1=================================== 2Generic Thermal Sysfs driver How To 3=================================== 4 5Written by Sujith Thomas <sujith.thomas@intel.com>, Zhang Rui <rui.zhang@intel.com> 6 7Copyright (c) 2008 Intel Corporation 8 9 100. Introduction 11=============== 12 13The generic thermal sysfs provides a set of interfaces for thermal zone 14devices (sensors) and thermal cooling devices (fan, processor...) to register 15with the thermal management solution and to be a part of it. 16 17This how-to focuses on enabling new thermal zone and cooling devices to 18participate in thermal management. 19This solution is platform independent and any type of thermal zone devices 20and cooling devices should be able to make use of the infrastructure. 21 22The main task of the thermal sysfs driver is to expose thermal zone attributes 23as well as cooling device attributes to the user space. 24An intelligent thermal management application can make decisions based on 25inputs from thermal zone attributes (the current temperature and trip point 26temperature) and throttle appropriate devices. 27 28- `[0-*]` denotes any positive number starting from 0 29- `[1-*]` denotes any positive number starting from 1 30 311. thermal sysfs driver interface functions 32=========================================== 33 341.1 thermal zone device interface 35--------------------------------- 36 37 :: 38 39 struct thermal_zone_device * 40 thermal_zone_device_register_with_trips(const char *type, 41 const struct thermal_trip *trips, 42 int num_trips, void *devdata, 43 const struct thermal_zone_device_ops *ops, 44 const struct thermal_zone_params *tzp, 45 unsigned int passive_delay, 46 unsigned int polling_delay) 47 48 This interface function adds a new thermal zone device (sensor) to the 49 /sys/class/thermal folder as `thermal_zone[0-*]`. It tries to bind all the 50 thermal cooling devices registered to it at the same time. 51 52 type: 53 the thermal zone type. 54 trips: 55 the table of trip points for this thermal zone. 56 devdata: 57 device private data 58 ops: 59 thermal zone device call-backs. 60 61 .bind: 62 bind the thermal zone device with a thermal cooling device. 63 .unbind: 64 unbind the thermal zone device with a thermal cooling device. 65 .get_temp: 66 get the current temperature of the thermal zone. 67 .set_trips: 68 set the trip points window. Whenever the current temperature 69 is updated, the trip points immediately below and above the 70 current temperature are found. 71 .change_mode: 72 change the mode (enabled/disabled) of the thermal zone. 73 .set_trip_temp: 74 set the temperature of a given trip point. 75 .get_crit_temp: 76 get the critical temperature for this thermal zone. 77 .set_emul_temp: 78 set the emulation temperature which helps in debugging 79 different threshold temperature points. 80 .get_trend: 81 get the trend of most recent zone temperature changes. 82 .hot: 83 hot trip point crossing handler. 84 .critical: 85 critical trip point crossing handler. 86 tzp: 87 thermal zone platform parameters. 88 passive_delay: 89 number of milliseconds to wait between polls when performing passive 90 cooling. 91 polling_delay: 92 number of milliseconds to wait between polls when checking 93 whether trip points have been crossed (0 for interrupt driven systems). 94 95 :: 96 97 void thermal_zone_device_unregister(struct thermal_zone_device *tz) 98 99 This interface function removes the thermal zone device. 100 It deletes the corresponding entry from /sys/class/thermal folder and 101 unbinds all the thermal cooling devices it uses. 102 103 :: 104 105 struct thermal_zone_device 106 *thermal_zone_of_sensor_register(struct device *dev, int sensor_id, 107 void *data, 108 const struct thermal_zone_of_device_ops *ops) 109 110 This interface adds a new sensor to a DT thermal zone. 111 This function will search the list of thermal zones described in 112 device tree and look for the zone that refer to the sensor device 113 pointed by dev->of_node as temperature providers. For the zone 114 pointing to the sensor node, the sensor will be added to the DT 115 thermal zone device. 116 117 The parameters for this interface are: 118 119 dev: 120 Device node of sensor containing valid node pointer in 121 dev->of_node. 122 sensor_id: 123 a sensor identifier, in case the sensor IP has more 124 than one sensors 125 data: 126 a private pointer (owned by the caller) that will be 127 passed back, when a temperature reading is needed. 