Commit | Line | Data |
---|---|---|
62052ab1 | 1 | Runtime Power Management Framework for I/O Devices |
5e928f77 | 2 | |
9659cc06 | 3 | (C) 2009-2011 Rafael J. Wysocki <rjw@sisk.pl>, Novell Inc. |
7490e442 | 4 | (C) 2010 Alan Stern <stern@rowland.harvard.edu> |
f71495f3 | 5 | (C) 2014 Intel Corp., Rafael J. Wysocki <rafael.j.wysocki@intel.com> |
5e928f77 RW |
6 | |
7 | 1. Introduction | |
8 | ||
62052ab1 | 9 | Support for runtime power management (runtime PM) of I/O devices is provided |
5e928f77 RW |
10 | at the power management core (PM core) level by means of: |
11 | ||
12 | * The power management workqueue pm_wq in which bus types and device drivers can | |
13 | put their PM-related work items. It is strongly recommended that pm_wq be | |
62052ab1 | 14 | used for queuing all work items related to runtime PM, because this allows |
5e928f77 RW |
15 | them to be synchronized with system-wide power transitions (suspend to RAM, |
16 | hibernation and resume from system sleep states). pm_wq is declared in | |
17 | include/linux/pm_runtime.h and defined in kernel/power/main.c. | |
18 | ||
62052ab1 | 19 | * A number of runtime PM fields in the 'power' member of 'struct device' (which |
5e928f77 | 20 | is of the type 'struct dev_pm_info', defined in include/linux/pm.h) that can |
62052ab1 | 21 | be used for synchronizing runtime PM operations with one another. |
5e928f77 | 22 | |
62052ab1 | 23 | * Three device runtime PM callbacks in 'struct dev_pm_ops' (defined in |
5e928f77 RW |
24 | include/linux/pm.h). |
25 | ||
26 | * A set of helper functions defined in drivers/base/power/runtime.c that can be | |
62052ab1 | 27 | used for carrying out runtime PM operations in such a way that the |
5e928f77 RW |
28 | synchronization between them is taken care of by the PM core. Bus types and |
29 | device drivers are encouraged to use these functions. | |
30 | ||
62052ab1 | 31 | The runtime PM callbacks present in 'struct dev_pm_ops', the device runtime PM |
5e928f77 | 32 | fields of 'struct dev_pm_info' and the core helper functions provided for |
62052ab1 | 33 | runtime PM are described below. |
5e928f77 | 34 | |
62052ab1 | 35 | 2. Device Runtime PM Callbacks |
5e928f77 | 36 | |
62052ab1 | 37 | There are three device runtime PM callbacks defined in 'struct dev_pm_ops': |
5e928f77 RW |
38 | |
39 | struct dev_pm_ops { | |
40 | ... | |
41 | int (*runtime_suspend)(struct device *dev); | |
42 | int (*runtime_resume)(struct device *dev); | |
e1b1903e | 43 | int (*runtime_idle)(struct device *dev); |
5e928f77 RW |
44 | ... |
45 | }; | |
46 | ||
2fb242ad | 47 | The ->runtime_suspend(), ->runtime_resume() and ->runtime_idle() callbacks |
5841eb64 RW |
48 | are executed by the PM core for the device's subsystem that may be either of |
49 | the following: | |
50 | ||
51 | 1. PM domain of the device, if the device's PM domain object, dev->pm_domain, | |
52 | is present. | |
53 | ||
54 | 2. Device type of the device, if both dev->type and dev->type->pm are present. | |
55 | ||
56 | 3. Device class of the device, if both dev->class and dev->class->pm are | |
57 | present. | |
58 | ||
59 | 4. Bus type of the device, if both dev->bus and dev->bus->pm are present. | |
60 | ||
35cd133c RW |
61 | If the subsystem chosen by applying the above rules doesn't provide the relevant |
62 | callback, the PM core will invoke the corresponding driver callback stored in | |
63 | dev->driver->pm directly (if present). | |
64 | ||
5841eb64 RW |
65 | The PM core always checks which callback to use in the order given above, so the |
66 | priority order of callbacks from high to low is: PM domain, device type, class | |
67 | and bus type. Moreover, the high-priority one will always take precedence over | |
68 | a low-priority one. The PM domain, bus type, device type and class callbacks | |
69 | are referred to as subsystem-level callbacks in what follows. | |
a6ab7aa9 | 70 | |
c7b61de5 | 71 | By default, the callbacks are always invoked in process context with interrupts |
35cd133c RW |
72 | enabled. However, the pm_runtime_irq_safe() helper function can be used to tell |
73 | the PM core that it is safe to run the ->runtime_suspend(), ->runtime_resume() | |
74 | and ->runtime_idle() callbacks for the given device in atomic context with | |
75 | interrupts disabled. This implies that the callback routines in question must | |
76 | not block or sleep, but it also means that the synchronous helper functions | |
77 | listed at the end of Section 4 may be used for that device within an interrupt | |
78 | handler or generally in an atomic context. | |
79 | ||
80 | The subsystem-level suspend callback, if present, is _entirely_ _responsible_ | |
81 | for handling the suspend of the device as appropriate, which may, but need not | |
82 | include executing the device driver's own ->runtime_suspend() callback (from the | |
5e928f77 | 83 | PM core's point of view it is not necessary to implement a ->runtime_suspend() |
a6ab7aa9 RW |
84 | callback in a device driver as long as the subsystem-level suspend callback |
85 | knows what to do to handle the device). | |
5e928f77 | 86 | |
35cd133c RW |
87 | * Once the subsystem-level suspend callback (or the driver suspend callback, |
88 | if invoked directly) has completed successfully for the given device, the PM | |
89 | core regards the device as suspended, which need not mean that it has been | |
90 | put into a low power state. It is supposed to mean, however, that the | |
91 | device will not process data and will not communicate with the CPU(s) and | |
92 | RAM until the appropriate resume callback is executed for it. The runtime | |
93 | PM status of a device after successful execution of the suspend callback is | |
94 | 'suspended'. | |
95 | ||
96 | * If the suspend callback returns -EBUSY or -EAGAIN, the device's runtime PM | |
97 | status remains 'active', which means that the device _must_ be fully | |
98 | operational afterwards. | |
99 | ||
100 | * If the suspend callback returns an error code different from -EBUSY and | |
101 | -EAGAIN, the PM core regards this as a fatal error and will refuse to run | |
102 | the helper functions described in Section 4 for the device until its status | |
103 | is directly set to either'active', or 'suspended' (the PM core provides | |
104 | special helper functions for this purpose). | |
105 | ||
106 | In particular, if the driver requires remote wakeup capability (i.e. hardware | |
a6ab7aa9 RW |
107 | mechanism allowing the device to request a change of its power state, such as |
108 | PCI PME) for proper functioning and device_run_wake() returns 'false' for the | |
109 | device, then ->runtime_suspend() should return -EBUSY. On the other hand, if | |
35cd133c RW |
110 | device_run_wake() returns 'true' for the device and the device is put into a |
111 | low-power state during the execution of the suspend callback, it is expected | |
