Merge branch 'pm-sleep'

* pm-sleep:
  PM / Freezer: Remove references to TIF_FREEZE in comments
  PM / Sleep: Add more wakeup source initialization routines
  PM / Hibernate: Enable usermodehelpers in hibernate() error path
  PM / Sleep: Make __pm_stay_awake() delete wakeup source timers
  PM / Sleep: Fix race conditions related to wakeup source timer function
  PM / Sleep: Fix possible infinite loop during wakeup source destruction
  PM / Hibernate: print physical addresses consistently with other parts of kernel
  PM: Add comment describing relationships between PM callbacks to pm.h
  PM / Sleep: Drop suspend_stats_update()
  PM / Sleep: Make enter_state() in kernel/power/suspend.c static
  PM / Sleep: Unify kerneldoc comments in kernel/power/suspend.c
  PM / Sleep: Remove unnecessary label from suspend_freeze_processes()
  PM / Sleep: Do not check wakeup too often in try_to_freeze_tasks()
  PM / Sleep: Initialize wakeup source locks in wakeup_source_add()
  PM / Hibernate: Refactor and simplify freezer_test_done
  PM / Hibernate: Thaw kernel threads in hibernation_snapshot() in error/test path
  PM / Freezer / Docs: Document the beauty of freeze/thaw semantics
  PM / Suspend: Avoid code duplication in suspend statistics update
  PM / Sleep: Introduce generic callbacks for new device PM phases
  PM / Sleep: Introduce "late suspend" and "early resume" of devices

Documentation/power/devices.txt

@@ -96,6 +96,12 @@ struct dev_pm_ops {
 	int (*thaw)(struct device *dev);
 	int (*poweroff)(struct device *dev);
 	int (*restore)(struct device *dev);
+	int (*suspend_late)(struct device *dev);
+	int (*resume_early)(struct device *dev);
+	int (*freeze_late)(struct device *dev);
+	int (*thaw_early)(struct device *dev);
+	int (*poweroff_late)(struct device *dev);
+	int (*restore_early)(struct device *dev);
 	int (*suspend_noirq)(struct device *dev);
 	int (*resume_noirq)(struct device *dev);
 	int (*freeze_noirq)(struct device *dev);
@@ -305,7 +311,7 @@ Entering System Suspend
 -----------------------
 When the system goes into the standby or memory sleep state, the phases are:
 
-		prepare, suspend, suspend_noirq.
+		prepare, suspend, suspend_late, suspend_noirq.
 
     1.	The prepare phase is meant to prevent races by preventing new devices
 	from being registered; the PM core would never know that all the
@@ -324,7 +330,12 @@ When the system goes into the standby or memory sleep state, the phases are:
 	appropriate low-power state, depending on the bus type the device is on,
 	and they may enable wakeup events.
 
-    3.	The suspend_noirq phase occurs after IRQ handlers have been disabled,
+    3	For a number of devices it is convenient to split suspend into the
+	"quiesce device" and "save device state" phases, in which cases
+	suspend_late is meant to do the latter.  It is always executed after
+	runtime power management has been disabled for all devices.
+
+    4.	The suspend_noirq phase occurs after IRQ handlers have been disabled,
 	which means that the driver's interrupt handler will not be called while
 	the callback method is running.  The methods should save the values of
 	the device's registers that weren't saved previously and finally put the
@@ -359,7 +370,7 @@ Leaving System Suspend
 ----------------------
 When resuming from standby or memory sleep, the phases are:
 
-		resume_noirq, resume, complete.
+		resume_noirq, resume_early, resume, complete.
 
     1.	The resume_noirq callback methods should perform any actions needed
 	before the driver's interrupt handlers are invoked.  This generally
@@ -375,14 +386,18 @@ When resuming from standby or memory sleep, the phases are:
 	device driver's ->pm.resume_noirq() method to perform device-specific
 	actions.
 
-    2.	The resume methods should bring the the device back to its operating
+    2.	The resume_early methods should prepare devices for the execution of
+	the resume methods.  This generally involves undoing the actions of the
+	preceding suspend_late phase.
+
+    3	The resume methods should bring the the device back to its operating
 	state, so that it can perform normal I/O.  This generally involves
 	undoing the actions of the suspend phase.
 
-    3.	The complete phase uses only a bus callback.  The method should undo the
+    4.	The complete phase should undo the actions of the prepare phase.  Note,
-	actions of the prepare phase.  Note, however, that new children may be
+	however, that new children may be registered below the device as soon as
-	registered below the device as soon as the resume callbacks occur; it's
+	the resume callbacks occur; it's not necessary to wait until the
-	not necessary to wait until the complete phase.
+	complete phase.
 
 At the end of these phases, drivers should be as functional as they were before
 suspending: I/O can be performed using DMA and IRQs, and the relevant clocks are
@@ -429,8 +444,8 @@ an image of the system memory while everything is stable, reactivate all
 devices (thaw), write the image to permanent storage, and finally shut down the
 system (poweroff).  The phases used to accomplish this are:
 
-	prepare, freeze, freeze_noirq, thaw_noirq, thaw, complete,
+	prepare, freeze, freeze_late, freeze_noirq, thaw_noirq, thaw_early,
-	prepare, poweroff, poweroff_noirq
+	thaw, complete, prepare, poweroff, poweroff_late, poweroff_noirq
 
     1.	The prepare phase is discussed in the "Entering System Suspend" section
 	above.
@@ -441,7 +456,11 @@ system (poweroff).  The phases used to accomplish this are:
 	save time it's best not to do so.  Also, the device should not be
 	prepared to generate wakeup events.
 
-    3.	The freeze_noirq phase is analogous to the suspend_noirq phase discussed
+    3.	The freeze_late phase is analogous to the suspend_late phase described
+	above, except that the device should not be put in a low-power state and
+	should not be allowed to generate wakeup events by it.
+
+    4.	The freeze_noirq phase is analogous to the suspend_noirq phase discussed
 	above, except again that the device should not be put in a low-power
 	state and should not be allowed to generate wakeup events.
 
@@ -449,15 +468,19 @@ At this point the system image is created.  All devices should be inactive and
 the contents of memory should remain undisturbed while this happens, so that the
 image forms an atomic snapshot of the system state.
 
-    4.	The thaw_noirq phase is analogous to the resume_noirq phase discussed
+    5.	The thaw_noirq phase is analogous to the resume_noirq phase discussed
 	above.  The main difference is that its methods can assume the device is
 	in the same state as at the end of the freeze_noirq phase.
 
-    5.	The thaw phase is analogous to the resume phase discussed above.  Its
+    6.	The thaw_early phase is analogous to the resume_early phase described
+	above.  Its methods should undo the actions of the preceding
+	freeze_late, if necessary.
+
+    7.	The thaw phase is analogous to the resume phase discussed above.  Its
 	methods should bring the device back to an operating state, so that it
 	can be used for saving the image if necessary.
 
-    6.	The complete phase is discussed in the "Leaving System Suspend" section
+    8.	The complete phase is discussed in the "Leaving System Suspend" section
 	above.
 
 At this point the system image is saved, and the devices then need to be
@@ -465,16 +488,19 @@ prepared for the upcoming system shutdown.  This is much like suspending them
 before putting the system into the standby or memory sleep state, and the phases
 are similar.
 
-    7.	The prepare phase is discussed above.
+    9.	The prepare phase is discussed above.
 
-    8.	The poweroff phase is analogous to the suspend phase.
+    10.	The poweroff phase is analogous to the suspend phase.
 
-    9.	The poweroff_noirq phase is analogous to the suspend_noirq phase.
+    11.	The poweroff_late phase is analogous to the suspend_late phase.
 