128 ops: 129 `struct thermal_zone_of_device_ops *`. 130 131 ============== ======================================= 132 get_temp a pointer to a function that reads the 133 sensor temperature. This is mandatory 134 callback provided by sensor driver. 135 set_trips a pointer to a function that sets a 136 temperature window. When this window is 137 left the driver must inform the thermal 138 core via thermal_zone_device_update. 139 get_trend a pointer to a function that reads the 140 sensor temperature trend. 141 set_emul_temp a pointer to a function that sets 142 sensor emulated temperature. 143 ============== ======================================= 144 145 The thermal zone temperature is provided by the get_temp() function 146 pointer of thermal_zone_of_device_ops. When called, it will 147 have the private pointer @data back. 148 149 It returns error pointer if fails otherwise valid thermal zone device 150 handle. Caller should check the return handle with IS_ERR() for finding 151 whether success or not. 152 153 :: 154 155 void thermal_zone_of_sensor_unregister(struct device *dev, 156 struct thermal_zone_device *tzd) 157 158 This interface unregisters a sensor from a DT thermal zone which was 159 successfully added by interface thermal_zone_of_sensor_register(). 160 This function removes the sensor callbacks and private data from the 161 thermal zone device registered with thermal_zone_of_sensor_register() 162 interface. It will also silent the zone by remove the .get_temp() and 163 get_trend() thermal zone device callbacks. 164 165 :: 166 167 struct thermal_zone_device 168 *devm_thermal_zone_of_sensor_register(struct device *dev, 169 int sensor_id, 170 void *data, 171 const struct thermal_zone_of_device_ops *ops) 172 173 This interface is resource managed version of 174 thermal_zone_of_sensor_register(). 175 176 All details of thermal_zone_of_sensor_register() described in 177 section 1.1.3 is applicable here. 178 179 The benefit of using this interface to register sensor is that it 180 is not require to explicitly call thermal_zone_of_sensor_unregister() 181 in error path or during driver unbinding as this is done by driver 182 resource manager. 183 184 :: 185 186 void devm_thermal_zone_of_sensor_unregister(struct device *dev, 187 struct thermal_zone_device *tzd) 188 189 This interface is resource managed version of 190 thermal_zone_of_sensor_unregister(). 191 All details of thermal_zone_of_sensor_unregister() described in 192 section 1.1.4 is applicable here. 193 Normally this function will not need to be called and the resource 194 management code will ensure that the resource is freed. 195 196 :: 197 198 int thermal_zone_get_slope(struct thermal_zone_device *tz) 199 200 This interface is used to read the slope attribute value 201 for the thermal zone device, which might be useful for platform 202 drivers for temperature calculations. 203 204 :: 205 206 int thermal_zone_get_offset(struct thermal_zone_device *tz) 207 208 This interface is used to read the offset attribute value 209 for the thermal zone device, which might be useful for platform 210 drivers for temperature calculations. 211 2121.2 thermal cooling device interface 213------------------------------------ 214 215 216 :: 217 218 struct thermal_cooling_device 219 *thermal_cooling_device_register(char *name, 220 void *devdata, struct thermal_cooling_device_ops *) 221 222 This interface function adds a new thermal cooling device (fan/processor/...) 223 to /sys/class/thermal/ folder as `cooling_device[0-*]`. It tries to bind itself 224 to all the thermal zone devices registered at the same time. 225 226 name: 227 the cooling device name. 228 devdata: 229 device private data. 230 ops: 231 thermal cooling devices call-backs. 232 233 .get_max_state: 234 get the Maximum throttle state of the cooling device. 235 .get_cur_state: 236 get the Currently requested throttle state of the 237 cooling device. 238 .set_cur_state: 239 set the Current throttle state of the cooling device. 240 241 :: 242 243 void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev) 244 245 This interface function removes the thermal cooling device. 246 It deletes the corresponding entry from /sys/class/thermal folder and 247 unbinds itself from all the thermal zone devices using it. 248 2491.3 interface for binding a thermal zone device with a thermal cooling device 250----------------------------------------------------------------------------- 251 252 :: 253 254 int thermal_zone_bind_cooling_device(struct thermal_zone_device *tz, 255 int trip, struct thermal_cooling_device *cdev, 256 unsigned long upper, unsigned long lower, unsigned int weight); 257 258 This interface function binds a thermal cooling device to a particular trip 259 point of a thermal zone device. 