112 | that remote wakeup will be enabled for the device. Generally, remote wakeup | |
113 | should be enabled for all input devices put into low-power states at run time. | |
114 | ||
115 | The subsystem-level resume callback, if present, is _entirely_ _responsible_ for | |
116 | handling the resume of the device as appropriate, which may, but need not | |
117 | include executing the device driver's own ->runtime_resume() callback (from the | |
118 | PM core's point of view it is not necessary to implement a ->runtime_resume() | |
119 | callback in a device driver as long as the subsystem-level resume callback knows | |
120 | what to do to handle the device). | |
121 | ||
122 | * Once the subsystem-level resume callback (or the driver resume callback, if | |
123 | invoked directly) has completed successfully, the PM core regards the device | |
124 | as fully operational, which means that the device _must_ be able to complete | |
125 | I/O operations as needed. The runtime PM status of the device is then | |
126 | 'active'. | |
127 | ||
128 | * If the resume callback returns an error code, the PM core regards this as a | |
129 | fatal error and will refuse to run the helper functions described in Section | |
130 | 4 for the device, until its status is directly set to either 'active', or | |
131 | 'suspended' (by means of special helper functions provided by the PM core | |
132 | for this purpose). | |
133 | ||
134 | The idle callback (a subsystem-level one, if present, or the driver one) is | |
135 | executed by the PM core whenever the device appears to be idle, which is | |
136 | indicated to the PM core by two counters, the device's usage counter and the | |
137 | counter of 'active' children of the device. | |
5e928f77 RW |
138 | |
139 | * If any of these counters is decreased using a helper function provided by | |
140 | the PM core and it turns out to be equal to zero, the other counter is | |
141 | checked. If that counter also is equal to zero, the PM core executes the | |
35cd133c | 142 | idle callback with the device as its argument. |
5e928f77 | 143 | |
35cd133c RW |
144 | The action performed by the idle callback is totally dependent on the subsystem |
145 | (or driver) in question, but the expected and recommended action is to check | |
a6ab7aa9 RW |
146 | if the device can be suspended (i.e. if all of the conditions necessary for |
147 | suspending the device are satisfied) and to queue up a suspend request for the | |
43d51af4 | 148 | device in that case. If there is no idle callback, or if the callback returns |
d66e6db2 UH |
149 | 0, then the PM core will attempt to carry out a runtime suspend of the device, |
150 | also respecting devices configured for autosuspend. In essence this means a | |
151 | call to pm_runtime_autosuspend() (do note that drivers needs to update the | |
152 | device last busy mark, pm_runtime_mark_last_busy(), to control the delay under | |
153 | this circumstance). To prevent this (for example, if the callback routine has | |
154 | started a delayed suspend), the routine must return a non-zero value. Negative | |
155 | error return codes are ignored by the PM core. | |
5e928f77 RW |
156 | |
157 | The helper functions provided by the PM core, described in Section 4, guarantee | |
35cd133c RW |
158 | that the following constraints are met with respect to runtime PM callbacks for |
159 | one device: | |
5e928f77 RW |
160 | |
161 | (1) The callbacks are mutually exclusive (e.g. it is forbidden to execute | |
162 | ->runtime_suspend() in parallel with ->runtime_resume() or with another | |
163 | instance of ->runtime_suspend() for the same device) with the exception that | |
164 | ->runtime_suspend() or ->runtime_resume() can be executed in parallel with | |
165 | ->runtime_idle() (although ->runtime_idle() will not be started while any | |
166 | of the other callbacks is being executed for the same device). | |
167 | ||
168 | (2) ->runtime_idle() and ->runtime_suspend() can only be executed for 'active' | |
169 | devices (i.e. the PM core will only execute ->runtime_idle() or | |
62052ab1 | 170 | ->runtime_suspend() for the devices the runtime PM status of which is |
5e928f77 RW |
171 | 'active'). |
172 | ||
173 | (3) ->runtime_idle() and ->runtime_suspend() can only be executed for a device | |
174 | the usage counter of which is equal to zero _and_ either the counter of | |
175 | 'active' children of which is equal to zero, or the 'power.ignore_children' | |
176 | flag of which is set. | |
177 | ||
178 | (4) ->runtime_resume() can only be executed for 'suspended' devices (i.e. the | |
62052ab1 | 179 | PM core will only execute ->runtime_resume() for the devices the runtime |
5e928f77 RW |
180 | PM status of which is 'suspended'). |
181 | ||
182 | Additionally, the helper functions provided by the PM core obey the following | |
183 | rules: | |
184 | ||
185 | * If ->runtime_suspend() is about to be executed or there's a pending request | |
186 | to execute it, ->runtime_idle() will not be executed for the same device. | |
187 | ||
188 | * A request to execute or to schedule the execution of ->runtime_suspend() | |
189 | will cancel any pending requests to execute ->runtime_idle() for the same | |
190 | device. | |
191 | ||
192 | * If ->runtime_resume() is about to be executed or there's a pending request | |
193 | to execute it, the other callbacks will not be executed for the same device. | |
194 | ||
195 | * A request to execute ->runtime_resume() will cancel any pending or | |
15bcb91d AS |
196 | scheduled requests to execute the other callbacks for the same device, |
197 | except for scheduled autosuspends. | |
5e928f77 | 198 | |
62052ab1 | 199 | 3. Runtime PM Device Fields |
5e928f77 | 200 | |
62052ab1 | 201 | The following device runtime PM fields are present in 'struct dev_pm_info', as |
5e928f77 RW |
202 | defined in include/linux/pm.h: |
203 | ||
204 | struct timer_list suspend_timer; | |
15bcb91d | 205 | - timer used for scheduling (delayed) suspend and autosuspend requests |
5e928f77 RW |
206 | |
207 | unsigned long timer_expires; | |
208 | - timer expiration time, in jiffies (if this is different from zero, the | |
209 | timer is running and will expire at that time, otherwise the timer is not | |
210 | running) | |
211 | ||
212 | struct work_struct work; | |
213 | - work structure used for queuing up requests (i.e. work items in pm_wq) | |
214 | ||
215 | wait_queue_head_t wait_queue; | |
216 | - wait queue used if any of the helper functions needs to wait for another | |
217 | one to complete | |
218 | ||
219 | spinlock_t lock; | |
220 | - lock used for synchronisation | |
221 | ||
222 | atomic_t usage_count; | |
223 | - the usage counter of the device | |
224 | ||
225 | atomic_t child_count; | |
226 | - the count of 'active' children of the device | |
227 | ||
228 | unsigned int ignore_children; | |
