關鍵詞:Zone、Cooling、Governor、Step Wise、Fair Share、trip等等。html
Linux Thermal的目的是控制系統運行過程當中採樣點溫度,避免溫度太高形成器件損壞,確保芯片長期穩定工做。node
整個Thermal框架能夠分爲四部分:數據結構
因此Thermal的工做流程是經過Thermal Driver獲取溫度,而後通過Thermal Governor決策,最後經過Thermal Cooling執行溫度控制。app
下面首先從整體詳細分析Thermal框架以及數據結構、API(1. Thermal框架分析),而後分別分析Thermal Driver實例(2. Thermal Driver實例)、Thermal Governor(Step Wise和Fair Share)(3. Thermal Governor分析)、以及Thermal Cooling實例(4. Thermal Cooling實例)。框架
最後將這些內容串起來,分析Thermal是如何控制溫度的。函數
struct thermal_zone_device是對獲取溫度設備的抽象,成員ops是對該Thermal Zone操做的抽象;governor是該Thermal Zone所使用的調溫策略;thermal_instances是該Thermal Zone下的Cooling Device列表。this
struct thermal_zone_device { int id; char type[THERMAL_NAME_LENGTH]; struct device device; struct thermal_attr *trip_temp_attrs; struct thermal_attr *trip_type_attrs; struct thermal_attr *trip_hyst_attrs; void *devdata; int trips;---------------------------------------------------------thermal zone支持的trip數目。 unsigned long trips_disabled; /* bitmap for disabled trips */ int passive_delay; int polling_delay;-------------------------------------------------輪詢讀取溫度的建個,0表示採用中斷形式。 int temperature;---------------------------------------------------當前溫度。 int last_temperature;----------------------------------------------最近一次溫度。 int emul_temperature; int passive; int prev_low_trip; int prev_high_trip; unsigned int forced_passive; atomic_t need_update; struct thermal_zone_device_ops *ops;------------------------------當前thermal zone操做函數集。 struct thermal_zone_params *tzp;----------------------------------當前thermal zone參數。 struct thermal_governor *governor; void *governor_data; struct list_head thermal_instances;-------------------------------當前thermal zone上thermal_instances列表。 struct idr idr; struct mutex lock; struct list_head node; struct delayed_work poll_queue; enum thermal_notify_event notify_event; }; struct thermal_zone_device_ops { int (*bind) (struct thermal_zone_device *, struct thermal_cooling_device *);------------------------將cooling device綁定到thermal zone中,二者經過struct thermal_instances在thermal_zone_bind_cooling_device()中綁定。 int (*unbind) (struct thermal_zone_device *, struct thermal_cooling_device *); int (*get_temp) (struct thermal_zone_device *, int *); int (*set_trips) (struct thermal_zone_device *, int, int); int (*get_mode) (struct thermal_zone_device *, enum thermal_device_mode *); int (*set_mode) (struct thermal_zone_device *, enum thermal_device_mode); int (*get_trip_type) (struct thermal_zone_device *, int, enum thermal_trip_type *); int (*get_trip_temp) (struct thermal_zone_device *, int, int *); int (*set_trip_temp) (struct thermal_zone_device *, int, int); int (*get_trip_hyst) (struct thermal_zone_device *, int, int *); int (*set_trip_hyst) (struct thermal_zone_device *, int, int); int (*get_crit_temp) (struct thermal_zone_device *, int *); int (*set_emul_temp) (struct thermal_zone_device *, int); int (*get_trend) (struct thermal_zone_device *, int, enum thermal_trend *); int (*notify) (struct thermal_zone_device *, int, enum thermal_trip_type); }; struct thermal_bind_params { struct thermal_cooling_device *cdev; int weight; int trip_mask; unsigned long *binding_limits; int (*match) (struct thermal_zone_device *tz, struct thermal_cooling_device *cdev); }; struct thermal_zone_params { char governor_name[THERMAL_NAME_LENGTH]; bool no_hwmon; int num_tbps; /* Number of tbp entries */ struct thermal_bind_params *tbp; ... int slope; int offset; }; struct thermal_zone_of_device_ops { int (*get_temp)(void *, int *); int (*get_trend)(void *, int, enum thermal_trend *); int (*set_trips)(void *, int, int); int (*set_emul_temp)(void *, int); int (*set_trip_temp)(void *, int, int); };
struct thermal_cooling_device是對降溫設備的抽象,對風扇設備就是不一樣的轉速,對CPU、DDR、GPU就是不一樣的電壓或者頻率。atom
struct thermal_cooling_device_ops是Cooling Device操做函數集,其中set_cur_state()是對設備進行溫度控制。spa
struct thermal_cooling_device { int id; char type[THERMAL_NAME_LENGTH]; struct device device; struct device_node *np; void *devdata; const struct thermal_cooling_device_ops *ops; bool updated; /* true if the cooling device does not need update */ struct mutex lock; /* protect thermal_instances list */ struct list_head thermal_instances; struct list_head node; }; struct thermal_cooling_device_ops { int (*get_max_state) (struct thermal_cooling_device *, unsigned long *); int (*get_cur_state) (struct thermal_cooling_device *, unsigned long *); int (*set_cur_state) (struct thermal_cooling_device *, unsigned long); ... };
strcut thermal_governor是對溫控策略的抽象,也就是根據Thermal Zone的trip來選擇Thermal Cooling設備的行爲。好比,溫度越高風扇轉速越快;溫度越高CPU運行在更低電壓和頻率上。debug
struct thermal_governor { char name[THERMAL_NAME_LENGTH]; int (*bind_to_tz)(struct thermal_zone_device *tz);---------------------將一個governor綁定到thermal zone得一個trip上。 void (*unbind_from_tz)(struct thermal_zone_device *tz);----------------將一個governor從thermal zone解綁。 int (*throttle)(struct thermal_zone_device *tz, int trip);-------------根據trip遍歷當前thermal zone下全部的cooling device執行溫控策略。 struct list_head governor_list;-------------------------------------thermal_governor_list上的一個列表元素。 };
全部的策略選擇都是經過throttle()函數進行的,不一樣的Governor的區別也主要在這裏。內核已經實現了Step Wise、User等等,而且還在演進中。
經過struct thermal_instances能夠將thermal zone和thermal cooling設備綁定起來。
struct thermal_instance { int id; char name[THERMAL_NAME_LENGTH]; struct thermal_zone_device *tz;-------------------------------------------綁定的thermal zone。 struct thermal_cooling_device *cdev;--------------------------------------綁定的thermal cooling設備。 int trip;-----------------------------------------------------------------對應的thermal zone的trip。 bool initialized; unsigned long upper; /* Highest cooling state for this trip point */---cooling設備的最高降溫狀態。 unsigned long lower; /* Lowest cooling state for this trip point */----cooling設備最低降溫狀態。 unsigned long target; /* expected cooling state */---------------------cooling設備的當前狀態,也是thermal_cooling_device_ops->set_cur_state()設置後的值。 char attr_name[THERMAL_NAME_LENGTH]; struct device_attribute attr; char weight_attr_name[THERMAL_NAME_LENGTH]; struct device_attribute weight_attr; struct list_head tz_node; /* node in tz->thermal_instances */-------------thermal_zone_device->thermal_instances上的節點。 struct list_head cdev_node; /* node in cdev->thermal_instances */---------thermal_cooling_device->thermal_instances上的節點。 unsigned int weight; /* The weight of the cooling device */ };