260 261 This function is usually called in the thermal zone device .bind callback. 262 263 tz: 264 the thermal zone device 265 cdev: 266 thermal cooling device 267 trip: 268 indicates which trip point in this thermal zone the cooling device 269 is associated with. 270 upper: 271 the Maximum cooling state for this trip point. 272 THERMAL_NO_LIMIT means no upper limit, 273 and the cooling device can be in max_state. 274 lower: 275 the Minimum cooling state can be used for this trip point. 276 THERMAL_NO_LIMIT means no lower limit, 277 and the cooling device can be in cooling state 0. 278 weight: 279 the influence of this cooling device in this thermal 280 zone. See 1.4.1 below for more information. 281 282 :: 283 284 int thermal_zone_unbind_cooling_device(struct thermal_zone_device *tz, 285 int trip, struct thermal_cooling_device *cdev); 286 287 This interface function unbinds a thermal cooling device from a particular 288 trip point of a thermal zone device. This function is usually called in 289 the thermal zone device .unbind callback. 290 291 tz: 292 the thermal zone device 293 cdev: 294 thermal cooling device 295 trip: 296 indicates which trip point in this thermal zone the cooling device 297 is associated with. 298 2991.4 Thermal Zone Parameters 300--------------------------- 301 302 :: 303 304 struct thermal_zone_params 305 306 This structure defines the platform level parameters for a thermal zone. 307 This data, for each thermal zone should come from the platform layer. 308 This is an optional feature where some platforms can choose not to 309 provide this data. 310 311 .governor_name: 312 Name of the thermal governor used for this zone 313 .no_hwmon: 314 a boolean to indicate if the thermal to hwmon sysfs interface 315 is required. when no_hwmon == false, a hwmon sysfs interface 316 will be created. when no_hwmon == true, nothing will be done. 317 In case the thermal_zone_params is NULL, the hwmon interface 318 will be created (for backward compatibility). 319 3202. sysfs attributes structure 321============================= 322 323== ================ 324RO read only value 325WO write only value 326RW read/write value 327== ================ 328 329Thermal sysfs attributes will be represented under /sys/class/thermal. 330Hwmon sysfs I/F extension is also available under /sys/class/hwmon 331if hwmon is compiled in or built as a module. 332 333Thermal zone device sys I/F, created once it's registered:: 334 335 /sys/class/thermal/thermal_zone[0-*]: 336 |---type: Type of the thermal zone 337 |---temp: Current temperature 338 |---mode: Working mode of the thermal zone 339 |---policy: Thermal governor used for this zone 340 |---available_policies: Available thermal governors for this zone 341 |---trip_point_[0-*]_temp: Trip point temperature 342 |---trip_point_[0-*]_type: Trip point type 343 |---trip_point_[0-*]_hyst: Hysteresis value for this trip point 344 |---emul_temp: Emulated temperature set node 345 |---sustainable_power: Sustainable dissipatable power 346 |---k_po: Proportional term during temperature overshoot 347 |---k_pu: Proportional term during temperature undershoot 348 |---k_i: PID's integral term in the power allocator gov 349 |---k_d: PID's derivative term in the power allocator 350 |---integral_cutoff: Offset above which errors are accumulated 351 |---slope: Slope constant applied as linear extrapolation 352 |---offset: Offset constant applied as linear extrapolation 353 354Thermal cooling device sys I/F, created once it's registered:: 355 356 /sys/class/thermal/cooling_device[0-*]: 357 |---type: Type of the cooling device(processor/fan/...) 358 |---max_state: Maximum cooling state of the cooling device 359 |---cur_state: Current cooling state of the cooling device 360 |---stats: Directory containing cooling device's statistics 361 |---stats/reset: Writing any value resets the statistics 362 |---stats/time_in_state_ms: Time (msec) spent in various cooling states 363 |---stats/total_trans: Total number of times cooling state is changed 364 |---stats/trans_table: Cooling state transition table 365 366 367Then next two dynamic attributes are created/removed in pairs. They represent 368the relationship between a thermal zone and its associated cooling device. 