-The poweroff and poweroff_noirq callbacks should do essentially the same things
+    12.	The poweroff_noirq phase is analogous to the suspend_noirq phase.
-as the suspend and suspend_noirq callbacks.  The only notable difference is that
+
-they need not store the device register values, because the registers should
+The poweroff, poweroff_late and poweroff_noirq callbacks should do essentially
-already have been stored during the freeze or freeze_noirq phases.
+the same things as the suspend, suspend_late and suspend_noirq callbacks,
+respectively.  The only notable difference is that they need not store the
+device register values, because the registers should already have been stored
+during the freeze, freeze_late or freeze_noirq phases.
 
 
 Leaving Hibernation
@@ -518,22 +544,25 @@ To achieve this, the image kernel must restore the devices' pre-hibernation
 functionality.  The operation is much like waking up from the memory sleep
 state, although it involves different phases:
 
-	restore_noirq, restore, complete
+	restore_noirq, restore_early, restore, complete
 
     1.	The restore_noirq phase is analogous to the resume_noirq phase.
 
-    2.	The restore phase is analogous to the resume phase.
+    2.	The restore_early phase is analogous to the resume_early phase.
 
-    3.	The complete phase is discussed above.
+    3.	The restore phase is analogous to the resume phase.
 
-The main difference from resume[_noirq] is that restore[_noirq] must assume the
+    4.	The complete phase is discussed above.
-device has been accessed and reconfigured by the boot loader or the boot kernel.
+
-Consequently the state of the device may be different from the state remembered
+The main difference from resume[_early|_noirq] is that restore[_early|_noirq]
-from the freeze and freeze_noirq phases.  The device may even need to be reset
+must assume the device has been accessed and reconfigured by the boot loader or
-and completely re-initialized.  In many cases this difference doesn't matter, so
+the boot kernel.  Consequently the state of the device may be different from the
-the resume[_noirq] and restore[_norq] method pointers can be set to the same
+state remembered from the freeze, freeze_late and freeze_noirq phases.  The
-routines.  Nevertheless, different callback pointers are used in case there is a
+device may even need to be reset and completely re-initialized.  In many cases
-situation where it actually matters.
+this difference doesn't matter, so the resume[_early|_noirq] and
+restore[_early|_norq] method pointers can be set to the same routines.
+Nevertheless, different callback pointers are used in case there is a situation
+where it actually does matter.
 
 
 Device Power Management Domains

Documentation/power/freezing-of-tasks.txt

@@ -63,6 +63,27 @@ devices have been reinitialized, the function thaw_processes() is called in
 order to clear the PF_FROZEN flag for each frozen task.  Then, the tasks that
 have been frozen leave __refrigerator() and continue running.
 
+
+Rationale behind the functions dealing with freezing and thawing of tasks:
+-------------------------------------------------------------------------
+
+freeze_processes():
+  - freezes only userspace tasks
+
+freeze_kernel_threads():
+  - freezes all tasks (including kernel threads) because we can't freeze
+    kernel threads without freezing userspace tasks
+
+thaw_kernel_threads():
+  - thaws only kernel threads; this is particularly useful if we need to do
+    anything special in between thawing of kernel threads and thawing of
+    userspace tasks, or if we want to postpone the thawing of userspace tasks
+
+thaw_processes():
+  - thaws all tasks (including kernel threads) because we can't thaw userspace
+    tasks without thawing kernel threads
+
+
 III. Which kernel threads are freezable?
 
 Kernel threads are not freezable by default.  However, a kernel thread may clear

arch/x86/kernel/apm_32.c

@@ -1234,8 +1234,7 @@ static int suspend(int vetoable)
 	struct apm_user	*as;
 
 	dpm_suspend_start(PMSG_SUSPEND);
-
+	dpm_suspend_end(PMSG_SUSPEND);
-	dpm_suspend_noirq(PMSG_SUSPEND);
 
 	local_irq_disable();
 	syscore_suspend();
@@ -1259,9 +1258,9 @@ static int suspend(int vetoable)
 	syscore_resume();
 	local_irq_enable();
 
-	dpm_resume_noirq(PMSG_RESUME);
+	dpm_resume_start(PMSG_RESUME);
-
 	dpm_resume_end(PMSG_RESUME);
+
 	queue_event(APM_NORMAL_RESUME, NULL);
 	spin_lock(&user_list_lock);
 	for (as = user_list; as != NULL; as = as->next) {
@@ -1277,7 +1276,7 @@ static void standby(void)
 {
 	int err;
 
-	dpm_suspend_noirq(PMSG_SUSPEND);
+	dpm_suspend_end(PMSG_SUSPEND);
 
 	local_irq_disable();
 	syscore_suspend();
@@ -1291,7 +1290,7 @@ static void standby(void)
 	syscore_resume();
 	local_irq_enable();
 
-	dpm_resume_noirq(PMSG_RESUME);
+	dpm_resume_start(PMSG_RESUME);
 }
 
 static apm_event_t get_event(void)

drivers/base/power/generic_ops.c

@@ -91,68 +91,39 @@ int pm_generic_prepare(struct device *dev)
 	return ret;
 }
 
-/**
- * __pm_generic_call - Generic suspend/freeze/poweroff/thaw subsystem callback.
- * @dev: Device to handle.
- * @event: PM transition of the system under way.
- * @bool: Whether or not this is the "noirq" stage.
- *
- * Execute the PM callback corresponding to @event provided by the driver of
- * @dev, if defined, and return its error code.    Return 0 if the callback is
- * not present.
- */
-static int __pm_generic_call(struct device *dev, int event, bool noirq)
-{
-	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
-	int (*callback)(struct device *);
-
-	if (!pm)
-		return 0;
-
-	switch (event) {
-	case PM_EVENT_SUSPEND:
-		callback = noirq ? pm->suspend_noirq : pm->suspend;
-		break;
-	case PM_EVENT_FREEZE:
-		callback = noirq ? pm->freeze_noirq : pm->freeze;
-		break;
-	case PM_EVENT_HIBERNATE:
-		callback = noirq ? pm->poweroff_noirq : pm->poweroff;
-		break;
-	case PM_EVENT_RESUME:
-		callback = noirq ? pm->resume_noirq : pm->resume;
-		break;
-	case PM_EVENT_THAW:
-		callback = noirq ? pm->thaw_noirq : pm->thaw;
-		break;
-	case PM_EVENT_RESTORE:
-		callback = noirq ? pm->restore_noirq : pm->restore;
-		break;
-	default:
-		callback = NULL;
-		break;
-	}
-
-	return callback ? callback(dev) : 0;
-}
-
 /**
  * pm_generic_suspend_noirq - Generic suspend_noirq callback for subsystems.
  * @dev: Device to suspend.
  */
 int pm_generic_suspend_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_SUSPEND, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->suspend_noirq ? pm->suspend_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_suspend_noirq);
 
+/**
+ * pm_generic_suspend_late - Generic suspend_late callback for subsystems.
+ * @dev: Device to suspend.
+ */
+int pm_generic_suspend_late(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->suspend_late ? pm->suspend_late(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_suspend_late);
+
 /**
  * pm_generic_suspend - Generic suspend callback for subsystems.
  * @dev: Device to suspend.
  */
 int pm_generic_suspend(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_SUSPEND, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->suspend ? pm->suspend(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_suspend);
 
@@ -162,17 +133,33 @@ EXPORT_SYMBOL_GPL(pm_generic_suspend);
  */
 int pm_generic_freeze_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_FREEZE, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->freeze_noirq ? pm->freeze_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_freeze_noirq);
 