229 | - if set, the value of child_count is ignored (but still updated) | |
230 | ||
231 | unsigned int disable_depth; | |
1f999d14 | 232 | - used for disabling the helper functions (they work normally if this is |
62052ab1 | 233 | equal to zero); the initial value of it is 1 (i.e. runtime PM is |
5e928f77 RW |
234 | initially disabled for all devices) |
235 | ||
e1a29760 | 236 | int runtime_error; |
5e928f77 | 237 | - if set, there was a fatal error (one of the callbacks returned error code |
1f999d14 | 238 | as described in Section 2), so the helper functions will not work until |
5e928f77 RW |
239 | this flag is cleared; this is the error code returned by the failing |
240 | callback | |
241 | ||
242 | unsigned int idle_notification; | |
243 | - if set, ->runtime_idle() is being executed | |
244 | ||
245 | unsigned int request_pending; | |
246 | - if set, there's a pending request (i.e. a work item queued up into pm_wq) | |
247 | ||
248 | enum rpm_request request; | |
249 | - type of request that's pending (valid if request_pending is set) | |
250 | ||
251 | unsigned int deferred_resume; | |
252 | - set if ->runtime_resume() is about to be run while ->runtime_suspend() is | |
253 | being executed for that device and it is not practical to wait for the | |
254 | suspend to complete; means "start a resume as soon as you've suspended" | |
255 | ||
7a1a8eb5 | 256 | unsigned int run_wake; |
62052ab1 | 257 | - set if the device is capable of generating runtime wake-up events |
7a1a8eb5 | 258 | |
5e928f77 | 259 | enum rpm_status runtime_status; |
62052ab1 | 260 | - the runtime PM status of the device; this field's initial value is |
5e928f77 RW |
261 | RPM_SUSPENDED, which means that each device is initially regarded by the |
262 | PM core as 'suspended', regardless of its real hardware status | |
263 | ||
87d1b3e6 RW |
264 | unsigned int runtime_auto; |
265 | - if set, indicates that the user space has allowed the device driver to | |
266 | power manage the device at run time via the /sys/devices/.../power/control | |
267 | interface; it may only be modified with the help of the pm_runtime_allow() | |
268 | and pm_runtime_forbid() helper functions | |
269 | ||
7490e442 | 270 | unsigned int no_callbacks; |
62052ab1 | 271 | - indicates that the device does not use the runtime PM callbacks (see |
7490e442 AS |
272 | Section 8); it may be modified only by the pm_runtime_no_callbacks() |
273 | helper function | |
274 | ||
c7b61de5 AS |
275 | unsigned int irq_safe; |
276 | - indicates that the ->runtime_suspend() and ->runtime_resume() callbacks | |
277 | will be invoked with the spinlock held and interrupts disabled | |
278 | ||
15bcb91d AS |
279 | unsigned int use_autosuspend; |
280 | - indicates that the device's driver supports delayed autosuspend (see | |
281 | Section 9); it may be modified only by the | |
282 | pm_runtime{_dont}_use_autosuspend() helper functions | |
283 | ||
284 | unsigned int timer_autosuspends; | |
285 | - indicates that the PM core should attempt to carry out an autosuspend | |
286 | when the timer expires rather than a normal suspend | |
287 | ||
288 | int autosuspend_delay; | |
289 | - the delay time (in milliseconds) to be used for autosuspend | |
290 | ||
291 | unsigned long last_busy; | |
292 | - the time (in jiffies) when the pm_runtime_mark_last_busy() helper | |
293 | function was last called for this device; used in calculating inactivity | |
294 | periods for autosuspend | |
295 | ||
5e928f77 RW |
296 | All of the above fields are members of the 'power' member of 'struct device'. |
297 | ||
62052ab1 | 298 | 4. Runtime PM Device Helper Functions |
5e928f77 | 299 | |
62052ab1 | 300 | The following runtime PM helper functions are defined in |
5e928f77 RW |
301 | drivers/base/power/runtime.c and include/linux/pm_runtime.h: |
302 | ||
303 | void pm_runtime_init(struct device *dev); | |
62052ab1 | 304 | - initialize the device runtime PM fields in 'struct dev_pm_info' |
5e928f77 RW |
305 | |
306 | void pm_runtime_remove(struct device *dev); | |
62052ab1 | 307 | - make sure that the runtime PM of the device will be disabled after |
5e928f77 RW |
308 | removing the device from device hierarchy |
309 | ||
310 | int pm_runtime_idle(struct device *dev); | |
43d51af4 AS |
311 | - execute the subsystem-level idle callback for the device; returns an |
312 | error code on failure, where -EINPROGRESS means that ->runtime_idle() is | |
313 | already being executed; if there is no callback or the callback returns 0 | |
d66e6db2 | 314 | then run pm_runtime_autosuspend(dev) and return its result |
5e928f77 RW |
315 | |
316 | int pm_runtime_suspend(struct device *dev); | |
a6ab7aa9 | 317 | - execute the subsystem-level suspend callback for the device; returns 0 on |
62052ab1 | 318 | success, 1 if the device's runtime PM status was already 'suspended', or |
5e928f77 | 319 | error code on failure, where -EAGAIN or -EBUSY means it is safe to attempt |
632e270e RW |
320 | to suspend the device again in future and -EACCES means that |
321 | 'power.disable_depth' is different from 0 | |
5e928f77 | 322 | |
15bcb91d AS |
323 | int pm_runtime_autosuspend(struct device *dev); |
324 | - same as pm_runtime_suspend() except that the autosuspend delay is taken | |
325 | into account; if pm_runtime_autosuspend_expiration() says the delay has | |
326 | not yet expired then an autosuspend is scheduled for the appropriate time | |
327 | and 0 is returned | |
328 | ||
5e928f77 | 329 | int pm_runtime_resume(struct device *dev); |
de8164fb | 330 | - execute the subsystem-level resume callback for the device; returns 0 on |
62052ab1 | 331 | success, 1 if the device's runtime PM status was already 'active' or |
5e928f77 RW |
332 | error code on failure, where -EAGAIN means it may be safe to attempt to |
333 | resume the device again in future, but 'power.runtime_error' should be | |
632e270e RW |
334 | checked additionally, and -EACCES means that 'power.disable_depth' is |
335 | different from 0 | |
5e928f77 RW |
336 | |
337 | int pm_request_idle(struct device *dev); | |
a6ab7aa9 RW |
338 | - submit a request to execute the subsystem-level idle callback for the |
339 | device (the request is represented by a work item in pm_wq); returns 0 on | |
340 | success or error code if the request has not been queued up | |
5e928f77 | 341 | |
15bcb91d AS |
342 | int pm_request_autosuspend(struct device *dev); |
343 | - schedule the execution of the subsystem-level suspend callback for the | |
344 | device when the autosuspend delay has expired; if the delay has already | |
345 | expired then the work item is queued up immediately | |
346 | ||
5e928f77 | 347 | int pm_schedule_suspend(struct device *dev, unsigned int delay); |
a6ab7aa9 RW |
348 | - schedule the execution of the subsystem-level suspend callback for the |
349 | device in future, where 'delay' is the time to wait before queuing up a | |