369They are created/removed for each successful execution of 370thermal_zone_bind_cooling_device/thermal_zone_unbind_cooling_device. 371 372:: 373 374 /sys/class/thermal/thermal_zone[0-*]: 375 |---cdev[0-*]: [0-*]th cooling device in current thermal zone 376 |---cdev[0-*]_trip_point: Trip point that cdev[0-*] is associated with 377 |---cdev[0-*]_weight: Influence of the cooling device in 378 this thermal zone 379 380Besides the thermal zone device sysfs I/F and cooling device sysfs I/F, 381the generic thermal driver also creates a hwmon sysfs I/F for each _type_ 382of thermal zone device. E.g. the generic thermal driver registers one hwmon 383class device and build the associated hwmon sysfs I/F for all the registered 384ACPI thermal zones. 385 386Please read Documentation/ABI/testing/sysfs-class-thermal for thermal 387zone and cooling device attribute details. 388 389:: 390 391 /sys/class/hwmon/hwmon[0-*]: 392 |---name: The type of the thermal zone devices 393 |---temp[1-*]_input: The current temperature of thermal zone [1-*] 394 |---temp[1-*]_critical: The critical trip point of thermal zone [1-*] 395 396Please read Documentation/hwmon/sysfs-interface.rst for additional information. 397 3983. A simple implementation 399========================== 400 401ACPI thermal zone may support multiple trip points like critical, hot, 402passive, active. If an ACPI thermal zone supports critical, passive, 403active[0] and active[1] at the same time, it may register itself as a 404thermal_zone_device (thermal_zone1) with 4 trip points in all. 405It has one processor and one fan, which are both registered as 406thermal_cooling_device. Both are considered to have the same 407effectiveness in cooling the thermal zone. 408 409If the processor is listed in _PSL method, and the fan is listed in _AL0 410method, the sys I/F structure will be built like this:: 411 412 /sys/class/thermal: 413 |thermal_zone1: 414 |---type: acpitz 415 |---temp: 37000 416 |---mode: enabled 417 |---policy: step_wise 418 |---available_policies: step_wise fair_share 419 |---trip_point_0_temp: 100000 420 |---trip_point_0_type: critical 421 |---trip_point_1_temp: 80000 422 |---trip_point_1_type: passive 423 |---trip_point_2_temp: 70000 424 |---trip_point_2_type: active0 425 |---trip_point_3_temp: 60000 426 |---trip_point_3_type: active1 427 |---cdev0: --->/sys/class/thermal/cooling_device0 428 |---cdev0_trip_point: 1 /* cdev0 can be used for passive */ 429 |---cdev0_weight: 1024 430 |---cdev1: --->/sys/class/thermal/cooling_device3 431 |---cdev1_trip_point: 2 /* cdev1 can be used for active[0]*/ 432 |---cdev1_weight: 1024 433 434 |cooling_device0: 435 |---type: Processor 436 |---max_state: 8 437 |---cur_state: 0 438 439 |cooling_device3: 440 |---type: Fan 441 |---max_state: 2 442 |---cur_state: 0 443 444 /sys/class/hwmon: 445 |hwmon0: 446 |---name: acpitz 447 |---temp1_input: 37000 448 |---temp1_crit: 100000 449 4504. Export Symbol APIs 451===================== 452 4534.1. get_tz_trend 454----------------- 455 456This function returns the trend of a thermal zone, i.e the rate of change 457of temperature of the thermal zone. Ideally, the thermal sensor drivers 458are supposed to implement the callback. If they don't, the thermal 459framework calculated the trend by comparing the previous and the current 460temperature values. 461 4624.2. get_thermal_instance 463------------------------- 464 465This function returns the thermal_instance corresponding to a given 466{thermal_zone, cooling_device, trip_point} combination. Returns NULL 467if such an instance does not exist. 468 4694.3. thermal_cdev_update 470------------------------ 471 472This function serves as an arbitrator to set the state of a cooling 473device. It sets the cooling device to the deepest cooling state if 474possible. 475 4765. thermal_emergency_poweroff 477============================= 478 479On an event of critical trip temperature crossing the thermal framework 480shuts down the system by calling hw_protection_shutdown(). The 481hw_protection_shutdown() first attempts to perform an orderly shutdown 482but accepts a delay after which it proceeds doing a forced power-off 483or as last resort an emergency_restart. 484 485The delay should be carefully profiled so as to give adequate time for 486orderly poweroff. 487 488If the delay is set to 0 emergency poweroff will not be supported. So a 489carefully profiled non-zero positive value is a must for emergency 490poweroff to be triggered. 491