Documentation/power/runtime_pm.txt

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