350 | suspend work item in pm_wq, in milliseconds (if 'delay' is zero, the work | |
351 | item is queued up immediately); returns 0 on success, 1 if the device's PM | |
62052ab1 | 352 | runtime status was already 'suspended', or error code if the request |
5e928f77 RW |
353 | hasn't been scheduled (or queued up if 'delay' is 0); if the execution of |
354 | ->runtime_suspend() is already scheduled and not yet expired, the new | |
355 | value of 'delay' will be used as the time to wait | |
356 | ||
357 | int pm_request_resume(struct device *dev); | |
a6ab7aa9 RW |
358 | - submit a request to execute the subsystem-level resume callback for the |
359 | device (the request is represented by a work item in pm_wq); returns 0 on | |
62052ab1 | 360 | success, 1 if the device's runtime PM status was already 'active', or |
5e928f77 RW |
361 | error code if the request hasn't been queued up |
362 | ||
363 | void pm_runtime_get_noresume(struct device *dev); | |
364 | - increment the device's usage counter | |
365 | ||
366 | int pm_runtime_get(struct device *dev); | |
367 | - increment the device's usage counter, run pm_request_resume(dev) and | |
368 | return its result | |
369 | ||
370 | int pm_runtime_get_sync(struct device *dev); | |
371 | - increment the device's usage counter, run pm_runtime_resume(dev) and | |
372 | return its result | |
373 | ||
a436b6a1 RW |
374 | int pm_runtime_get_if_in_use(struct device *dev); |
375 | - return -EINVAL if 'power.disable_depth' is nonzero; otherwise, if the | |
376 | runtime PM status is RPM_ACTIVE and the runtime PM usage counter is | |
377 | nonzero, increment the counter and return 1; otherwise return 0 without | |
378 | changing the counter | |
379 | ||
5e928f77 RW |
380 | void pm_runtime_put_noidle(struct device *dev); |
381 | - decrement the device's usage counter | |
382 | ||
383 | int pm_runtime_put(struct device *dev); | |
15bcb91d AS |
384 | - decrement the device's usage counter; if the result is 0 then run |
385 | pm_request_idle(dev) and return its result | |
386 | ||
387 | int pm_runtime_put_autosuspend(struct device *dev); | |
388 | - decrement the device's usage counter; if the result is 0 then run | |
389 | pm_request_autosuspend(dev) and return its result | |
5e928f77 RW |
390 | |
391 | int pm_runtime_put_sync(struct device *dev); | |
15bcb91d AS |
392 | - decrement the device's usage counter; if the result is 0 then run |
393 | pm_runtime_idle(dev) and return its result | |
394 | ||
c7b61de5 AS |
395 | int pm_runtime_put_sync_suspend(struct device *dev); |
396 | - decrement the device's usage counter; if the result is 0 then run | |
397 | pm_runtime_suspend(dev) and return its result | |
398 | ||
15bcb91d AS |
399 | int pm_runtime_put_sync_autosuspend(struct device *dev); |
400 | - decrement the device's usage counter; if the result is 0 then run | |
401 | pm_runtime_autosuspend(dev) and return its result | |
5e928f77 RW |
402 | |
403 | void pm_runtime_enable(struct device *dev); | |
e358bad7 | 404 | - decrement the device's 'power.disable_depth' field; if that field is equal |
62052ab1 | 405 | to zero, the runtime PM helper functions can execute subsystem-level |
e358bad7 | 406 | callbacks described in Section 2 for the device |
5e928f77 RW |
407 | |
408 | int pm_runtime_disable(struct device *dev); | |
e358bad7 RW |
409 | - increment the device's 'power.disable_depth' field (if the value of that |
410 | field was previously zero, this prevents subsystem-level runtime PM | |
91e63cc0 GU |
411 | callbacks from being run for the device), make sure that all of the |
412 | pending runtime PM operations on the device are either completed or | |
413 | canceled; returns 1 if there was a resume request pending and it was | |
414 | necessary to execute the subsystem-level resume callback for the device | |
415 | to satisfy that request, otherwise 0 is returned | |
5e928f77 | 416 | |
e358bad7 RW |
417 | int pm_runtime_barrier(struct device *dev); |
418 | - check if there's a resume request pending for the device and resume it | |
419 | (synchronously) in that case, cancel any other pending runtime PM requests | |
420 | regarding it and wait for all runtime PM operations on it in progress to | |
421 | complete; returns 1 if there was a resume request pending and it was | |
422 | necessary to execute the subsystem-level resume callback for the device to | |
423 | satisfy that request, otherwise 0 is returned | |
424 | ||
5e928f77 RW |
425 | void pm_suspend_ignore_children(struct device *dev, bool enable); |
426 | - set/unset the power.ignore_children flag of the device | |
427 | ||
428 | int pm_runtime_set_active(struct device *dev); | |
62052ab1 | 429 | - clear the device's 'power.runtime_error' flag, set the device's runtime |
5e928f77 RW |
430 | PM status to 'active' and update its parent's counter of 'active' |
431 | children as appropriate (it is only valid to use this function if | |
432 | 'power.runtime_error' is set or 'power.disable_depth' is greater than | |
433 | zero); it will fail and return error code if the device has a parent | |
434 | which is not active and the 'power.ignore_children' flag of which is unset | |
435 | ||
436 | void pm_runtime_set_suspended(struct device *dev); | |
62052ab1 | 437 | - clear the device's 'power.runtime_error' flag, set the device's runtime |
5e928f77 RW |
438 | PM status to 'suspended' and update its parent's counter of 'active' |
439 | children as appropriate (it is only valid to use this function if | |
440 | 'power.runtime_error' is set or 'power.disable_depth' is greater than | |
441 | zero) | |
442 | ||
fbadc58d SL |
443 | bool pm_runtime_active(struct device *dev); |
444 | - return true if the device's runtime PM status is 'active' or its | |
445 | 'power.disable_depth' field is not equal to zero, or false otherwise | |
446 | ||
d690b2cd | 447 | bool pm_runtime_suspended(struct device *dev); |
f08f5a0a RW |
448 | - return true if the device's runtime PM status is 'suspended' and its |
449 | 'power.disable_depth' field is equal to zero, or false otherwise | |
d690b2cd | 450 | |
f3393b62 KH |
451 | bool pm_runtime_status_suspended(struct device *dev); |
452 | - return true if the device's runtime PM status is 'suspended' | |
453 | ||
87d1b3e6 RW |
454 | void pm_runtime_allow(struct device *dev); |
455 | - set the power.runtime_auto flag for the device and decrease its usage | |
456 | counter (used by the /sys/devices/.../power/control interface to | |
457 | effectively allow the device to be power managed at run time) | |
458 | ||
459 | void pm_runtime_forbid(struct device *dev); | |
460 | - unset the power.runtime_auto flag for the device and increase its usage | |
461 | counter (used by the /sys/devices/.../power/control interface to | |
462 | effectively prevent the device from being power managed at run time) | |
463 | ||
7490e442 | 464 | void pm_runtime_no_callbacks(struct device *dev); |