+/**
+ * pm_generic_freeze_late - Generic freeze_late callback for subsystems.
+ * @dev: Device to freeze.
+ */
+int pm_generic_freeze_late(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->freeze_late ? pm->freeze_late(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_freeze_late);
+
 /**
  * pm_generic_freeze - Generic freeze callback for subsystems.
  * @dev: Device to freeze.
  */
 int pm_generic_freeze(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_FREEZE, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->freeze ? pm->freeze(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_freeze);
 
@@ -182,17 +169,33 @@ EXPORT_SYMBOL_GPL(pm_generic_freeze);
  */
 int pm_generic_poweroff_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_HIBERNATE, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->poweroff_noirq ? pm->poweroff_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_poweroff_noirq);
 
+/**
+ * pm_generic_poweroff_late - Generic poweroff_late callback for subsystems.
+ * @dev: Device to handle.
+ */
+int pm_generic_poweroff_late(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->poweroff_late ? pm->poweroff_late(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_poweroff_late);
+
 /**
  * pm_generic_poweroff - Generic poweroff callback for subsystems.
  * @dev: Device to handle.
  */
 int pm_generic_poweroff(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_HIBERNATE, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->poweroff ? pm->poweroff(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_poweroff);
 
@@ -202,17 +205,33 @@ EXPORT_SYMBOL_GPL(pm_generic_poweroff);
  */
 int pm_generic_thaw_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_THAW, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->thaw_noirq ? pm->thaw_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_thaw_noirq);
 
+/**
+ * pm_generic_thaw_early - Generic thaw_early callback for subsystems.
+ * @dev: Device to thaw.
+ */
+int pm_generic_thaw_early(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->thaw_early ? pm->thaw_early(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_thaw_early);
+
 /**
  * pm_generic_thaw - Generic thaw callback for subsystems.
  * @dev: Device to thaw.
  */
 int pm_generic_thaw(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_THAW, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->thaw ? pm->thaw(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_thaw);
 
@@ -222,17 +241,33 @@ EXPORT_SYMBOL_GPL(pm_generic_thaw);
  */
 int pm_generic_resume_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_RESUME, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->resume_noirq ? pm->resume_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_resume_noirq);
 
+/**
+ * pm_generic_resume_early - Generic resume_early callback for subsystems.
+ * @dev: Device to resume.
+ */
+int pm_generic_resume_early(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->resume_early ? pm->resume_early(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_resume_early);
+
 /**
  * pm_generic_resume - Generic resume callback for subsystems.
  * @dev: Device to resume.
  */
 int pm_generic_resume(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_RESUME, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->resume ? pm->resume(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_resume);
 
@@ -242,17 +277,33 @@ EXPORT_SYMBOL_GPL(pm_generic_resume);
  */
 int pm_generic_restore_noirq(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_RESTORE, true);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->restore_noirq ? pm->restore_noirq(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_restore_noirq);
 
+/**
+ * pm_generic_restore_early - Generic restore_early callback for subsystems.
+ * @dev: Device to resume.
+ */
+int pm_generic_restore_early(struct device *dev)
+{
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->restore_early ? pm->restore_early(dev) : 0;
+}
+EXPORT_SYMBOL_GPL(pm_generic_restore_early);
+
 /**
  * pm_generic_restore - Generic restore callback for subsystems.
  * @dev: Device to restore.
  */
 int pm_generic_restore(struct device *dev)
 {
-	return __pm_generic_call(dev, PM_EVENT_RESTORE, false);
+	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+	return pm && pm->restore ? pm->restore(dev) : 0;
 }
 EXPORT_SYMBOL_GPL(pm_generic_restore);
 

drivers/base/power/main.c

@@ -47,6 +47,7 @@ typedef int (*pm_callback_t)(struct device *);
 LIST_HEAD(dpm_list);
 LIST_HEAD(dpm_prepared_list);
 LIST_HEAD(dpm_suspended_list);
+LIST_HEAD(dpm_late_early_list);
 LIST_HEAD(dpm_noirq_list);
 
 struct suspend_stats suspend_stats;
@@ -245,6 +246,40 @@ static pm_callback_t pm_op(const struct dev_pm_ops *ops, pm_message_t state)
 	return NULL;
 }
 
+/**
+ * pm_late_early_op - Return the PM operation appropriate for given PM event.
+ * @ops: PM operations to choose from.
+ * @state: PM transition of the system being carried out.
+ *
+ * Runtime PM is disabled for @dev while this function is being executed.
+ */
+static pm_callback_t pm_late_early_op(const struct dev_pm_ops *ops,
+				      pm_message_t state)
+{
+	switch (state.event) {
+#ifdef CONFIG_SUSPEND
+	case PM_EVENT_SUSPEND:
+		return ops->suspend_late;
+	case PM_EVENT_RESUME:
+		return ops->resume_early;
+#endif /* CONFIG_SUSPEND */
+#ifdef CONFIG_HIBERNATE_CALLBACKS
+	case PM_EVENT_FREEZE:
+	case PM_EVENT_QUIESCE:
+		return ops->freeze_late;
+	case PM_EVENT_HIBERNATE:
+		return ops->poweroff_late;
+	case PM_EVENT_THAW:
+	case PM_EVENT_RECOVER:
+		return ops->thaw_early;
+	case PM_EVENT_RESTORE:
+		return ops->restore_early;
+#endif /* CONFIG_HIBERNATE_CALLBACKS */
+	}
+
+	return NULL;
+}
+
 /**
  * pm_noirq_op - Return the PM operation appropriate for given PM event.
  * @ops: PM operations to choose from.
@@ -374,21 +409,21 @@ static int device_resume_noirq(struct device *dev, pm_message_t state)
 	TRACE_RESUME(0);
 
 	if (dev->pm_domain) {
-		info = "EARLY power domain ";
+		info = "noirq power domain ";
 		callback = pm_noirq_op(&dev->pm_domain->ops, state);
 	} else if (dev->type && dev->type->pm) {
-		info = "EARLY type ";
+		info = "noirq type ";
 		callback = pm_noirq_op(dev->type->pm, state);
 	} else if (dev->class && dev->class->pm) {
-		info = "EARLY class ";
+		info = "noirq class ";
 		callback = pm_noirq_op(dev->class->pm, state);
 	} else if (dev->bus && dev->bus->pm) {
-		info = "EARLY bus ";
+		info = "noirq bus ";
 		callback = pm_noirq_op(dev->bus->pm, state);
 	}
 
 	if (!callback && dev->driver && dev->driver->pm) {
-		info = "EARLY driver ";
+		info = "noirq driver ";
 		callback = pm_noirq_op(dev->driver->pm, state);
 	}
 
@@ -399,13 +434,13 @@ static int device_resume_noirq(struct device *dev, pm_message_t state)
 }
 
 /**
- * dpm_resume_noirq - Execute "early resume" callbacks for non-sysdev devices.
+ * dpm_resume_noirq - Execute "noirq resume" callbacks for all devices.
  * @state: PM transition of the system being carried out.
  *
- * Call the "noirq" resume handlers for all devices marked as DPM_OFF_IRQ and
+ * Call the "noirq" resume handlers for all devices in dpm_noirq_list and
  * enable device drivers to receive interrupts.
  */
-void dpm_resume_noirq(pm_message_t state)
+static void dpm_resume_noirq(pm_message_t state)
 {
 	ktime_t starttime = ktime_get();
 