thermal_device_mode表示當前的thermal zone是否使能。
thermal_trip_type表示thermal zone的當前trip類型,其中ACTIVE和PASSIVE屬於non-critical類型,交由Governor進行處理;HOT和CRITICAL屬於critical類型,其中CRITICAL會執行orderly_poweroff()。
thermal_trend表示thermal zone的溫度趨勢,是平緩、上升、降低仍是跳躍式的,這就給Governor選擇trip提供依據。
enum thermal_device_mode { THERMAL_DEVICE_DISABLED = 0, THERMAL_DEVICE_ENABLED, }; enum thermal_trip_type { THERMAL_TRIP_ACTIVE = 0, THERMAL_TRIP_PASSIVE, THERMAL_TRIP_HOT, THERMAL_TRIP_CRITICAL, }; enum thermal_trend { THERMAL_TREND_STABLE, /* temperature is stable */-----------------------表示溫度平穩。 THERMAL_TREND_RAISING, /* temperature is raising */---------------------表示當前溫度趨勢是升高的。 THERMAL_TREND_DROPPING, /* temperature is dropping */-------------------表示當前溫度趨勢是下降的。 THERMAL_TREND_RAISE_FULL, /* apply highest cooling action */------------直接應用upper對應的trip。 THERMAL_TREND_DROP_FULL, /* apply lowest cooling action */--------------直接應用lower對應的trip。 }; /* Thermal notification reason */ enum thermal_notify_event { THERMAL_EVENT_UNSPECIFIED, /* Unspecified event */ THERMAL_EVENT_TEMP_SAMPLE, /* New Temperature sample */ THERMAL_TRIP_VIOLATED, /* TRIP Point violation */ THERMAL_TRIP_CHANGED, /* TRIP Point temperature changed */ THERMAL_DEVICE_DOWN, /* Thermal device is down */ THERMAL_DEVICE_UP, /* Thermal device is up after a down event */ THERMAL_DEVICE_POWER_CAPABILITY_CHANGED, /* power capability changed */ };
Thermal core是Thermal Zone、Thermal Cooling、ThermalGovernor的粘合劑。
經過Thermal core提供的API,將這三者相互關聯起來;從Thermal Zone設備獲取溫度,選擇對應的Thermal Governor,Thermal Governor設置Thermal Cooling的狀態,進而達到控制溫度的目的。
經過thermal_zone_device_register()註冊thermal zone設備,建立一系列sysfs節點,而且和governor、cooling進行綁定。
struct thermal_zone_device *thermal_zone_device_register(const char *type, int trips, int mask, void *devdata, struct thermal_zone_device_ops *ops, struct thermal_zone_params *tzp, int passive_delay, int polling_delay) { struct thermal_zone_device *tz; enum thermal_trip_type trip_type; int trip_temp; int result; int count; int passive = 0; struct thermal_governor *governor; if (type && strlen(type) >= THERMAL_NAME_LENGTH) return ERR_PTR(-EINVAL); if (trips > THERMAL_MAX_TRIPS || trips < 0 || mask >> trips) return ERR_PTR(-EINVAL); if (!ops) return ERR_PTR(-EINVAL); if (trips > 0 && (!ops->get_trip_type || !ops->get_trip_temp)) return ERR_PTR(-EINVAL); tz = kzalloc(sizeof(struct thermal_zone_device), GFP_KERNEL); if (!tz) return ERR_PTR(-ENOMEM); INIT_LIST_HEAD(&tz->thermal_instances);------------------------------初始化thermal_instances鏈表,放置struct thermal_instances實例。經過thermal_instances能夠關聯thermal zone和thermal cooling。 idr_init(&tz->idr); mutex_init(&tz->lock); result = get_idr(&thermal_tz_idr, &thermal_idr_lock, &tz->id); if (result) { kfree(tz); return ERR_PTR(result); } strlcpy(tz->type, type ? : "", sizeof(tz->type)); tz->ops = ops; tz->tzp = tzp; tz->device.class = &thermal_class;------------------------------------建立的設備會在/sys/class/thermal下面有個連接。 tz->devdata = devdata; tz->trips = trips; tz->passive_delay = passive_delay; tz->polling_delay = polling_delay; /* A new thermal zone needs to be updated anyway. */ atomic_set(&tz->need_update, 1); dev_set_name(&tz->device, "thermal_zone%d", tz->id); result = device_register(&tz->device);--------------------------------建立/sys/devices/virtual/thermal/thermal_zone*設備。 if (result) { release_idr(&thermal_tz_idr, &thermal_idr_lock, tz->id); kfree(tz); return ERR_PTR(result); } /* sys I/F */---------------------------------------------------------分別建立type、temp、mode、trip等sysfs節點。 if (type) { result = device_create_file(&tz->device, &dev_attr_type); if (result) goto unregister; } ... result = create_trip_attrs(tz, mask);-----------------------爲每一個trip建立trip_point_*_temp/hyst/type節點。 if (result) goto unregister; ... /* Update 'this' zone's governor information */ mutex_lock(&thermal_governor_lock); if (tz->tzp)-------------------------------------------------若是指定thermal zone的governor則經過__find_governor()選定;不然使用默認def_governor。 governor = __find_governor(tz->tzp->governor_name); else governor = def_governor; result = thermal_set_governor(tz, governor);-----------------將governor綁定到tz上,優先使用bind_to_tz()執行綁定;不然直接指定tz->governor爲governor。 if (result) { mutex_unlock(&thermal_governor_lock); goto unregister; } mutex_unlock(&thermal_governor_lock); if (!tz->tzp || !tz->tzp->no_hwmon) { result = thermal_add_hwmon_sysfs(tz); if (result) goto unregister; } mutex_lock(&thermal_list_lock); list_add_tail(&tz->node, &thermal_tz_list);------------------------將當前thermal zone加入到thermal_tz_list列表上。 mutex_unlock(&thermal_list_lock); /* Bind cooling devices for this zone */ bind_tz(tz);-------------------------------------------------------調用tz->ops->bind()將thermal_cdev_list上的cooling設備綁定到tz上。 INIT_DELAYED_WORK(&(tz->poll_queue), thermal_zone_device_check); thermal_zone_device_reset(tz);-------------------------------------對thermal zone的溫度等復位。 /* Update the new thermal zone and mark it as already updated. */ if (atomic_cmpxchg(&tz->need_update, 1, 0)) thermal_zone_device_update(tz, THERMAL_EVENT_UNSPECIFIED); return tz; unregister: release_idr(&thermal_tz_idr, &thermal_idr_lock, tz->id); device_unregister(&tz->device); return ERR_PTR(result); } static int thermal_set_governor(struct thermal_zone_device *tz, struct thermal_governor *new_gov) { int ret = 0; if (tz->governor && tz->governor->unbind_from_tz) tz->governor->unbind_from_tz(tz);------------------------------先調用當前governor進行unbind()。 