62052ab1 | 465 | - set the power.no_callbacks flag for the device and remove the runtime |
7490e442 AS |
466 | PM attributes from /sys/devices/.../power (or prevent them from being |
467 | added when the device is registered) | |
468 | ||
c7b61de5 AS |
469 | void pm_runtime_irq_safe(struct device *dev); |
470 | - set the power.irq_safe flag for the device, causing the runtime-PM | |
64584eb9 | 471 | callbacks to be invoked with interrupts off |
c7b61de5 | 472 | |
3fb1581e KK |
473 | bool pm_runtime_is_irq_safe(struct device *dev); |
474 | - return true if power.irq_safe flag was set for the device, causing | |
475 | the runtime-PM callbacks to be invoked with interrupts off | |
476 | ||
15bcb91d AS |
477 | void pm_runtime_mark_last_busy(struct device *dev); |
478 | - set the power.last_busy field to the current time | |
479 | ||
480 | void pm_runtime_use_autosuspend(struct device *dev); | |
481 | - set the power.use_autosuspend flag, enabling autosuspend delays | |
482 | ||
483 | void pm_runtime_dont_use_autosuspend(struct device *dev); | |
484 | - clear the power.use_autosuspend flag, disabling autosuspend delays | |
485 | ||
486 | void pm_runtime_set_autosuspend_delay(struct device *dev, int delay); | |
487 | - set the power.autosuspend_delay value to 'delay' (expressed in | |
62052ab1 | 488 | milliseconds); if 'delay' is negative then runtime suspends are |
15bcb91d AS |
489 | prevented |
490 | ||
491 | unsigned long pm_runtime_autosuspend_expiration(struct device *dev); | |
492 | - calculate the time when the current autosuspend delay period will expire, | |
493 | based on power.last_busy and power.autosuspend_delay; if the delay time | |
494 | is 1000 ms or larger then the expiration time is rounded up to the | |
495 | nearest second; returns 0 if the delay period has already expired or | |
496 | power.use_autosuspend isn't set, otherwise returns the expiration time | |
497 | in jiffies | |
498 | ||
5e928f77 RW |
499 | It is safe to execute the following helper functions from interrupt context: |
500 | ||
501 | pm_request_idle() | |
15bcb91d | 502 | pm_request_autosuspend() |
5e928f77 RW |
503 | pm_schedule_suspend() |
504 | pm_request_resume() | |
505 | pm_runtime_get_noresume() | |
506 | pm_runtime_get() | |
507 | pm_runtime_put_noidle() | |
508 | pm_runtime_put() | |
15bcb91d AS |
509 | pm_runtime_put_autosuspend() |
510 | pm_runtime_enable() | |
5e928f77 RW |
511 | pm_suspend_ignore_children() |
512 | pm_runtime_set_active() | |
513 | pm_runtime_set_suspended() | |
15bcb91d AS |
514 | pm_runtime_suspended() |
515 | pm_runtime_mark_last_busy() | |
516 | pm_runtime_autosuspend_expiration() | |
5e928f77 | 517 | |
c7b61de5 AS |
518 | If pm_runtime_irq_safe() has been called for a device then the following helper |
519 | functions may also be used in interrupt context: | |
520 | ||
2e6ba515 | 521 | pm_runtime_idle() |
c7b61de5 AS |
522 | pm_runtime_suspend() |
523 | pm_runtime_autosuspend() | |
524 | pm_runtime_resume() | |
525 | pm_runtime_get_sync() | |
02b26774 | 526 | pm_runtime_put_sync() |
c7b61de5 | 527 | pm_runtime_put_sync_suspend() |
311aab73 | 528 | pm_runtime_put_sync_autosuspend() |
c7b61de5 | 529 | |
62052ab1 | 530 | 5. Runtime PM Initialization, Device Probing and Removal |
5e928f77 | 531 | |
62052ab1 | 532 | Initially, the runtime PM is disabled for all devices, which means that the |
1f999d14 | 533 | majority of the runtime PM helper functions described in Section 4 will return |
5e928f77 RW |
534 | -EAGAIN until pm_runtime_enable() is called for the device. |
535 | ||
62052ab1 | 536 | In addition to that, the initial runtime PM status of all devices is |
5e928f77 RW |
537 | 'suspended', but it need not reflect the actual physical state of the device. |
538 | Thus, if the device is initially active (i.e. it is able to process I/O), its | |
62052ab1 | 539 | runtime PM status must be changed to 'active', with the help of |
5e928f77 RW |
540 | pm_runtime_set_active(), before pm_runtime_enable() is called for the device. |
541 | ||
62052ab1 | 542 | However, if the device has a parent and the parent's runtime PM is enabled, |
5e928f77 RW |
543 | calling pm_runtime_set_active() for the device will affect the parent, unless |
544 | the parent's 'power.ignore_children' flag is set. Namely, in that case the | |
545 | parent won't be able to suspend at run time, using the PM core's helper | |
546 | functions, as long as the child's status is 'active', even if the child's | |
62052ab1 | 547 | runtime PM is still disabled (i.e. pm_runtime_enable() hasn't been called for |
5e928f77 RW |
548 | the child yet or pm_runtime_disable() has been called for it). For this reason, |
549 | once pm_runtime_set_active() has been called for the device, pm_runtime_enable() | |
62052ab1 | 550 | should be called for it too as soon as reasonably possible or its runtime PM |
5e928f77 RW |
551 | status should be changed back to 'suspended' with the help of |
552 | pm_runtime_set_suspended(). | |
553 | ||
62052ab1 | 554 | If the default initial runtime PM status of the device (i.e. 'suspended') |
5e928f77 RW |
555 | reflects the actual state of the device, its bus type's or its driver's |
556 | ->probe() callback will likely need to wake it up using one of the PM core's | |
557 | helper functions described in Section 4. In that case, pm_runtime_resume() | |
62052ab1 | 558 | should be used. Of course, for this purpose the device's runtime PM has to be |
5e928f77 RW |
559 | enabled earlier by calling pm_runtime_enable(). |
560 | ||
f6a2fbb9 BD |
561 | Note, if the device may execute pm_runtime calls during the probe (such as |
562 | if it is registers with a subsystem that may call back in) then the | |
563 | pm_runtime_get_sync() call paired with a pm_runtime_put() call will be | |
564 | appropriate to ensure that the device is not put back to sleep during the | |
565 | probe. This can happen with systems such as the network device layer. | |
566 | ||
ea309944 UH |
567 | It may be desirable to suspend the device once ->probe() has finished. |
568 | Therefore the driver core uses the asyncronous pm_request_idle() to submit a | |
569 | request to execute the subsystem-level idle callback for the device at that | |
570 | time. A driver that makes use of the runtime autosuspend feature, may want to | |
571 | update the last busy mark before returning from ->probe(). | |
f5da24db RW |
572 | |
573 | Moreover, the driver core prevents runtime PM callbacks from racing with the bus | |
574 | notifier callback in __device_release_driver(), which is necessary, because the | |
575 | notifier is used by some subsystems to carry out operations affecting the | |
576 | runtime PM functionality. It does so by calling pm_runtime_get_sync() before | |
577 | driver_sysfs_remove() and the BUS_NOTIFY_UNBIND_DRIVER notifications. This | |
578 | resumes the device if it's in the suspended state and prevents it from | |