@@ -415,7 +450,7 @@ void dpm_resume_noirq(pm_message_t state)
 		int error;
 
 		get_device(dev);
-		list_move_tail(&dev->power.entry, &dpm_suspended_list);
+		list_move_tail(&dev->power.entry, &dpm_late_early_list);
 		mutex_unlock(&dpm_list_mtx);
 
 		error = device_resume_noirq(dev, state);
@@ -423,6 +458,80 @@ void dpm_resume_noirq(pm_message_t state)
 			suspend_stats.failed_resume_noirq++;
 			dpm_save_failed_step(SUSPEND_RESUME_NOIRQ);
 			dpm_save_failed_dev(dev_name(dev));
+			pm_dev_err(dev, state, " noirq", error);
+		}
+
+		mutex_lock(&dpm_list_mtx);
+		put_device(dev);
+	}
+	mutex_unlock(&dpm_list_mtx);
+	dpm_show_time(starttime, state, "noirq");
+	resume_device_irqs();
+}
+
+/**
+ * device_resume_early - Execute an "early resume" callback for given device.
+ * @dev: Device to handle.
+ * @state: PM transition of the system being carried out.
+ *
+ * Runtime PM is disabled for @dev while this function is being executed.
+ */
+static int device_resume_early(struct device *dev, pm_message_t state)
+{
+	pm_callback_t callback = NULL;
+	char *info = NULL;
+	int error = 0;
+
+	TRACE_DEVICE(dev);
+	TRACE_RESUME(0);
+
+	if (dev->pm_domain) {
+		info = "early power domain ";
+		callback = pm_late_early_op(&dev->pm_domain->ops, state);
+	} else if (dev->type && dev->type->pm) {
+		info = "early type ";
+		callback = pm_late_early_op(dev->type->pm, state);
+	} else if (dev->class && dev->class->pm) {
+		info = "early class ";
+		callback = pm_late_early_op(dev->class->pm, state);
+	} else if (dev->bus && dev->bus->pm) {
+		info = "early bus ";
+		callback = pm_late_early_op(dev->bus->pm, state);
+	}
+
+	if (!callback && dev->driver && dev->driver->pm) {
+		info = "early driver ";
+		callback = pm_late_early_op(dev->driver->pm, state);
+	}
+
+	error = dpm_run_callback(callback, dev, state, info);
+
+	TRACE_RESUME(error);
+	return error;
+}
+
+/**
+ * dpm_resume_early - Execute "early resume" callbacks for all devices.
+ * @state: PM transition of the system being carried out.
+ */
+static void dpm_resume_early(pm_message_t state)
+{
+	ktime_t starttime = ktime_get();
+
+	mutex_lock(&dpm_list_mtx);
+	while (!list_empty(&dpm_late_early_list)) {
+		struct device *dev = to_device(dpm_late_early_list.next);
+		int error;
+
+		get_device(dev);
+		list_move_tail(&dev->power.entry, &dpm_suspended_list);
+		mutex_unlock(&dpm_list_mtx);
+
+		error = device_resume_early(dev, state);
+		if (error) {
+			suspend_stats.failed_resume_early++;
+			dpm_save_failed_step(SUSPEND_RESUME_EARLY);
+			dpm_save_failed_dev(dev_name(dev));
 			pm_dev_err(dev, state, " early", error);
 		}
 
@@ -431,9 +540,18 @@ void dpm_resume_noirq(pm_message_t state)
 	}
 	mutex_unlock(&dpm_list_mtx);
 	dpm_show_time(starttime, state, "early");
-	resume_device_irqs();
 }
-EXPORT_SYMBOL_GPL(dpm_resume_noirq);
+
+/**
+ * dpm_resume_start - Execute "noirq" and "early" device callbacks.
+ * @state: PM transition of the system being carried out.
+ */
+void dpm_resume_start(pm_message_t state)
+{
+	dpm_resume_noirq(state);
+	dpm_resume_early(state);
+}
+EXPORT_SYMBOL_GPL(dpm_resume_start);
 
 /**
  * device_resume - Execute "resume" callbacks for given device.
@@ -716,21 +834,21 @@ static int device_suspend_noirq(struct device *dev, pm_message_t state)
 	char *info = NULL;
 
 	if (dev->pm_domain) {
-		info = "LATE power domain ";
+		info = "noirq power domain ";
 		callback = pm_noirq_op(&dev->pm_domain->ops, state);
 	} else if (dev->type && dev->type->pm) {
-		info = "LATE type ";
+		info = "noirq type ";
 		callback = pm_noirq_op(dev->type->pm, state);
 	} else if (dev->class && dev->class->pm) {
-		info = "LATE class ";
+		info = "noirq class ";
 		callback = pm_noirq_op(dev->class->pm, state);
 	} else if (dev->bus && dev->bus->pm) {
-		info = "LATE bus ";
+		info = "noirq bus ";
 		callback = pm_noirq_op(dev->bus->pm, state);
 	}
 
 	if (!callback && dev->driver && dev->driver->pm) {
-		info = "LATE driver ";
+		info = "noirq driver ";
 		callback = pm_noirq_op(dev->driver->pm, state);
 	}
 
@@ -738,21 +856,21 @@ static int device_suspend_noirq(struct device *dev, pm_message_t state)
 }
 
 /**
- * dpm_suspend_noirq - Execute "late suspend" callbacks for non-sysdev devices.
+ * dpm_suspend_noirq - Execute "noirq suspend" callbacks for all devices.
  * @state: PM transition of the system being carried out.
  *
  * Prevent device drivers from receiving interrupts and call the "noirq" suspend
  * handlers for all non-sysdev devices.
  */
-int dpm_suspend_noirq(pm_message_t state)
+static int dpm_suspend_noirq(pm_message_t state)
 {
 	ktime_t starttime = ktime_get();
 	int error = 0;
 
 	suspend_device_irqs();
 	mutex_lock(&dpm_list_mtx);
-	while (!list_empty(&dpm_suspended_list)) {
+	while (!list_empty(&dpm_late_early_list)) {
-		struct device *dev = to_device(dpm_suspended_list.prev);
+		struct device *dev = to_device(dpm_late_early_list.prev);
 
 		get_device(dev);
 		mutex_unlock(&dpm_list_mtx);
@@ -761,7 +879,7 @@ int dpm_suspend_noirq(pm_message_t state)
 