if (new_gov && new_gov->bind_to_tz) { ret = new_gov->bind_to_tz(tz);---------------------------------使用當前governor進行bind()。 if (ret) { bind_previous_governor(tz, new_gov->name); return ret; } } tz->governor = new_gov;--------------------------------------------更新tz->governor。 return ret; } static void bind_tz(struct thermal_zone_device *tz) { int i, ret; struct thermal_cooling_device *pos = NULL; const struct thermal_zone_params *tzp = tz->tzp; if (!tzp && !tz->ops->bind) return; mutex_lock(&thermal_list_lock); /* If there is ops->bind, try to use ops->bind */ if (tz->ops->bind) { list_for_each_entry(pos, &thermal_cdev_list, node) {-----------遍歷thermal_cdev_list的cooling設備,而後和當前thermal zone進行綁定。 ret = tz->ops->bind(tz, pos); if (ret) print_bind_err_msg(tz, pos, ret); } goto exit; } ... exit: mutex_unlock(&thermal_list_lock); } static void thermal_zone_device_check(struct work_struct *work) { struct thermal_zone_device *tz = container_of(work, struct thermal_zone_device, poll_queue.work); thermal_zone_device_update(tz, THERMAL_EVENT_UNSPECIFIED); }
thermal_zone_device_unregister()則執行相反的操做,將thermal zone從thermal_tz_list上摘除,而且和cooling設備去綁定,以及刪除一系列sysfs節點。
void thermal_zone_device_unregister(struct thermal_zone_device *tz) { int i; const struct thermal_zone_params *tzp; struct thermal_cooling_device *cdev; struct thermal_zone_device *pos = NULL; if (!tz) return; tzp = tz->tzp; mutex_lock(&thermal_list_lock); list_for_each_entry(pos, &thermal_tz_list, node) if (pos == tz) break; if (pos != tz) { /* thermal zone device not found */ mutex_unlock(&thermal_list_lock); return; } list_del(&tz->node); /* Unbind all cdevs associated with 'this' thermal zone */ list_for_each_entry(cdev, &thermal_cdev_list, node) { if (tz->ops->unbind) { tz->ops->unbind(tz, cdev); continue; } ... } ... return; }
thermal_cooling_device_register()建立cooling設備並放入thermal_cdev_list中,以及相關sysfs節點,並將cooling設備和thermal zone綁定。
thermal_cooling_device_unregister()則進行相反的操做。
struct thermal_cooling_device * thermal_cooling_device_register(char *type, void *devdata, const struct thermal_cooling_device_ops *ops) { return __thermal_cooling_device_register(NULL, type, devdata, ops); } static struct thermal_cooling_device * __thermal_cooling_device_register(struct device_node *np, char *type, void *devdata, const struct thermal_cooling_device_ops *ops) { struct thermal_cooling_device *cdev; struct thermal_zone_device *pos = NULL; int result; if (type && strlen(type) >= THERMAL_NAME_LENGTH) return ERR_PTR(-EINVAL); if (!ops || !ops->get_max_state || !ops->get_cur_state || !ops->set_cur_state) return ERR_PTR(-EINVAL); cdev = kzalloc(sizeof(struct thermal_cooling_device), GFP_KERNEL); if (!cdev) return ERR_PTR(-ENOMEM); result = get_idr(&thermal_cdev_idr, &thermal_idr_lock, &cdev->id); if (result) { kfree(cdev); return ERR_PTR(result); } strlcpy(cdev->type, type ? : "", sizeof(cdev->type)); mutex_init(&cdev->lock); INIT_LIST_HEAD(&cdev->thermal_instances); cdev->np = np; cdev->ops = ops; cdev->updated = false; cdev->device.class = &thermal_class;---------------------------------cooling設備一樣會在/sys/class/thermal下建立連接。 cdev->device.groups = cooling_device_attr_groups;--------------------建立cur_state、max_state、type三個sysfs節點。 cdev->devdata = devdata; dev_set_name(&cdev->device, "cooling_device%d", cdev->id); result = device_register(&cdev->device);-----------------------------建立/sys/devices/virtual/thermal/cooling_device*設備節點。 if (result) { release_idr(&thermal_cdev_idr, &thermal_idr_lock, cdev->id); kfree(cdev); return ERR_PTR(result); } /* Add 'this' new cdev to the global cdev list */ mutex_lock(&thermal_list_lock); list_add(&cdev->node, &thermal_cdev_list);---------------------------將設備放入thermal_cdev_list設備鏈表。 mutex_unlock(&thermal_list_lock); /* Update binding information for 'this' new cdev */ bind_cdev(cdev);-----------------------------------------------------遍歷thermal_tz_list,將cdev綁定到上面的thermal zone。 mutex_lock(&thermal_list_lock); list_for_each_entry(pos, &thermal_tz_list, node) if (atomic_cmpxchg(&pos->need_update, 1, 0)) thermal_zone_device_update(pos, THERMAL_EVENT_UNSPECIFIED); mutex_unlock(&thermal_list_lock); return cdev; } void thermal_cooling_device_unregister(struct thermal_cooling_device *cdev) { int i; const struct thermal_zone_params *tzp; struct thermal_zone_device *tz; struct thermal_cooling_device *pos = NULL; if (!cdev) return; mutex_lock(&thermal_list_lock); list_for_each_entry(pos, &thermal_cdev_list, node) if (pos == cdev) break; if (pos != cdev) { /* thermal cooling device not found */ mutex_unlock(&thermal_list_lock); return; } list_del(&cdev->node); /* Unbind all thermal zones associated with 'this' cdev */ list_for_each_entry(tz, &thermal_tz_list, node) { if (tz->ops->unbind) { tz->ops->unbind(tz, cdev); continue; } if (!tz->tzp || !tz->tzp->tbp) continue; tzp = tz->tzp; for (i = 0; i < tzp->num_tbps; i++) { if (tzp->tbp[i].cdev == cdev) { __unbind(tz, tzp->tbp[i].trip_mask, cdev); tzp->tbp[i].cdev = NULL; } } } mutex_unlock(&thermal_list_lock); if (cdev->type[0]) device_remove_file(&cdev->device, &dev_attr_cdev_type); device_remove_file(&cdev->device, &dev_attr_max_state); device_remove_file(&cdev->device, &dev_attr_cur_state); release_idr(&thermal_cdev_idr, &thermal_idr_lock, cdev->id); device_unregister(&cdev->device); return; }
thermal_register_governor()首先判斷thermal_governor_list上是否有同名governor,而後更新thermal_tz_list上未指定governor的thermal zone。
thermal_unregister_governor()則相反,將governor和thermal zone調用unbind_from_tz()並置空;最後從thermal_go上摘除。