579 | being suspended again while those routines are being executed. | |
580 | ||
581 | To allow bus types and drivers to put devices into the suspended state by | |
582 | calling pm_runtime_suspend() from their ->remove() routines, the driver core | |
583 | executes pm_runtime_put_sync() after running the BUS_NOTIFY_UNBIND_DRIVER | |
584 | notifications in __device_release_driver(). This requires bus types and | |
585 | drivers to make their ->remove() callbacks avoid races with runtime PM directly, | |
586 | but also it allows of more flexibility in the handling of devices during the | |
587 | removal of their drivers. | |
f1212ae1 | 588 | |
8fd2910e KK |
589 | Drivers in ->remove() callback should undo the runtime PM changes done |
590 | in ->probe(). Usually this means calling pm_runtime_disable(), | |
591 | pm_runtime_dont_use_autosuspend() etc. | |
592 | ||
87d1b3e6 RW |
593 | The user space can effectively disallow the driver of the device to power manage |
594 | it at run time by changing the value of its /sys/devices/.../power/control | |
595 | attribute to "on", which causes pm_runtime_forbid() to be called. In principle, | |
596 | this mechanism may also be used by the driver to effectively turn off the | |
62052ab1 RW |
597 | runtime power management of the device until the user space turns it on. |
598 | Namely, during the initialization the driver can make sure that the runtime PM | |
87d1b3e6 RW |
599 | status of the device is 'active' and call pm_runtime_forbid(). It should be |
600 | noted, however, that if the user space has already intentionally changed the | |
601 | value of /sys/devices/.../power/control to "auto" to allow the driver to power | |
602 | manage the device at run time, the driver may confuse it by using | |
603 | pm_runtime_forbid() this way. | |
604 | ||
62052ab1 | 605 | 6. Runtime PM and System Sleep |
f1212ae1 | 606 | |
62052ab1 | 607 | Runtime PM and system sleep (i.e., system suspend and hibernation, also known |
f1212ae1 AS |
608 | as suspend-to-RAM and suspend-to-disk) interact with each other in a couple of |
609 | ways. If a device is active when a system sleep starts, everything is | |
610 | straightforward. But what should happen if the device is already suspended? | |
611 | ||
62052ab1 RW |
612 | The device may have different wake-up settings for runtime PM and system sleep. |
613 | For example, remote wake-up may be enabled for runtime suspend but disallowed | |
f1212ae1 AS |
614 | for system sleep (device_may_wakeup(dev) returns 'false'). When this happens, |
615 | the subsystem-level system suspend callback is responsible for changing the | |
616 | device's wake-up setting (it may leave that to the device driver's system | |
617 | suspend routine). It may be necessary to resume the device and suspend it again | |
618 | in order to do so. The same is true if the driver uses different power levels | |
62052ab1 | 619 | or other settings for runtime suspend and system sleep. |
f1212ae1 | 620 | |
455716e9 RW |
621 | During system resume, the simplest approach is to bring all devices back to full |
622 | power, even if they had been suspended before the system suspend began. There | |
623 | are several reasons for this, including: | |
f1212ae1 AS |
624 | |
625 | * The device might need to switch power levels, wake-up settings, etc. | |
626 | ||
627 | * Remote wake-up events might have been lost by the firmware. | |
628 | ||
629 | * The device's children may need the device to be at full power in order | |
630 | to resume themselves. | |
631 | ||
632 | * The driver's idea of the device state may not agree with the device's | |
633 | physical state. This can happen during resume from hibernation. | |
634 | ||
635 | * The device might need to be reset. | |
636 | ||
637 | * Even though the device was suspended, if its usage counter was > 0 then most | |
62052ab1 | 638 | likely it would need a runtime resume in the near future anyway. |
f1212ae1 | 639 | |
455716e9 | 640 | If the device had been suspended before the system suspend began and it's |
62052ab1 | 641 | brought back to full power during resume, then its runtime PM status will have |
455716e9 RW |
642 | to be updated to reflect the actual post-system sleep status. The way to do |
643 | this is: | |
f1212ae1 AS |
644 | |
645 | pm_runtime_disable(dev); | |
646 | pm_runtime_set_active(dev); | |
647 | pm_runtime_enable(dev); | |
648 | ||
62052ab1 | 649 | The PM core always increments the runtime usage counter before calling the |
1e2ef05b | 650 | ->suspend() callback and decrements it after calling the ->resume() callback. |
62052ab1 | 651 | Hence disabling runtime PM temporarily like this will not cause any runtime |
1e2ef05b RW |
652 | suspend attempts to be permanently lost. If the usage count goes to zero |
653 | following the return of the ->resume() callback, the ->runtime_idle() callback | |
654 | will be invoked as usual. | |
655 | ||
455716e9 RW |
656 | On some systems, however, system sleep is not entered through a global firmware |
657 | or hardware operation. Instead, all hardware components are put into low-power | |
658 | states directly by the kernel in a coordinated way. Then, the system sleep | |
659 | state effectively follows from the states the hardware components end up in | |
660 | and the system is woken up from that state by a hardware interrupt or a similar | |
661 | mechanism entirely under the kernel's control. As a result, the kernel never | |
662 | gives control away and the states of all devices during resume are precisely | |
663 | known to it. If that is the case and none of the situations listed above takes | |
664 | place (in particular, if the system is not waking up from hibernation), it may | |
665 | be more efficient to leave the devices that had been suspended before the system | |
666 | suspend began in the suspended state. | |
667 | ||
f71495f3 RW |
668 | To this end, the PM core provides a mechanism allowing some coordination between |
669 | different levels of device hierarchy. Namely, if a system suspend .prepare() | |
670 | callback returns a positive number for a device, that indicates to the PM core | |
671 | that the device appears to be runtime-suspended and its state is fine, so it | |
672 | may be left in runtime suspend provided that all of its descendants are also | |
673 | left in runtime suspend. If that happens, the PM core will not execute any | |
674 | system suspend and resume callbacks for all of those devices, except for the | |
675 | complete callback, which is then entirely responsible for handling the device | |
676 | as appropriate. This only applies to system suspend transitions that are not | |
677 | related to hibernation (see Documentation/power/devices.txt for more | |
678 | information). | |
679 | ||
1e2ef05b RW |
680 | The PM core does its best to reduce the probability of race conditions between |