 		mutex_lock(&dpm_list_mtx);
 		if (error) {
-			pm_dev_err(dev, state, " late", error);
+			pm_dev_err(dev, state, " noirq", error);
 			suspend_stats.failed_suspend_noirq++;
 			dpm_save_failed_step(SUSPEND_SUSPEND_NOIRQ);
 			dpm_save_failed_dev(dev_name(dev));
@@ -776,10 +894,95 @@ int dpm_suspend_noirq(pm_message_t state)
 	if (error)
 		dpm_resume_noirq(resume_event(state));
 	else
-		dpm_show_time(starttime, state, "late");
+		dpm_show_time(starttime, state, "noirq");
 	return error;
 }
-EXPORT_SYMBOL_GPL(dpm_suspend_noirq);
+
+/**
+ * device_suspend_late - Execute a "late suspend" callback for given device.
+ * @dev: Device to handle.
+ * @state: PM transition of the system being carried out.
+ *
+ * Runtime PM is disabled for @dev while this function is being executed.
+ */
+static int device_suspend_late(struct device *dev, pm_message_t state)
+{
+	pm_callback_t callback = NULL;
+	char *info = NULL;
+
+	if (dev->pm_domain) {
+		info = "late power domain ";
+		callback = pm_late_early_op(&dev->pm_domain->ops, state);
+	} else if (dev->type && dev->type->pm) {
+		info = "late type ";
+		callback = pm_late_early_op(dev->type->pm, state);
+	} else if (dev->class && dev->class->pm) {
+		info = "late class ";
+		callback = pm_late_early_op(dev->class->pm, state);
+	} else if (dev->bus && dev->bus->pm) {
+		info = "late bus ";
+		callback = pm_late_early_op(dev->bus->pm, state);
+	}
+
+	if (!callback && dev->driver && dev->driver->pm) {
+		info = "late driver ";
+		callback = pm_late_early_op(dev->driver->pm, state);
+	}
+
+	return dpm_run_callback(callback, dev, state, info);
+}
+
+/**
+ * dpm_suspend_late - Execute "late suspend" callbacks for all devices.
+ * @state: PM transition of the system being carried out.
+ */
+static int dpm_suspend_late(pm_message_t state)
+{
+	ktime_t starttime = ktime_get();
+	int error = 0;
+
+	mutex_lock(&dpm_list_mtx);
+	while (!list_empty(&dpm_suspended_list)) {
+		struct device *dev = to_device(dpm_suspended_list.prev);
+
+		get_device(dev);
+		mutex_unlock(&dpm_list_mtx);
+
+		error = device_suspend_late(dev, state);
+
+		mutex_lock(&dpm_list_mtx);
+		if (error) {
+			pm_dev_err(dev, state, " late", error);
+			suspend_stats.failed_suspend_late++;
+			dpm_save_failed_step(SUSPEND_SUSPEND_LATE);
+			dpm_save_failed_dev(dev_name(dev));
+			put_device(dev);
+			break;
+		}
+		if (!list_empty(&dev->power.entry))
+			list_move(&dev->power.entry, &dpm_late_early_list);
+		put_device(dev);
+	}
+	mutex_unlock(&dpm_list_mtx);
+	if (error)
+		dpm_resume_early(resume_event(state));
+	else
+		dpm_show_time(starttime, state, "late");
+
+	return error;
+}
+
+/**
+ * dpm_suspend_end - Execute "late" and "noirq" device suspend callbacks.
+ * @state: PM transition of the system being carried out.
+ */
+int dpm_suspend_end(pm_message_t state)
+{
+	int error = dpm_suspend_late(state);
+
+	return error ? : dpm_suspend_noirq(state);
+}
+EXPORT_SYMBOL_GPL(dpm_suspend_end);
 
 /**
  * legacy_suspend - Execute a legacy (bus or class) suspend callback for device.

drivers/base/power/wakeup.c

@@ -52,6 +52,23 @@ static void pm_wakeup_timer_fn(unsigned long data);
 
 static LIST_HEAD(wakeup_sources);
 
+/**
+ * wakeup_source_prepare - Prepare a new wakeup source for initialization.
+ * @ws: Wakeup source to prepare.
+ * @name: Pointer to the name of the new wakeup source.
+ *
+ * Callers must ensure that the @name string won't be freed when @ws is still in
+ * use.
+ */
+void wakeup_source_prepare(struct wakeup_source *ws, const char *name)
+{
+	if (ws) {
+		memset(ws, 0, sizeof(*ws));
+		ws->name = name;
+	}
+}
+EXPORT_SYMBOL_GPL(wakeup_source_prepare);
+
 /**
  * wakeup_source_create - Create a struct wakeup_source object.
  * @name: Name of the new wakeup source.
@@ -60,37 +77,44 @@ struct wakeup_source *wakeup_source_create(const char *name)
 {
 	struct wakeup_source *ws;
 
-	ws = kzalloc(sizeof(*ws), GFP_KERNEL);
+	ws = kmalloc(sizeof(*ws), GFP_KERNEL);
 	if (!ws)
 		return NULL;
 
-	spin_lock_init(&ws->lock);
+	wakeup_source_prepare(ws, name ? kstrdup(name, GFP_KERNEL) : NULL);
-	if (name)
-		ws->name = kstrdup(name, GFP_KERNEL);
-
 	return ws;
 }
 EXPORT_SYMBOL_GPL(wakeup_source_create);
 
+/**
+ * wakeup_source_drop - Prepare a struct wakeup_source object for destruction.
+ * @ws: Wakeup source to prepare for destruction.
+ *
+ * Callers must ensure that __pm_stay_awake() or __pm_wakeup_event() will never
+ * be run in parallel with this function for the same wakeup source object.
+ */
+void wakeup_source_drop(struct wakeup_source *ws)
+{
+	if (!ws)
+		return;
+
+	del_timer_sync(&ws->timer);
+	__pm_relax(ws);
+}
+EXPORT_SYMBOL_GPL(wakeup_source_drop);
+
 /**
  * wakeup_source_destroy - Destroy a struct wakeup_source object.
  * @ws: Wakeup source to destroy.
+ *
+ * Use only for wakeup source objects created with wakeup_source_create().
  */
 void wakeup_source_destroy(struct wakeup_source *ws)
 {
 	if (!ws)
 		return;
 
-	spin_lock_irq(&ws->lock);
+	wakeup_source_drop(ws);
-	while (ws->active) {
-		spin_unlock_irq(&ws->lock);
-
-		schedule_timeout_interruptible(msecs_to_jiffies(TIMEOUT));
-
-		spin_lock_irq(&ws->lock);
-	}
-	spin_unlock_irq(&ws->lock);
-
 	kfree(ws->name);
 	kfree(ws);
 }
@@ -105,6 +129,7 @@ void wakeup_source_add(struct wakeup_source *ws)
 	if (WARN_ON(!ws))
 		return;
 
+	spin_lock_init(&ws->lock);
 	setup_timer(&ws->timer, pm_wakeup_timer_fn, (unsigned long)ws);
 	ws->active = false;
 
@@ -152,8 +177,10 @@ EXPORT_SYMBOL_GPL(wakeup_source_register);
  */
 void wakeup_source_unregister(struct wakeup_source *ws)
 {
-	wakeup_source_remove(ws);
+	if (ws) {
-	wakeup_source_destroy(ws);
+		wakeup_source_remove(ws);
+		wakeup_source_destroy(ws);
+	}
 }
 EXPORT_SYMBOL_GPL(wakeup_source_unregister);
 
@@ -349,7 +376,6 @@ static void wakeup_source_activate(struct wakeup_source *ws)
 {
 	ws->active = true;
 	ws->active_count++;
-	ws->timer_expires = jiffies;
 	ws->last_time = ktime_get();
 
 	/* Increment the counter of events in progress. */
@@ -370,9 +396,14 @@ void __pm_stay_awake(struct wakeup_source *ws)
 		return;
 
 	spin_lock_irqsave(&ws->lock, flags);
+
 	ws->event_count++;
 	if (!ws->active)
 		wakeup_source_activate(ws);
+
+	del_timer(&ws->timer);
+	ws->timer_expires = 0;
+
 	spin_unlock_irqrestore(&ws->lock, flags);
 }
 EXPORT_SYMBOL_GPL(__pm_stay_awake);
@@ -438,6 +469,7 @@ static void wakeup_source_deactivate(struct wakeup_source *ws)
 		ws->max_time = duration;
 
 	del_timer(&ws->timer);
+	ws->timer_expires = 0;
 
 	/*
 	 * Increment the counter of registered wakeup events and decrement the
@@ -492,11 +524,22 @@ EXPORT_SYMBOL_GPL(pm_relax);
  * pm_wakeup_timer_fn - Delayed finalization of a wakeup event.
  * @data: Address of the wakeup source object associated with the event source.
  *
- * Call __pm_relax() for the wakeup source whose address is stored in @data.
+ * Call wakeup_source_deactivate() for the wakeup source whose address is stored
+ * in @data if it is currently active and its timer has not been canceled and
+ * the expiration time of the timer is not in future.
  */
 static void pm_wakeup_timer_fn(unsigned long data)
 {
-	__pm_relax((struct wakeup_source *)data);
+	struct wakeup_source *ws = (struct wakeup_source *)data;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ws->lock, flags);
+
+	if (ws->active && ws->timer_expires
+	    && time_after_eq(jiffies, ws->timer_expires))
+		wakeup_source_deactivate(ws);
+
+	spin_unlock_irqrestore(&ws->lock, flags);
 }
 