int thermal_register_governor(struct thermal_governor *governor) { int err; const char *name; struct thermal_zone_device *pos; if (!governor) return -EINVAL; mutex_lock(&thermal_governor_lock); err = -EBUSY; if (__find_governor(governor->name) == NULL) {--------------------檢查此governor是否已經在thermal_governor_list中,若是不在則加入thermal_governor_list。而且判斷是否爲def_governor。 err = 0; list_add(&governor->governor_list, &thermal_governor_list); if (!def_governor && !strncmp(governor->name, DEFAULT_THERMAL_GOVERNOR, THERMAL_NAME_LENGTH)) def_governor = governor; } mutex_lock(&thermal_list_lock); list_for_each_entry(pos, &thermal_tz_list, node) { if (pos->governor)--------------------------------------------若是thermal zone已經制定governor,則跳過。 continue; name = pos->tzp->governor_name; if (!strncasecmp(name, governor->name, THERMAL_NAME_LENGTH)) { int ret; ret = thermal_set_governor(pos, governor);----------------給當前thermal zone制定governor。 if (ret) dev_err(&pos->device, "Failed to set governor %s for thermal zone %s: %d\n", governor->name, pos->type, ret); } } mutex_unlock(&thermal_list_lock); mutex_unlock(&thermal_governor_lock); return err; } void thermal_unregister_governor(struct thermal_governor *governor) { struct thermal_zone_device *pos; if (!governor) return; mutex_lock(&thermal_governor_lock); if (__find_governor(governor->name) == NULL) goto exit; mutex_lock(&thermal_list_lock); list_for_each_entry(pos, &thermal_tz_list, node) { if (!strncasecmp(pos->governor->name, governor->name, THERMAL_NAME_LENGTH)) thermal_set_governor(pos, NULL); } mutex_unlock(&thermal_list_lock); list_del(&governor->governor_list); exit: mutex_unlock(&thermal_governor_lock); return; }
thermal_zone_bind_cooling_device()經過建立thermal_instances設備將Thermal Zone和Thermal Cooling綁定,這樣Thermal Zone就能夠根據溫度處理Thermal Cooling設備。
thermal_zone_unbind_cooling_device() 則將關聯Thermal Zone和Thermal Cooling的thermal_instances從二者的鏈表上摘除。
int thermal_zone_bind_cooling_device(struct thermal_zone_device *tz, int trip, struct thermal_cooling_device *cdev, unsigned long upper, unsigned long lower, unsigned int weight) { struct thermal_instance *dev; struct thermal_instance *pos; struct thermal_zone_device *pos1; struct thermal_cooling_device *pos2; unsigned long max_state; int result, ret; if (trip >= tz->trips || (trip < 0 && trip != THERMAL_TRIPS_NONE)) return -EINVAL; list_for_each_entry(pos1, &thermal_tz_list, node) { if (pos1 == tz) break; } list_for_each_entry(pos2, &thermal_cdev_list, node) { if (pos2 == cdev) break; } if (tz != pos1 || cdev != pos2) return -EINVAL; ret = cdev->ops->get_max_state(cdev, &max_state);----------------------從Cooling設備操做函數get_max_state()獲取max_state,進而決定thermal_instances的lower和upper範圍。 if (ret) return ret; /* lower default 0, upper default max_state */ lower = lower == THERMAL_NO_LIMIT ? 0 : lower; upper = upper == THERMAL_NO_LIMIT ? max_state : upper; if (lower > upper || upper > max_state) return -EINVAL; dev = kzalloc(sizeof(struct thermal_instance), GFP_KERNEL); if (!dev) return -ENOMEM; dev->tz = tz; dev->cdev = cdev; dev->trip = trip; dev->upper = upper; dev->lower = lower; dev->target = THERMAL_NO_TARGET; dev->weight = weight; result = get_idr(&tz->idr, &tz->lock, &dev->id); if (result) goto free_mem; sprintf(dev->name, "cdev%d", dev->id); result = sysfs_create_link(&tz->device.kobj, &cdev->device.kobj, dev->name);------cdevx鏈接到cooling_devicex。 if (result) goto release_idr; sprintf(dev->attr_name, "cdev%d_trip_point", dev->id);-----------------------建立cdevx_trip_point和cdevx_weight節點。 sysfs_attr_init(&dev->attr.attr); dev->attr.attr.name = dev->attr_name; dev->attr.attr.mode = 0444; dev->attr.show = thermal_cooling_device_trip_point_show; result = device_create_file(&tz->device, &dev->attr); if (result) goto remove_symbol_link; sprintf(dev->weight_attr_name, "cdev%d_weight", dev->id); sysfs_attr_init(&dev->weight_attr.attr); dev->weight_attr.attr.name = dev->weight_attr_name; dev->weight_attr.attr.mode = S_IWUSR | S_IRUGO; dev->weight_attr.show = thermal_cooling_device_weight_show; dev->weight_attr.store = thermal_cooling_device_weight_store; result = device_create_file(&tz->device, &dev->weight_attr); if (result) goto remove_trip_file;... } int thermal_zone_unbind_cooling_device(struct thermal_zone_device *tz, int trip, struct thermal_cooling_device *cdev) { struct thermal_instance *pos, *next; mutex_lock(&tz->lock); mutex_lock(&cdev->lock); list_for_each_entry_safe(pos, next, &tz->thermal_instances, tz_node) { if (pos->tz == tz && pos->trip == trip && pos->cdev == cdev) { list_del(&pos->tz_node); list_del(&pos->cdev_node); mutex_unlock(&cdev->lock); mutex_unlock(&tz->lock); goto unbind; } } mutex_unlock(&cdev->lock); mutex_unlock(&tz->lock); return -ENODEV; unbind: device_remove_file(&tz->device, &pos->weight_attr); device_remove_file(&tz->device, &pos->attr); sysfs_remove_link(&tz->device.kobj, pos->name); release_idr(&tz->idr, &tz->lock, pos->id); kfree(pos); return 0; }
thermal_zone_device_update()通常由Thermal驅動調用,有多是polling或者中斷觸發。
而後更新當前Thermal Zone的溫度,最後根據溫度值經過handle_thermal_trip()進行處理。
monitor_thermal_zone()根據passive和polling的設置決定是否啓動thermal_zone_device->pool_queue這個delayed_work。
整個polling流程由thermal_zone_device_update()觸發,依次流程爲:handle_thermal_trip()中啓動monitor_thermal_zone(),monitor_thermal_zone()中調用mod_delayed_work()進行poll_queue延時值的更新。若是thermal zone有多個trip,poll_queue延時值可能被屢次更新。poll_queue放入system_freezable_wq後,達到時間後調用thermal_zone_device_check(),進而調用thermal_zone_device_update()完成周期性循環。