681 | the runtime PM and system suspend/resume (and hibernation) callbacks by carrying | |
682 | out the following operations: | |
683 | ||
4ec6a9cc RW |
684 | * During system suspend pm_runtime_get_noresume() is called for every device |
685 | right before executing the subsystem-level .prepare() callback for it and | |
686 | pm_runtime_barrier() is called for every device right before executing the | |
687 | subsystem-level .suspend() callback for it. In addition to that the PM core | |
688 | calls __pm_runtime_disable() with 'false' as the second argument for every | |
689 | device right before executing the subsystem-level .suspend_late() callback | |
690 | for it. | |
691 | ||
692 | * During system resume pm_runtime_enable() and pm_runtime_put() are called for | |
693 | every device right after executing the subsystem-level .resume_early() | |
694 | callback and right after executing the subsystem-level .complete() callback | |
9f6d8f6a | 695 | for it, respectively. |
1e2ef05b | 696 | |
d690b2cd RW |
697 | 7. Generic subsystem callbacks |
698 | ||
699 | Subsystems may wish to conserve code space by using the set of generic power | |
700 | management callbacks provided by the PM core, defined in | |
701 | driver/base/power/generic_ops.c: | |
702 | ||
d690b2cd RW |
703 | int pm_generic_runtime_suspend(struct device *dev); |
704 | - invoke the ->runtime_suspend() callback provided by the driver of this | |
39c29f3d | 705 | device and return its result, or return 0 if not defined |
d690b2cd RW |
706 | |
707 | int pm_generic_runtime_resume(struct device *dev); | |
708 | - invoke the ->runtime_resume() callback provided by the driver of this | |
39c29f3d | 709 | device and return its result, or return 0 if not defined |
d690b2cd RW |
710 | |
711 | int pm_generic_suspend(struct device *dev); | |
712 | - if the device has not been suspended at run time, invoke the ->suspend() | |
713 | callback provided by its driver and return its result, or return 0 if not | |
714 | defined | |
715 | ||
e5291928 RW |
716 | int pm_generic_suspend_noirq(struct device *dev); |
717 | - if pm_runtime_suspended(dev) returns "false", invoke the ->suspend_noirq() | |
718 | callback provided by the device's driver and return its result, or return | |
719 | 0 if not defined | |
720 | ||
d690b2cd RW |
721 | int pm_generic_resume(struct device *dev); |
722 | - invoke the ->resume() callback provided by the driver of this device and, | |
723 | if successful, change the device's runtime PM status to 'active' | |
724 | ||
e5291928 RW |
725 | int pm_generic_resume_noirq(struct device *dev); |
726 | - invoke the ->resume_noirq() callback provided by the driver of this device | |
727 | ||
d690b2cd RW |
728 | int pm_generic_freeze(struct device *dev); |
729 | - if the device has not been suspended at run time, invoke the ->freeze() | |
730 | callback provided by its driver and return its result, or return 0 if not | |
731 | defined | |
732 | ||
e5291928 RW |
733 | int pm_generic_freeze_noirq(struct device *dev); |
734 | - if pm_runtime_suspended(dev) returns "false", invoke the ->freeze_noirq() | |
735 | callback provided by the device's driver and return its result, or return | |
736 | 0 if not defined | |
737 | ||
d690b2cd RW |
738 | int pm_generic_thaw(struct device *dev); |
739 | - if the device has not been suspended at run time, invoke the ->thaw() | |
740 | callback provided by its driver and return its result, or return 0 if not | |
741 | defined | |
742 | ||
e5291928 RW |
743 | int pm_generic_thaw_noirq(struct device *dev); |
744 | - if pm_runtime_suspended(dev) returns "false", invoke the ->thaw_noirq() | |
745 | callback provided by the device's driver and return its result, or return | |
746 | 0 if not defined | |
747 | ||
d690b2cd RW |
748 | int pm_generic_poweroff(struct device *dev); |
749 | - if the device has not been suspended at run time, invoke the ->poweroff() | |
750 | callback provided by its driver and return its result, or return 0 if not | |
751 | defined | |
752 | ||
e5291928 RW |
753 | int pm_generic_poweroff_noirq(struct device *dev); |
754 | - if pm_runtime_suspended(dev) returns "false", run the ->poweroff_noirq() | |
755 | callback provided by the device's driver and return its result, or return | |
756 | 0 if not defined | |
757 | ||
d690b2cd RW |
758 | int pm_generic_restore(struct device *dev); |
759 | - invoke the ->restore() callback provided by the driver of this device and, | |
760 | if successful, change the device's runtime PM status to 'active' | |
761 | ||
e5291928 RW |
762 | int pm_generic_restore_noirq(struct device *dev); |
763 | - invoke the ->restore_noirq() callback provided by the device's driver | |
764 | ||
fd6fe826 GU |
765 | These functions are the defaults used by the PM core, if a subsystem doesn't |
766 | provide its own callbacks for ->runtime_idle(), ->runtime_suspend(), | |
e5291928 RW |
767 | ->runtime_resume(), ->suspend(), ->suspend_noirq(), ->resume(), |
768 | ->resume_noirq(), ->freeze(), ->freeze_noirq(), ->thaw(), ->thaw_noirq(), | |
fd6fe826 GU |
769 | ->poweroff(), ->poweroff_noirq(), ->restore(), ->restore_noirq() in the |
770 | subsystem-level dev_pm_ops structure. | |
d690b2cd RW |
771 | |
772 | Device drivers that wish to use the same function as a system suspend, freeze, | |
62052ab1 RW |
773 | poweroff and runtime suspend callback, and similarly for system resume, thaw, |
774 | restore, and runtime resume, can achieve this with the help of the | |
d690b2cd RW |
775 | UNIVERSAL_DEV_PM_OPS macro defined in include/linux/pm.h (possibly setting its |
776 | last argument to NULL). | |
7490e442 AS |
777 | |
778 | 8. "No-Callback" Devices | |
779 | ||
780 | Some "devices" are only logical sub-devices of their parent and cannot be | |
781 | power-managed on their own. (The prototype example is a USB interface. Entire | |
782 | USB devices can go into low-power mode or send wake-up requests, but neither is | |
783 | possible for individual interfaces.) The drivers for these devices have no | |
62052ab1 | 784 | need of runtime PM callbacks; if the callbacks did exist, ->runtime_suspend() |
7490e442 AS |
785 | and ->runtime_resume() would always return 0 without doing anything else and |
786 | ->runtime_idle() would always call pm_runtime_suspend(). | |
787 | ||
788 | Subsystems can tell the PM core about these devices by calling | |
789 | pm_runtime_no_callbacks(). This should be done after the device structure is | |
790 | initialized and before it is registered (although after device registration is | |
791 | also okay). The routine will set the device's power.no_callbacks flag and | |
62052ab1 | 792 | prevent the non-debugging runtime PM sysfs attributes from being created. |
7490e442 AS |
793 | |
794 | When power.no_callbacks is set, the PM core will not invoke the | |
795 | ->runtime_idle(), ->runtime_suspend(), or ->runtime_resume() callbacks. | |