 /**
@@ -534,7 +577,7 @@ void __pm_wakeup_event(struct wakeup_source *ws, unsigned int msec)
 	if (!expires)
 		expires = 1;
 
-	if (time_after(expires, ws->timer_expires)) {
+	if (!ws->timer_expires || time_after(expires, ws->timer_expires)) {
 		mod_timer(&ws->timer, expires);
 		ws->timer_expires = expires;
 	}

drivers/xen/manage.c

@@ -129,9 +129,9 @@ static void do_suspend(void)
 	printk(KERN_DEBUG "suspending xenstore...\n");
 	xs_suspend();
 
-	err = dpm_suspend_noirq(PMSG_FREEZE);
+	err = dpm_suspend_end(PMSG_FREEZE);
 	if (err) {
-		printk(KERN_ERR "dpm_suspend_noirq failed: %d\n", err);
+		printk(KERN_ERR "dpm_suspend_end failed: %d\n", err);
 		goto out_resume;
 	}
 
@@ -149,7 +149,7 @@ static void do_suspend(void)
 
 	err = stop_machine(xen_suspend, &si, cpumask_of(0));
 
-	dpm_resume_noirq(si.cancelled ? PMSG_THAW : PMSG_RESTORE);
+	dpm_resume_start(si.cancelled ? PMSG_THAW : PMSG_RESTORE);
 
 	if (err) {
 		printk(KERN_ERR "failed to start xen_suspend: %d\n", err);

include/linux/pm.h

@@ -110,6 +110,10 @@ typedef struct pm_message {
  *	Subsystem-level @suspend() is executed for all devices after invoking
  *	subsystem-level @prepare() for all of them.
  *
+ * @suspend_late: Continue operations started by @suspend().  For a number of
+ *	devices @suspend_late() may point to the same callback routine as the
+ *	runtime suspend callback.
+ *
  * @resume: Executed after waking the system up from a sleep state in which the
  *	contents of main memory were preserved.  The exact action to perform
  *	depends on the device's subsystem, but generally the driver is expected
@@ -122,6 +126,10 @@ typedef struct pm_message {
  *	Subsystem-level @resume() is executed for all devices after invoking
  *	subsystem-level @resume_noirq() for all of them.
  *
+ * @resume_early: Prepare to execute @resume().  For a number of devices
+ *	@resume_early() may point to the same callback routine as the runtime
+ *	resume callback.
+ *
  * @freeze: Hibernation-specific, executed before creating a hibernation image.
  *	Analogous to @suspend(), but it should not enable the device to signal
  *	wakeup events or change its power state.  The majority of subsystems
@@ -131,6 +139,10 @@ typedef struct pm_message {
  *	Subsystem-level @freeze() is executed for all devices after invoking
  *	subsystem-level @prepare() for all of them.
  *
+ * @freeze_late: Continue operations started by @freeze().  Analogous to
+ *	@suspend_late(), but it should not enable the device to signal wakeup
+ *	events or change its power state.
+ *
  * @thaw: Hibernation-specific, executed after creating a hibernation image OR
  *	if the creation of an image has failed.  Also executed after a failing
  *	attempt to restore the contents of main memory from such an image.
@@ -140,15 +152,23 @@ typedef struct pm_message {
  *	subsystem-level @thaw_noirq() for all of them.  It also may be executed
  *	directly after @freeze() in case of a transition error.
  *
+ * @thaw_early: Prepare to execute @thaw().  Undo the changes made by the
+ *	preceding @freeze_late().
+ *
  * @poweroff: Hibernation-specific, executed after saving a hibernation image.
  *	Analogous to @suspend(), but it need not save the device's settings in
  *	memory.
  *	Subsystem-level @poweroff() is executed for all devices after invoking
  *	subsystem-level @prepare() for all of them.
  *
+ * @poweroff_late: Continue operations started by @poweroff().  Analogous to
+ *	@suspend_late(), but it need not save the device's settings in memory.
+ *
  * @restore: Hibernation-specific, executed after restoring the contents of main
  *	memory from a hibernation image, analogous to @resume().
  *
+ * @restore_early: Prepare to execute @restore(), analogous to @resume_early().
+ *
  * @suspend_noirq: Complete the actions started by @suspend().  Carry out any
  *	additional operations required for suspending the device that might be
  *	racing with its driver's interrupt handler, which is guaranteed not to
@@ -158,9 +178,10 @@ typedef struct pm_message {
  *	@suspend_noirq() has returned successfully.  If the device can generate
  *	system wakeup signals and is enabled to wake up the system, it should be
  *	configured to do so at that time.  However, depending on the platform
- *	and device's subsystem, @suspend() may be allowed to put the device into
+ *	and device's subsystem, @suspend() or @suspend_late() may be allowed to
- *	the low-power state and configure it to generate wakeup signals, in
+ *	put the device into the low-power state and configure it to generate
- *	which case it generally is not necessary to define @suspend_noirq().
+ *	wakeup signals, in which case it generally is not necessary to define
+ *	@suspend_noirq().
  *
  * @resume_noirq: Prepare for the execution of @resume() by carrying out any
  *	operations required for resuming the device that might be racing with
@@ -171,9 +192,9 @@ typedef struct pm_message {
  *	additional operations required for freezing the device that might be
  *	racing with its driver's interrupt handler, which is guaranteed not to
  *	run while @freeze_noirq() is being executed.
- *	The power state of the device should not be changed by either @freeze()
+ *	The power state of the device should not be changed by either @freeze(),
- *	or @freeze_noirq() and it should not be configured to signal system
+ *	or @freeze_late(), or @freeze_noirq() and it should not be configured to
- *	wakeup by any of these callbacks.
+ *	signal system wakeup by any of these callbacks.
  *
  * @thaw_noirq: Prepare for the execution of @thaw() by carrying out any
  *	operations required for thawing the device that might be racing with its
@@ -249,6 +270,12 @@ struct dev_pm_ops {
 	int (*thaw)(struct device *dev);
 	int (*poweroff)(struct device *dev);
 	int (*restore)(struct device *dev);
+	int (*suspend_late)(struct device *dev);
+	int (*resume_early)(struct device *dev);
+	int (*freeze_late)(struct device *dev);
+	int (*thaw_early)(struct device *dev);
+	int (*poweroff_late)(struct device *dev);
+	int (*restore_early)(struct device *dev);
 	int (*suspend_noirq)(struct device *dev);
 	int (*resume_noirq)(struct device *dev);
 	int (*freeze_noirq)(struct device *dev);
@@ -293,6 +320,15 @@ const struct dev_pm_ops name = { \
 /*
  * Use this for defining a set of PM operations to be used in all situations
  * (sustem suspend, hibernation or runtime PM).
+ * NOTE: In general, system suspend callbacks, .suspend() and .resume(), should
+ * be different from the corresponding runtime PM callbacks, .runtime_suspend(),
+ * and .runtime_resume(), because .runtime_suspend() always works on an already
+ * quiescent device, while .suspend() should assume that the device may be doing
+ * something when it is called (it should ensure that the device will be
+ * quiescent after it has returned).  Therefore it's better to point the "late"
+ * suspend and "early" resume callback pointers, .suspend_late() and
+ * .resume_early(), to the same routines as .runtime_suspend() and
+ * .runtime_resume(), respectively (and analogously for hibernation).
  */
 #define UNIVERSAL_DEV_PM_OPS(name, suspend_fn, resume_fn, idle_fn) \
 const struct dev_pm_ops name = { \
@@ -584,13 +620,13 @@ struct dev_pm_domain {
 