void thermal_zone_device_update(struct thermal_zone_device *tz, enum thermal_notify_event event) { int count; if (atomic_read(&in_suspend)) return; if (!tz->ops->get_temp) return; update_temperature(tz); thermal_zone_set_trips(tz); tz->notify_event = event; for (count = 0; count < tz->trips; count++) handle_thermal_trip(tz, count); } static void update_temperature(struct thermal_zone_device *tz) { int temp, ret; ret = thermal_zone_get_temp(tz, &temp); if (ret) { if (ret != -EAGAIN) dev_warn(&tz->device, "failed to read out thermal zone (%d)\n", ret); return; } mutex_lock(&tz->lock); tz->last_temperature = tz->temperature; tz->temperature = temp; mutex_unlock(&tz->lock); trace_thermal_temperature(tz); if (tz->last_temperature == THERMAL_TEMP_INVALID) dev_dbg(&tz->device, "last_temperature N/A, current_temperature=%d\n", tz->temperature); else dev_dbg(&tz->device, "last_temperature=%d, current_temperature=%d\n", tz->last_temperature, tz->temperature); } void thermal_zone_set_trips(struct thermal_zone_device *tz) { int low = -INT_MAX; int high = INT_MAX; int trip_temp, hysteresis; int i, ret; mutex_lock(&tz->lock); if (!tz->ops->set_trips || !tz->ops->get_trip_hyst) goto exit; for (i = 0; i < tz->trips; i++) { int trip_low; tz->ops->get_trip_temp(tz, i, &trip_temp); tz->ops->get_trip_hyst(tz, i, &hysteresis); trip_low = trip_temp - hysteresis; if (trip_low < tz->temperature && trip_low > low) low = trip_low; if (trip_temp > tz->temperature && trip_temp < high) high = trip_temp; } /* No need to change trip points */ if (tz->prev_low_trip == low && tz->prev_high_trip == high) goto exit; tz->prev_low_trip = low; tz->prev_high_trip = high; dev_dbg(&tz->device, "new temperature boundaries: %d < x < %d\n", low, high); ret = tz->ops->set_trips(tz, low, high); if (ret) dev_err(&tz->device, "Failed to set trips: %d\n", ret); exit: mutex_unlock(&tz->lock); } static void handle_thermal_trip(struct thermal_zone_device *tz, int trip) { enum thermal_trip_type type; /* Ignore disabled trip points */ if (test_bit(trip, &tz->trips_disabled)) return; tz->ops->get_trip_type(tz, trip, &type); if (type == THERMAL_TRIP_CRITICAL || type == THERMAL_TRIP_HOT) handle_critical_trips(tz, trip, type); else handle_non_critical_trips(tz, trip, type); /* * Alright, we handled this trip successfully. * So, start monitoring again. */ monitor_thermal_zone(tz); } static void handle_critical_trips(struct thermal_zone_device *tz, int trip, enum thermal_trip_type trip_type) { int trip_temp; tz->ops->get_trip_temp(tz, trip, &trip_temp); /* If we have not crossed the trip_temp, we do not care. */ if (trip_temp <= 0 || tz->temperature < trip_temp) return; trace_thermal_zone_trip(tz, trip, trip_type); if (tz->ops->notify) tz->ops->notify(tz, trip, trip_type); if (trip_type == THERMAL_TRIP_CRITICAL) { dev_emerg(&tz->device, "critical temperature reached(%d C),shutting down\n", tz->temperature / 1000); orderly_poweroff(true); } } static void handle_non_critical_trips(struct thermal_zone_device *tz, int trip, enum thermal_trip_type trip_type) { tz->governor ? tz->governor->throttle(tz, trip) : def_governor->throttle(tz, trip); } static void monitor_thermal_zone(struct thermal_zone_device *tz) { mutex_lock(&tz->lock); if (tz->passive)-----------------------------------分別設置passive和polling兩種延時工做。 thermal_zone_device_set_polling(tz, tz->passive_delay); else if (tz->polling_delay) thermal_zone_device_set_polling(tz, tz->polling_delay); else thermal_zone_device_set_polling(tz, 0); mutex_unlock(&tz->lock); } static void thermal_zone_device_set_polling(struct thermal_zone_device *tz, int delay) { if (delay > 1000)----------------------------------將poll_queue放入system_freezable_wq工做隊列上,屢次調用mod_delayed_work()在超時前只有最後一次生效。 mod_delayed_work(system_freezable_wq, &tz->poll_queue, round_jiffies(msecs_to_jiffies(delay))); else if (delay) mod_delayed_work(system_freezable_wq, &tz->poll_queue, msecs_to_jiffies(delay)); else cancel_delayed_work(&tz->poll_queue);----------若是delay爲0,則取消poll_queue延時工做。 }
thermal_cdev_update()是由Governor調用進行cooling device設置。
void thermal_cdev_update(struct thermal_cooling_device *cdev) { struct thermal_instance *instance; unsigned long target = 0; mutex_lock(&cdev->lock); /* cooling device is updated*/ if (cdev->updated) { mutex_unlock(&cdev->lock); return; } /* Make sure cdev enters the deepest cooling state */ list_for_each_entry(instance, &cdev->thermal_instances, cdev_node) {----------遍歷當前cooling device上全部的thermal zone。 dev_dbg(&cdev->device, "zone%d->target=%lu\n", instance->tz->id, instance->target); if (instance->target == THERMAL_NO_TARGET) continue; if (instance->target > target) target = instance->target;---------------------------------------------確保cooling設備選擇最高cooling狀態,而後調用cooling設備的set_cur_state()進行降溫。 } cdev->ops->set_cur_state(cdev, target); cdev->updated = true; mutex_unlock(&cdev->lock); trace_cdev_update(cdev, target); dev_dbg(&cdev->device, "set to state %lu\n", target); }
thermal_init()在內核fs_initcall()階段調用,進行governor、thermal_class、Generic Netlink註冊等操做。
static int __init thermal_init(void) { int result; result = thermal_register_governors();---------------註冊平臺支持的全部governor。 if (result) goto error; result = class_register(&thermal_class);-------------註冊thermal_class。 if (result) goto unregister_governors; result = genetlink_init();---------------------------註冊Generic Netlink。 if (result) goto unregister_class;... return result; } static void __exit thermal_exit(void) { unregister_pm_notifier(&thermal_pm_nb); of_thermal_destroy_zones(); genetlink_exit(); class_unregister(&thermal_class); thermal_unregister_governors(); ... } fs_initcall(thermal_init); module_exit(thermal_exit);
下面首先簡單看一下Temp Sensor的硬件,而後分析DTS,最後分析驅動的實現。
對Temp Sensor的配置能夠經過APB BUS進行,包括兩個Temp Sensor,每一個Temp Sensor中包括3個Trip觸發點設置,以及一個Alarm配置。
Trip達到後會觸發中斷,CPU的INTC收到中斷後,進行中斷處理;Alarm達到後直接致使CPU復位或者關閉PLL。
Temp Sensor默認使用32K時鐘,每32768個時鐘採樣一次。還能夠根據狀況選擇24M做爲時鐘輸入。
一個重要工做就是根據實際狀況,選定Trip溫度以及Alarm溫度。
另外一個核心的工做就是肯定如何根據Data寄存器的值計算出溫度值。這就須要計算兩個參數A和B。
經過其餘測量手段讀出溫度值,以及當前溫度值下的DBN<11:0>。這獲取一系列數據以後,經過直線數據擬合,得出A和B的值。
DTS是對硬件的抽象,包括寄存器配置地址和範圍、中斷、3個trip溫度、一個alarm溫度。
sensor0: sensor0@0xfc20a000 { compatible = "vsi,dp1000-thermal"; reg = <0xfc20a000 0x20>; interrupts = <38>; vsi,temp0 = <90>; vsi,temp1 = <95>; vsi,temp2 = <100>; vsi,alarm_temp = <120>; vsi,alarm_en; };
從DTS能夠看出,經過配置不一樣trip和alarm的溫度,中斷觸發後,CPU會讀取溫度進行相應處理。
Temp Sensor的驅動首先解析DTS,並進行iomem映射;而後註冊中斷以及下半部workqueue處理;再進行硬件設置;最後註冊thermal zone設備。
在設備正常工做中,根據配置的trip和alarm值觸發中斷,而後進行work處理;中間會用到struct thermal_zone_device_ops提供的成員函數獲取溫度、和cooling設備綁定等等操做。