796 | Instead it will assume that suspends and resumes always succeed and that idle | |
797 | devices should be suspended. | |
798 | ||
799 | As a consequence, the PM core will never directly inform the device's subsystem | |
62052ab1 | 800 | or driver about runtime power changes. Instead, the driver for the device's |
7490e442 AS |
801 | parent must take responsibility for telling the device's driver when the |
802 | parent's power state changes. | |
15bcb91d AS |
803 | |
804 | 9. Autosuspend, or automatically-delayed suspends | |
805 | ||
806 | Changing a device's power state isn't free; it requires both time and energy. | |
807 | A device should be put in a low-power state only when there's some reason to | |
808 | think it will remain in that state for a substantial time. A common heuristic | |
809 | says that a device which hasn't been used for a while is liable to remain | |
810 | unused; following this advice, drivers should not allow devices to be suspended | |
62052ab1 | 811 | at runtime until they have been inactive for some minimum period. Even when |
15bcb91d AS |
812 | the heuristic ends up being non-optimal, it will still prevent devices from |
813 | "bouncing" too rapidly between low-power and full-power states. | |
814 | ||
815 | The term "autosuspend" is an historical remnant. It doesn't mean that the | |
816 | device is automatically suspended (the subsystem or driver still has to call | |
62052ab1 | 817 | the appropriate PM routines); rather it means that runtime suspends will |
15bcb91d AS |
818 | automatically be delayed until the desired period of inactivity has elapsed. |
819 | ||
820 | Inactivity is determined based on the power.last_busy field. Drivers should | |
821 | call pm_runtime_mark_last_busy() to update this field after carrying out I/O, | |
822 | typically just before calling pm_runtime_put_autosuspend(). The desired length | |
823 | of the inactivity period is a matter of policy. Subsystems can set this length | |
824 | initially by calling pm_runtime_set_autosuspend_delay(), but after device | |
825 | registration the length should be controlled by user space, using the | |
826 | /sys/devices/.../power/autosuspend_delay_ms attribute. | |
827 | ||
828 | In order to use autosuspend, subsystems or drivers must call | |
829 | pm_runtime_use_autosuspend() (preferably before registering the device), and | |
830 | thereafter they should use the various *_autosuspend() helper functions instead | |
831 | of the non-autosuspend counterparts: | |
832 | ||
833 | Instead of: pm_runtime_suspend use: pm_runtime_autosuspend; | |
834 | Instead of: pm_schedule_suspend use: pm_request_autosuspend; | |
835 | Instead of: pm_runtime_put use: pm_runtime_put_autosuspend; | |
836 | Instead of: pm_runtime_put_sync use: pm_runtime_put_sync_autosuspend. | |
837 | ||
838 | Drivers may also continue to use the non-autosuspend helper functions; they | |
839 | will behave normally, not taking the autosuspend delay into account. | |
840 | Similarly, if the power.use_autosuspend field isn't set then the autosuspend | |
841 | helper functions will behave just like the non-autosuspend counterparts. | |
842 | ||
886486b7 AS |
843 | Under some circumstances a driver or subsystem may want to prevent a device |
844 | from autosuspending immediately, even though the usage counter is zero and the | |
845 | autosuspend delay time has expired. If the ->runtime_suspend() callback | |
846 | returns -EAGAIN or -EBUSY, and if the next autosuspend delay expiration time is | |
847 | in the future (as it normally would be if the callback invoked | |
848 | pm_runtime_mark_last_busy()), the PM core will automatically reschedule the | |
849 | autosuspend. The ->runtime_suspend() callback can't do this rescheduling | |
850 | itself because no suspend requests of any kind are accepted while the device is | |
851 | suspending (i.e., while the callback is running). | |
852 | ||
15bcb91d AS |
853 | The implementation is well suited for asynchronous use in interrupt contexts. |
854 | However such use inevitably involves races, because the PM core can't | |
855 | synchronize ->runtime_suspend() callbacks with the arrival of I/O requests. | |
856 | This synchronization must be handled by the driver, using its private lock. | |
857 | Here is a schematic pseudo-code example: | |
858 | ||
859 | foo_read_or_write(struct foo_priv *foo, void *data) | |
860 | { | |
861 | lock(&foo->private_lock); | |
862 | add_request_to_io_queue(foo, data); | |
863 | if (foo->num_pending_requests++ == 0) | |
864 | pm_runtime_get(&foo->dev); | |
865 | if (!foo->is_suspended) | |
866 | foo_process_next_request(foo); | |
867 | unlock(&foo->private_lock); | |
868 | } | |
869 | ||
870 | foo_io_completion(struct foo_priv *foo, void *req) | |
871 | { | |
872 | lock(&foo->private_lock); | |
873 | if (--foo->num_pending_requests == 0) { | |
874 | pm_runtime_mark_last_busy(&foo->dev); | |
875 | pm_runtime_put_autosuspend(&foo->dev); | |
876 | } else { | |
877 | foo_process_next_request(foo); | |
878 | } | |
879 | unlock(&foo->private_lock); | |
880 | /* Send req result back to the user ... */ | |
881 | } | |
882 | ||
883 | int foo_runtime_suspend(struct device *dev) | |
884 | { | |
885 | struct foo_priv foo = container_of(dev, ...); | |
886 | int ret = 0; | |
887 | ||
888 | lock(&foo->private_lock); | |
889 | if (foo->num_pending_requests > 0) { | |
890 | ret = -EBUSY; | |
891 | } else { | |
892 | /* ... suspend the device ... */ | |
893 | foo->is_suspended = 1; | |
894 | } | |
895 | unlock(&foo->private_lock); | |
896 | return ret; | |
897 | } | |
898 | ||
899 | int foo_runtime_resume(struct device *dev) | |
900 | { | |
901 | struct foo_priv foo = container_of(dev, ...); | |
902 | ||
903 | lock(&foo->private_lock); | |
904 | /* ... resume the device ... */ | |
905 | foo->is_suspended = 0; | |
906 | pm_runtime_mark_last_busy(&foo->dev); | |
907 | if (foo->num_pending_requests > 0) | |
fe982450 | 908 | foo_process_next_request(foo); |
15bcb91d AS |
909 | unlock(&foo->private_lock); |
910 | return 0; | |
911 | } | |
912 | ||
913 | The important point is that after foo_io_completion() asks for an autosuspend, | |
914 | the foo_runtime_suspend() callback may race with foo_read_or_write(). | |
915 | Therefore foo_runtime_suspend() has to check whether there are any pending I/O | |
916 | requests (while holding the private lock) before allowing the suspend to | |
917 | proceed. | |
918 | ||
919 | In addition, the power.autosuspend_delay field can be changed by user space at | |
920 | any time. If a driver cares about this, it can call | |
921 | pm_runtime_autosuspend_expiration() from within the ->runtime_suspend() | |
922 | callback while holding its private lock. If the function returns a nonzero | |
923 | value then the delay has not yet expired and the callback should return | |
924 | -EAGAIN. |