 #ifdef CONFIG_PM_SLEEP
 extern void device_pm_lock(void);
-extern void dpm_resume_noirq(pm_message_t state);
+extern void dpm_resume_start(pm_message_t state);
 extern void dpm_resume_end(pm_message_t state);
 extern void dpm_resume(pm_message_t state);
 extern void dpm_complete(pm_message_t state);
 
 extern void device_pm_unlock(void);
-extern int dpm_suspend_noirq(pm_message_t state);
+extern int dpm_suspend_end(pm_message_t state);
 extern int dpm_suspend_start(pm_message_t state);
 extern int dpm_suspend(pm_message_t state);
 extern int dpm_prepare(pm_message_t state);
@@ -605,17 +641,23 @@ extern void __suspend_report_result(const char *function, void *fn, int ret);
 extern int device_pm_wait_for_dev(struct device *sub, struct device *dev);
 
 extern int pm_generic_prepare(struct device *dev);
+extern int pm_generic_suspend_late(struct device *dev);
 extern int pm_generic_suspend_noirq(struct device *dev);
 extern int pm_generic_suspend(struct device *dev);
+extern int pm_generic_resume_early(struct device *dev);
 extern int pm_generic_resume_noirq(struct device *dev);
 extern int pm_generic_resume(struct device *dev);
 extern int pm_generic_freeze_noirq(struct device *dev);
+extern int pm_generic_freeze_late(struct device *dev);
 extern int pm_generic_freeze(struct device *dev);
 extern int pm_generic_thaw_noirq(struct device *dev);
+extern int pm_generic_thaw_early(struct device *dev);
 extern int pm_generic_thaw(struct device *dev);
 extern int pm_generic_restore_noirq(struct device *dev);
+extern int pm_generic_restore_early(struct device *dev);
 extern int pm_generic_restore(struct device *dev);
 extern int pm_generic_poweroff_noirq(struct device *dev);
+extern int pm_generic_poweroff_late(struct device *dev);
 extern int pm_generic_poweroff(struct device *dev);
 extern void pm_generic_complete(struct device *dev);
 

include/linux/pm_wakeup.h

@@ -41,7 +41,7 @@
  * @active: Status of the wakeup source.
  */
 struct wakeup_source {
-	char 			*name;
+	const char 		*name;
 	struct list_head	entry;
 	spinlock_t		lock;
 	struct timer_list	timer;
@@ -73,7 +73,9 @@ static inline bool device_may_wakeup(struct device *dev)
 }
 
 /* drivers/base/power/wakeup.c */
+extern void wakeup_source_prepare(struct wakeup_source *ws, const char *name);
 extern struct wakeup_source *wakeup_source_create(const char *name);
+extern void wakeup_source_drop(struct wakeup_source *ws);
 extern void wakeup_source_destroy(struct wakeup_source *ws);
 extern void wakeup_source_add(struct wakeup_source *ws);
 extern void wakeup_source_remove(struct wakeup_source *ws);
@@ -103,11 +105,16 @@ static inline bool device_can_wakeup(struct device *dev)
 	return dev->power.can_wakeup;
 }
 
+static inline void wakeup_source_prepare(struct wakeup_source *ws,
+					 const char *name) {}
+
 static inline struct wakeup_source *wakeup_source_create(const char *name)
 {
 	return NULL;
 }
 
+static inline void wakeup_source_drop(struct wakeup_source *ws) {}
+
 static inline void wakeup_source_destroy(struct wakeup_source *ws) {}
 
 static inline void wakeup_source_add(struct wakeup_source *ws) {}
@@ -165,4 +172,17 @@ static inline void pm_wakeup_event(struct device *dev, unsigned int msec) {}
 
 #endif /* !CONFIG_PM_SLEEP */
 
+static inline void wakeup_source_init(struct wakeup_source *ws,
+				      const char *name)
+{
+	wakeup_source_prepare(ws, name);
+	wakeup_source_add(ws);
+}
+
+static inline void wakeup_source_trash(struct wakeup_source *ws)
+{
+	wakeup_source_remove(ws);
+	wakeup_source_drop(ws);
+}
+
 #endif /* _LINUX_PM_WAKEUP_H */

include/linux/suspend.h

@@ -42,8 +42,10 @@ enum suspend_stat_step {
 	SUSPEND_FREEZE = 1,
 	SUSPEND_PREPARE,
 	SUSPEND_SUSPEND,
+	SUSPEND_SUSPEND_LATE,
 	SUSPEND_SUSPEND_NOIRQ,
 	SUSPEND_RESUME_NOIRQ,
+	SUSPEND_RESUME_EARLY,
 	SUSPEND_RESUME
 };
 
@@ -53,8 +55,10 @@ struct suspend_stats {
 	int	failed_freeze;
 	int	failed_prepare;
 	int	failed_suspend;
+	int	failed_suspend_late;
 	int	failed_suspend_noirq;
 	int	failed_resume;
+	int	failed_resume_early;
 	int	failed_resume_noirq;
 #define	REC_FAILED_NUM	2
 	int	last_failed_dev;

kernel/exit.c

@@ -424,7 +424,7 @@ void daemonize(const char *name, ...)
 	 */
 	exit_mm(current);
 	/*
-	 * We don't want to have TIF_FREEZE set if the system-wide hibernation
+	 * We don't want to get frozen, in case system-wide hibernation
 	 * or suspend transition begins right now.
 	 */
 	current->flags |= (PF_NOFREEZE | PF_KTHREAD);

kernel/freezer.c

@@ -99,9 +99,9 @@ static void fake_signal_wake_up(struct task_struct *p)
  * freeze_task - send a freeze request to given task
  * @p: task to send the request to
  *
- * If @p is freezing, the freeze request is sent by setting %TIF_FREEZE
+ * If @p is freezing, the freeze request is sent either by sending a fake
- * flag and either sending a fake signal to it or waking it up, depending
+ * signal (if it's not a kernel thread) or waking it up (if it's a kernel
- * on whether it has %PF_FREEZER_NOSIG set.
+ * thread).
  *
  * RETURNS:
  * %false, if @p is not freezing or already frozen; %true, otherwise

kernel/kexec.c

@@ -1546,13 +1546,13 @@ int kernel_kexec(void)
 		if (error)
 			goto Resume_console;
 		/* At this point, dpm_suspend_start() has been called,
-		 * but *not* dpm_suspend_noirq(). We *must* call
+		 * but *not* dpm_suspend_end(). We *must* call
-		 * dpm_suspend_noirq() now.  Otherwise, drivers for
+		 * dpm_suspend_end() now.  Otherwise, drivers for
 		 * some devices (e.g. interrupt controllers) become
 		 * desynchronized with the actual state of the
 		 * hardware at resume time, and evil weirdness ensues.
 		 */
-		error = dpm_suspend_noirq(PMSG_FREEZE);
+		error = dpm_suspend_end(PMSG_FREEZE);
 		if (error)
 			goto Resume_devices;
 		error = disable_nonboot_cpus();
@@ -1579,7 +1579,7 @@ int kernel_kexec(void)
 		local_irq_enable();
  Enable_cpus:
 		enable_nonboot_cpus();
-		dpm_resume_noirq(PMSG_RESTORE);
+		dpm_resume_start(PMSG_RESTORE);
  Resume_devices:
 		dpm_resume_end(PMSG_RESTORE);
  Resume_console:

kernel/power/hibernate.c

@@ -245,8 +245,8 @@ void swsusp_show_speed(struct timeval *start, struct timeval *stop,
  * create_image - Create a hibernation image.
  * @platform_mode: Whether or not to use the platform driver.
  *
- * Execute device drivers' .freeze_noirq() callbacks, create a hibernation image
+ * Execute device drivers' "late" and "noirq" freeze callbacks, create a
- * and execute the drivers' .thaw_noirq() callbacks.
+ * hibernation image and run the drivers' "noirq" and "early" thaw callbacks.
  *
  * Control reappears in this routine after the subsequent restore.
  */
@@ -254,7 +254,7 @@ static int create_image(int platform_mode)
 {
 	int error;
 
