selecting P states below the turbo |
| 224 | frequency range. |
| 225 | |
| 226 | More details can be found in Documentation/cpu-freq/intel-pstate.txt |
Sudeep Holla | 246246c | 2014-09-30 14:48:25 +0100 | [diff] [blame] | 227 | |
| 228 | What: /sys/devices/system/cpu/cpu*/cache/index*/<set_of_attributes_mentioned_below> |
| 229 | Date: July 2014(documented, existed before August 2008) |
| 230 | Contact: Sudeep Holla <sudeep.holla@arm.com> |
| 231 | Linux kernel mailing list <linux-kernel@vger.kernel.org> |
| 232 | Description: Parameters for the CPU cache attributes |
| 233 | |
| 234 | allocation_policy: |
| 235 | - WriteAllocate: allocate a memory location to a cache line |
| 236 | on a cache miss because of a write |
| 237 | - ReadAllocate: allocate a memory location to a cache line |
| 238 | on a cache miss because of a read |
| 239 | - ReadWriteAllocate: both writeallocate and readallocate |
| 240 | |
| 241 | attributes: LEGACY used only on IA64 and is same as write_policy |
| 242 | |
| 243 | coherency_line_size: the minimum amount of data in bytes that gets |
| 244 | transferred from memory to cache |
| 245 | |
Will Deacon | 2539b25 | 2015-05-08 14:45:34 +0100 | [diff] [blame] | 246 | level: the cache hierarchy in the multi-level cache configuration |
Sudeep Holla | 246246c | 2014-09-30 14:48:25 +0100 | [diff] [blame] | 247 | |
| 248 | number_of_sets: total number of sets in the cache, a set is a |
| 249 | collection of cache lines with the same cache index |
| 250 | |
| 251 | physical_line_partition: number of physical cache line per cache tag |
| 252 | |
| 253 | shared_cpu_list: the list of logical cpus sharing the cache |
| 254 | |
| 255 | shared_cpu_map: logical cpu mask containing the list of cpus sharing |
| 256 | the cache |
| 257 | |
| 258 | size: the total cache size in kB |
| 259 | |
| 260 | type: |
| 261 | - Instruction: cache that only holds instructions |
| 262 | - Data: cache that only caches data |
| 263 | - Unified: cache that holds both data and instructions |
| 264 | |
| 265 | ways_of_associativity: degree of freedom in placing a particular block |
| 266 | of memory in the cache |
| 267 | |
| 268 | write_policy: |
| 269 | - WriteThrough: data is written to both the cache line |
| 270 | and to the block in the lower-level memory |
| 271 | - WriteBack: data is written only to the cache line and |
| 272 | the modified cache line is written to main |
| 273 | memory only when it is replaced |