static int dp1000_thermal_probe(struct platform_device *pdev) { struct dp1000_thermal_priv *priv; struct resource *res; int ret; priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); if (!priv) return -ENOMEM; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); priv->reg_base = devm_ioremap_resource(&pdev->dev, res); if (IS_ERR(priv->reg_base)) return PTR_ERR(priv->reg_base); if (device_property_read_u32(&pdev->dev, "vsi,temp0", &priv->temp0) < 0) { dev_dbg(&pdev->dev, "\"temp0\" property is missing, using default value.\n"); priv->temp0 = 0; } ... INIT_DELAYED_WORK(&priv->work, dp1000_thermal_work);---------------------建立delayed_work,對應的處理函數是dp1000_thermal_work()。 priv->irq = platform_get_irq(pdev, 0); ... ret = devm_request_irq(&pdev->dev, priv->irq, dp1000_thermal_interrupt, 0, dev_name(&pdev->dev), priv);-------------------------------irq註冊,中斷處理函數爲dp1000_thermal_interrupt()。 ... dp1000_init_thermal(priv);------------------------------------------------硬件初始化。 priv->zone = thermal_zone_device_register("dp1000_thermal", DP1000_THERMAL_TRIPS, 0, priv, &dp1000_thermal_zone_ops, NULL, 0, 0);-------Thermal Zone註冊。 #ifdef DP1000_THERMAL_STUB if(priv->zone->id == 0) thermal_zone_0 = priv->zone; else if(priv->zone->id == 1) thermal_zone_1 = priv->zone; #endif ... return 0; } static int dp1000_thermal_remove(struct platform_device *pdev) { struct dp1000_thermal_priv *priv = dev_get_drvdata(&pdev->dev); ... return 0; } static const struct of_device_id dp1000_thermal_id_table[] = { { .compatible = "vsi,dp1000-thermal" },-----------------------------------和dts匹配。 { } }; MODULE_DEVICE_TABLE(of, dp1000_thermal_id_table); static struct platform_driver dp1000_thermal_driver = { .probe = dp1000_thermal_probe, .remove = dp1000_thermal_remove, .driver = { .name = "dp1000_thermal", .of_match_table = dp1000_thermal_id_table, }, }; module_platform_driver(dp1000_thermal_driver);
dp1000_thermal_zone_ops函數集是Thermal Sensor的核心,對Thermal Zone的操做都是經過調用這些函數實現的。
static struct thermal_zone_device_ops dp1000_thermal_zone_ops = { .bind = dp1000_thermal_bind, .unbind = dp1000_thermal_unbind, .get_trip_type = dp1000_thermal_get_trip_type, .get_trip_hyst = dp1000_thermal_get_trip_hyst, .get_temp = dp1000_thermal_get_temp, .set_trip_temp = dp1000_thermal_set_trip_temp, .get_trip_temp = dp1000_thermal_get_trip_temp, .get_crit_temp = dp1000_thermal_get_crit_temp, };
當Thermal Sensor的溫度達到trip值時,會觸發中斷。
而後進入dp1000_thermal_interrupt(),在延時300ms進行dp1000_thermal_work()處理。
static irqreturn_t dp1000_thermal_interrupt(int irq, void *id) { struct dp1000_thermal_priv *priv = (struct dp1000_thermal_priv *)id; unsigned int status; if (status == 0) return IRQ_NONE; else { schedule_delayed_work(&priv->work, msecs_to_jiffies(300)); } return IRQ_HANDLED; } static void dp1000_thermal_work(struct work_struct *work) { struct dp1000_thermal_priv *priv; priv = container_of(work, struct dp1000_thermal_priv, work.work); thermal_zone_device_update(priv->zone, THERMAL_EVENT_UNSPECIFIED); }
最終的工做交給thermal_zone_device_update()進行,讀取溫度,根據溫度選擇trip。
下面簡單分析兩個Governor:Step Wise和Fair Share。
首先看一下Step Wise的註冊。
static struct thermal_governor thermal_gov_step_wise = { .name = "step_wise", .throttle = step_wise_throttle, }; int thermal_gov_step_wise_register(void) { return thermal_register_governor(&thermal_gov_step_wise); } void thermal_gov_step_wise_unregister(void) { thermal_unregister_governor(&thermal_gov_step_wise); }
在handle_non_critical_trips()中,首先選用當前thermal zone的throttle()進行處理。
對於Step Wise governor來講,對外的接口只有step_wise_throttle()。
static int step_wise_throttle(struct thermal_zone_device *tz, int trip) { struct thermal_instance *instance; thermal_zone_trip_update(tz, trip);-----------------------------根據當前溫度和上次溫度對比,獲得溫度趨勢;而後根據溫度趨勢得出Cooling設備對應的state。 if (tz->forced_passive) thermal_zone_trip_update(tz, THERMAL_TRIPS_NONE); mutex_lock(&tz->lock); list_for_each_entry(instance, &tz->thermal_instances, tz_node) thermal_cdev_update(instance->cdev);------------------------遍歷cdev->thermal_instances選擇最深的cooling狀態。而後調用cdev->ops->set_cur_state()中。 mutex_unlock(&tz->lock); return 0; } static void thermal_zone_trip_update(struct thermal_zone_device *tz, int trip) { int trip_temp; enum thermal_trip_type trip_type; enum thermal_trend trend; struct thermal_instance *instance; bool throttle = false; int old_target; if (trip == THERMAL_TRIPS_NONE) { trip_temp = tz->forced_passive; trip_type = THERMAL_TRIPS_NONE; } else { tz->ops->get_trip_temp(tz, trip, &trip_temp); tz->ops->get_trip_type(tz, trip, &trip_type); } trend = get_tz_trend(tz, trip);----------------------------------------根據當前溫度tz->temperature和tz->last_temperature對比,斷定tend是STABLE/RAISING/DROPPING等中的一種。 if (tz->temperature >= trip_temp) { throttle = true;---------------------------------------------------throttle爲true表示須要節流,即降溫。 trace_thermal_zone_trip(tz, trip, trip_type); } mutex_lock(&tz->lock); list_for_each_entry(instance, &tz->thermal_instances, tz_node) { if (instance->trip != trip)----------------------------------------相同trip不作處理。 continue; old_target = instance->target; instance->target = get_target_state(instance, trend, throttle);----instance->target是將要設置到Cooling設備的狀態。 if (instance->initialized && old_target == instance->target) continue; /* Activate a passive thermal instance */ if (old_target == THERMAL_NO_TARGET && instance->target != THERMAL_NO_TARGET) update_passive_instance(tz, trip_type, 1); /* Deactivate a passive thermal instance */ else if (old_target != THERMAL_NO_TARGET && instance->target == THERMAL_NO_TARGET) update_passive_instance(tz, trip_type, -1); instance->initialized = true; mutex_lock(&instance->cdev->lock); instance->cdev->updated = false;------------------------------------updated爲false表示Cooling設備須要更新狀態,在thermal_cdev_update()中會進行判斷。 mutex_unlock(&instance->cdev->lock); } mutex_unlock(&tz->lock); } static unsigned long get_target_state(struct thermal_instance *instance, enum thermal_trend trend, bool throttle) { struct thermal_cooling_device *cdev = instance->cdev; unsigned long cur_state; unsigned long next_target; cdev->ops->get_cur_state(cdev, &cur_state); next_target = instance->target; dev_dbg(&cdev->device, "cur_state=%ld\n", cur_state); if (!instance->initialized) { if (throttle) { next_target = (cur_state + 1) >= instance->upper ? instance->upper : ((cur_state + 1) < instance->lower ? instance->lower : (cur_state + 1)); } else { next_target = THERMAL_NO_TARGET; } return next_target; } switch (trend) { case THERMAL_TREND_RAISING:------------------------------------升溫狀態下,next_target爲cur_state+1,可是不超過instance->upper。 if (throttle) { next_target = cur_state < instance->upper ? (cur_state + 1) : instance->upper; if (next_target < instance->lower) next_target = instance->lower; } break; case THERMAL_TREND_RAISE_FULL: if (throttle) next_target = instance->upper; break; case THERMAL_TREND_DROPPING:------------------------------------降溫狀態下,next_target爲cur_state-1,但不低於instance->lower。存在特殊狀況爲THERMAL_NO_TARGET。 if (cur_state <= instance->lower) { if (!throttle) next_target = THERMAL_NO_TARGET; } else { next_target = cur_state - 1; if (next_target > instance->upper) next_target = instance->upper; } break; case THERMAL_TREND_DROP_FULL: if (cur_state == instance->lower) { if (!throttle) next_target = THERMAL_NO_TARGET; } else next_target = instance->lower; break; default:--------------------------------------------------------stable狀態,不改變target值。 break; } return next_target; } static void update_passive_instance(struct thermal_zone_device *tz, enum thermal_trip_type type, int value) { if (type == THERMAL_TRIP_PASSIVE || type == THERMAL_TRIPS_NONE) tz->passive += value; }
Step Wise在中斷觸發後根據溫度的變化趨勢選擇Cooling狀態。而不是根據trip值選擇Cooling狀態。
* If the temperature is higher than a trip point, * a. if the trend is THERMAL_TREND_RAISING, use higher cooling * state for this trip point * b. if the trend is THERMAL_TREND_DROPPING, use lower cooling * state for this trip point * c. if the trend is THERMAL_TREND_RAISE_FULL, use upper limit * for this trip point * d. if the trend is THERMAL_TREND_DROP_FULL, use lower limit * for this trip point * If the temperature is lower than a trip point, * a. if the trend is THERMAL_TREND_RAISING, do nothing * b. if the trend is THERMAL_TREND_DROPPING, use lower cooling * state for this trip point, if the cooling state already * equals lower limit, deactivate the thermal instance * c. if the trend is THERMAL_TREND_RAISE_FULL, do nothing * d. if the trend is THERMAL_TREND_DROP_FULL, use lower limit, * if the cooling state already equals lower limit, * deactivate the thermal instance
如上是step_wise.c中關於Step Wise governor的溫控策略。分別對高於或低於trip溫度下不一樣趨勢行爲作出瞭解釋。
FairShare引入了weight概念。若是一個thermal zone中存在多個Cooling設備,不一樣的設備降溫效果可能不一樣,用weight表示降溫的能力。
weight大的設備得分較高,所以能夠選擇更深的Cooling狀態。
static struct thermal_governor thermal_gov_fair_share = { .name = "fair_share", .throttle = fair_share_throttle, }; int thermal_gov_fair_share_register(void) { return thermal_register_governor(&thermal_gov_fair_share); } void thermal_gov_fair_share_unregister(void) { thermal_unregister_governor(&thermal_gov_fair_share); }
fair_share_throttle()首先根據溫度得出當前trip等級,而後綜合不一樣Cooling的weight等計算出每一個Cooling設備的target。
static int fair_share_throttle(struct thermal_zone_device *tz, int trip) { struct thermal_instance *instance; int total_weight = 0; int total_instance = 0; int cur_trip_level = get_trip_level(tz);------------------------------------根據溫度獲取對應trip等級。 list_for_each_entry(instance, &tz->thermal_instances, tz_node) { if (instance->trip != trip) continue; total_weight += instance->weight; total_instance++; } list_for_each_entry(instance, &tz->thermal_instances, tz_node) { int percentage; struct thermal_cooling_device *cdev = instance->cdev; if (instance->trip != trip) continue; if (!total_weight) percentage = 100 / total_instance;----------------------------------在都沒有定義weight的狀況下,每一個Cooling設備一樣percentage。 else percentage = (instance->weight * 100) / total_weight;---------------若是存在weight的狀況下,根據權重來劃分percentage。 instance->target = get_target_state(tz, cdev, percentage, cur_trip_level);------------------------------------獲取當前Cooling設備對應的state。 mutex_lock(&instance->cdev->lock); instance->cdev->updated = false; mutex_unlock(&instance->cdev->lock); thermal_cdev_update(cdev); } return 0; } static int get_trip_level(struct thermal_zone_device *tz) { int count = 0; int trip_temp; enum thermal_trip_type trip_type; if (tz->trips == 0 || !tz->ops->get_trip_temp) return 0; for (count = 0; count < tz->trips; count++) { tz->ops->get_trip_temp(tz, count, &trip_temp); if (tz->temperature < trip_temp)----------------------------------------根據thermal zone的溫度值,選擇合適的trip等級。 break; } if (count > 0) { tz->ops->get_trip_type(tz, count - 1, &trip_type);----------------------僅是更新thermal trace point。 trace_thermal_zone_trip(tz, count - 1, trip_type); } return count; } static long get_target_state(struct thermal_zone_device *tz, struct thermal_cooling_device *cdev, int percentage, int level) { unsigned long max_state; cdev->ops->get_max_state(cdev, &max_state); return (long)(percentage * level * max_state) / (100 * tz->trips);-------------- }
在weight爲0的狀況下,不一樣Cooling設備state均等映射到trip。在只有一個Cooling設備狀況下,若是Cooling最大狀態和ThermalZone trip最大值相等,怎能夠trip和狀態一一對應。
* Parameters used for Throttling: * P1. max_state: Maximum throttle state exposed by the cooling device. * P2. percentage[i]/100: * How 'effective' the 'i'th device is, in cooling the given zone. * P3. cur_trip_level/max_no_of_trips: * This describes the extent to which the devices should be throttled. * We do not want to throttle too much when we trip a lower temperature, * whereas the throttling is at full swing if we trip critical levels. * (Heavily assumes the trip points are in ascending order) * new_state of cooling device = P3 * P2 * P1
fair_share.c中給出了計算Cooling設備狀態的計算公式,new_state=percentage*cur_trip_level*max_state/(100*max_no_of_trips)。
建立一個Dummy Cooling驅動表示Cooling設備,經過thermal_cooling_device_register()註冊Thermal Cooling設備,將其和Thermal Zone綁定。在Thermal Zone中斷出發後,經過Governor選擇state,而後經過set_cur_state()執行溫控操做。
/* bind to generic thermal layer as cooling device*/ static struct thermal_cooling_device_ops dummy_cooling_ops = { .get_max_state = dummy_cooling_get_max_state,---------------------------Cooling設備最深降溫狀態。 .get_cur_state = dummy_cooling_get_cur_state,---------------------------當前Cooling狀態。 .set_cur_state = dummy_cooling_set_cur_state,---------------------------根據狀態,執行溫控操做。 }; static int __init dummy_cooling_init(void) { int retval; dummy_cooling_dev = thermal_cooling_device_register("dummy_cooling", NULL, &dummy_cooling_ops); if (IS_ERR(dummy_cooling_dev)) { retval = -ENODEV; } return retval; } module_init(dummy_cooling_init); static void __exit dummy_cooling_exit(void) { thermal_cooling_device_unregister(dummy_cooling_dev); } module_exit(dummy_cooling_exit);
首先使能已有Thermal調試手段,並添加proc節點模擬中斷觸發;而後基於log分析Thermal流程。
對Termal的調試能夠有兩種方式:
爲了模擬溫度變化,添加proc節點,而後使用腳本模擬溫度觸發流程。
#ifdef DP1000_THERMAL_STUB static int dp1000_temp_stub = 0; struct thermal_zone_device *thermal_zone_0, *thermal_zone_1; struct proc_dir_entry *dp1000_temp_proc = NULL; #endif static int dp1000_thermal_get_temp(struct thermal_zone_device *zone, int *temp) { #ifdef DP1000_THERMAL_STUB *temp = dp1000_temp_stub;--------------------------------------------------------替代從寄存器獲取溫度流程,使用/proc/dp1000_temp_stub輸入的溫度值。 #else ... #endif return 0; } #ifdef DP1000_THERMAL_STUB static int dp1000_temp_stub_proc_show(struct seq_file *m, void *v) { seq_printf(m, "%d\n", dp1000_temp_stub); return 0; } static int dp1000_temp_stub_proc_open(struct inode *inode, struct file *file) { return single_open(file, dp1000_temp_stub_proc_show, NULL); } static ssize_t dp1000_temp_stub_proc_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos) { int rc; rc = kstrtoint_from_user(buffer, count, 0, &dp1000_temp_stub); if (rc) return rc; thermal_zone_device_update(thermal_zone_0, THERMAL_EVENT_UNSPECIFIED);------------對/proc/dp1000_temp_stub寫入溫度,觸發流程。模擬中斷觸發流程。 // thermal_zone_device_update(thermal_zone_1, THERMAL_EVENT_UNSPECIFIED); return count; } static const struct file_operations dp1000_temp_stub_proc_fops = { .open = dp1000_temp_stub_proc_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, .write = dp1000_temp_stub_proc_write, }; #endif static void dp1000_init_thermal(struct dp1000_thermal_priv *priv) { ... #ifdef DP1000_THERMAL_STUB if(!dp1000_temp_proc) dp1000_temp_proc = proc_create("dp1000_temp_stub", 0, NULL, &dp1000_temp_stub_proc_fops);------------建立/proc/dp1000_temp_stub節點。 #endif ... } static int dp1000_thermal_probe(struct platform_device *pdev) { ... priv->zone = thermal_zone_device_register("dp1000_thermal", DP1000_THERMAL_TRIPS, 0, priv, &dp1000_thermal_zone_ops, NULL, 0, 0); #ifdef DP1000_THERMAL_STUB if(priv->zone->id == 0) thermal_zone_0 = priv->zone; else if(priv->zone->id == 1) thermal_zone_1 = priv->zone; #endif ... }
使用以下腳本進行Thermal流程調試:
echo 0 > /sys/kernel/debug/tracing/events/enable echo 1 > /sys/kernel/debug/tracing/events/thermal/enable echo > /sys/kernel/debug/tracing/trace for i in 89 95 100 95 90 95 100 do echo -e "\n" echo $i > /proc/dp1000_temp_stub sleep 1 done cat /sys/kernel/debug/tracing/trace
獲得的結果以下:
[ 35.900013] thermal thermal_zone0: last_temperature=0, current_temperature=89 [ 35.907277] thermal thermal_zone0: Trip0[type=0,temp=89]:trend=1,throttle=1--------------89度達到trip0的觸發溫度,並且是升溫狀態。trip1和trip2都沒有throttle。 [ 35.914290] thermal cooling_device0: cur_state=0 [ 35.918933] thermal cooling_device0: old_target=-1, target=1 [ 35.924619] thermal cooling_device0: zone0->target=1 [ 35.929608] thermal cooling_device0: zone1->target=4294967295 [ 35.935383] thermal cooling_device0: set to state 1--------------------------------------Cooling設備當前狀態時0,因此要將狀態設置爲1。 [ 35.940293] thermal thermal_zone0: Trip1[type=0,temp=94]:trend=1,throttle=0 [ 35.947286] thermal thermal_zone0: Trip2[type=0,temp=99]:trend=1,throttle=0 [ 36.999977] thermal thermal_zone0: last_temperature=89, current_temperature=95 [ 37.007326] thermal thermal_zone0: Trip0[type=0,temp=89]:trend=1,throttle=1 [ 37.014332] thermal cooling_device0: cur_state=1 [ 37.018973] thermal cooling_device0: old_target=1, target=2 [ 37.024570] thermal cooling_device0: zone0->target=2 [ 37.029558] thermal cooling_device0: zone1->target=4294967295 [ 37.035336] thermal cooling_device0: set to state 2--------------------------------------95度是trip1的觸發溫度,因此Cooling狀態從當前的1設置到2。trip2沒有throttle。 [ 37.040248] thermal thermal_zone0: Trip1[type=0,temp=94]:trend=1,throttle=1 [ 37.047240] thermal thermal_zone0: Trip2[type=0,temp=99]:trend=1,throttle=0
...
從thermal trace能夠纔看出,首先獲取溫度,而後選擇trip,最後設置Cooling設備。
# TASK-PID CPU# |||| TIMESTAMP FUNCTION # | | | |||| | | sh-156 [000] .... 35.899971: thermal_temperature: thermal_zone=dp1000_thermal id=0 temp_prev=0 temp=89 sh-156 [000] .... 35.907265: thermal_zone_trip: thermal_zone=dp1000_thermal id=0 trip=0 trip_type=ACTIVE sh-156 [000] .n.. 35.935374: cdev_update: type=dummy_cooling target=1 sh-156 [000] .... 36.999933: thermal_temperature: thermal_zone=dp1000_thermal id=0 temp_prev=89 temp=95 sh-156 [000] .n.. 37.007312: thermal_zone_trip: thermal_zone=dp1000_thermal id=0 trip=0 trip_type=ACTIVE sh-156 [000] .n.. 37.035327: cdev_update: type=dummy_cooling target=2 sh-156 [000] .n.. 37.040238: thermal_zone_trip: thermal_zone=dp1000_thermal id=0 trip=1 trip_type=ACTIVE sh-156 [000] .... 38.079912: thermal_temperature: thermal_zone=dp1000_thermal id=0 temp_prev=95 temp=100 sh-156 [000] .n.. 38.087374: thermal_zone_trip: thermal_zone=dp1000_thermal id=0 trip=0 trip_type=ACTIVE sh-156 [000] .n.. 38.115385: cdev_update: type=dummy_cooling target=3 ...
Thermal Framework一共能夠分爲四部分,Thermal Core、Thermal Zone、Thermal Governor、Thermal Cooling。
其中Core很穩定,主要是會使用;Governor也比較穩定,已有的Governor能覆蓋大部分場景;須要開發的主要有Thermal Zone的Driver和降溫設備Cooling。
在開發過程當中,能夠藉助Trace point等措施進行問題定位。