-	error = dpm_suspend_noirq(PMSG_FREEZE);
+	error = dpm_suspend_end(PMSG_FREEZE);
 	if (error) {
 		printk(KERN_ERR "PM: Some devices failed to power down, "
 			"aborting hibernation\n");
@@ -306,7 +306,7 @@ static int create_image(int platform_mode)
  Platform_finish:
 	platform_finish(platform_mode);
 
-	dpm_resume_noirq(in_suspend ?
+	dpm_resume_start(in_suspend ?
 		(error ? PMSG_RECOVER : PMSG_THAW) : PMSG_RESTORE);
 
 	return error;
@@ -343,13 +343,13 @@ int hibernation_snapshot(int platform_mode)
 		 * successful freezer test.
 		 */
 		freezer_test_done = true;
-		goto Cleanup;
+		goto Thaw;
 	}
 
 	error = dpm_prepare(PMSG_FREEZE);
 	if (error) {
 		dpm_complete(PMSG_RECOVER);
-		goto Cleanup;
+		goto Thaw;
 	}
 
 	suspend_console();
@@ -385,6 +385,8 @@ int hibernation_snapshot(int platform_mode)
 	platform_end(platform_mode);
 	return error;
 
+ Thaw:
+	thaw_kernel_threads();
  Cleanup:
 	swsusp_free();
 	goto Close;
@@ -394,16 +396,16 @@ int hibernation_snapshot(int platform_mode)
  * resume_target_kernel - Restore system state from a hibernation image.
  * @platform_mode: Whether or not to use the platform driver.
  *
- * Execute device drivers' .freeze_noirq() callbacks, restore the contents of
+ * Execute device drivers' "noirq" and "late" freeze callbacks, restore the
- * highmem that have not been restored yet from the image and run the low-level
+ * contents of highmem that have not been restored yet from the image and run
- * code that will restore the remaining contents of memory and switch to the
+ * the low-level code that will restore the remaining contents of memory and
- * just restored target kernel.
+ * switch to the just restored target kernel.
  */
 static int resume_target_kernel(bool platform_mode)
 {
 	int error;
 
-	error = dpm_suspend_noirq(PMSG_QUIESCE);
+	error = dpm_suspend_end(PMSG_QUIESCE);
 	if (error) {
 		printk(KERN_ERR "PM: Some devices failed to power down, "
 			"aborting resume\n");
@@ -460,7 +462,7 @@ static int resume_target_kernel(bool platform_mode)
  Cleanup:
 	platform_restore_cleanup(platform_mode);
 
-	dpm_resume_noirq(PMSG_RECOVER);
+	dpm_resume_start(PMSG_RECOVER);
 
 	return error;
 }
@@ -518,7 +520,7 @@ int hibernation_platform_enter(void)
 		goto Resume_devices;
 	}
 
-	error = dpm_suspend_noirq(PMSG_HIBERNATE);
+	error = dpm_suspend_end(PMSG_HIBERNATE);
 	if (error)
 		goto Resume_devices;
 
@@ -549,7 +551,7 @@ int hibernation_platform_enter(void)
  Platform_finish:
 	hibernation_ops->finish();
 
-	dpm_resume_noirq(PMSG_RESTORE);
+	dpm_resume_start(PMSG_RESTORE);
 
  Resume_devices:
 	entering_platform_hibernation = false;
@@ -616,7 +618,7 @@ int hibernate(void)
 	/* Allocate memory management structures */
 	error = create_basic_memory_bitmaps();
 	if (error)
-		goto Exit;
+		goto Enable_umh;
 
 	printk(KERN_INFO "PM: Syncing filesystems ... ");
 	sys_sync();
@@ -624,15 +626,11 @@ int hibernate(void)
 
 	error = freeze_processes();
 	if (error)
-		goto Finish;
+		goto Free_bitmaps;
 
 	error = hibernation_snapshot(hibernation_mode == HIBERNATION_PLATFORM);
-	if (error)
+	if (error || freezer_test_done)
 		goto Thaw;
-	if (freezer_test_done) {
-		freezer_test_done = false;
-		goto Thaw;
-	}
 
 	if (in_suspend) {
 		unsigned int flags = 0;
@@ -657,8 +655,13 @@ int hibernate(void)
 
  Thaw:
 	thaw_processes();
- Finish:
+
+	/* Don't bother checking whether freezer_test_done is true */
+	freezer_test_done = false;
+
+ Free_bitmaps:
 	free_basic_memory_bitmaps();
+ Enable_umh:
 	usermodehelper_enable();
  Exit:
 	pm_notifier_call_chain(PM_POST_HIBERNATION);

kernel/power/main.c

@@ -165,16 +165,20 @@ static int suspend_stats_show(struct seq_file *s, void *unused)
 	last_errno %= REC_FAILED_NUM;
 	last_step = suspend_stats.last_failed_step + REC_FAILED_NUM - 1;
 	last_step %= REC_FAILED_NUM;
-	seq_printf(s, "%s: %d\n%s: %d\n%s: %d\n%s: %d\n"
+	seq_printf(s, "%s: %d\n%s: %d\n%s: %d\n%s: %d\n%s: %d\n"
-			"%s: %d\n%s: %d\n%s: %d\n%s: %d\n",
+			"%s: %d\n%s: %d\n%s: %d\n%s: %d\n%s: %d\n",
 			"success", suspend_stats.success,
 			"fail", suspend_stats.fail,
 			"failed_freeze", suspend_stats.failed_freeze,
 			"failed_prepare", suspend_stats.failed_prepare,
 			"failed_suspend", suspend_stats.failed_suspend,
+			"failed_suspend_late",
+				suspend_stats.failed_suspend_late,
 			"failed_suspend_noirq",
 				suspend_stats.failed_suspend_noirq,
 			"failed_resume", suspend_stats.failed_resume,
+			"failed_resume_early",
+				suspend_stats.failed_resume_early,
 			"failed_resume_noirq",
 				suspend_stats.failed_resume_noirq);
 	seq_printf(s,	"failures:\n  last_failed_dev:\t%-s\n",
@@ -287,16 +291,10 @@ static ssize_t state_store(struct kobject *kobj, struct kobj_attribute *attr,
 
 #ifdef CONFIG_SUSPEND
 	for (s = &pm_states[state]; state < PM_SUSPEND_MAX; s++, state++) {
-		if (*s && len == strlen(*s) && !strncmp(buf, *s, len))
+		if (*s && len == strlen(*s) && !strncmp(buf, *s, len)) {
+			error = pm_suspend(state);
 			break;
-	}
+		}
-	if (state < PM_SUSPEND_MAX && *s) {
-		error = enter_state(state);
-		if (error) {
-			suspend_stats.fail++;
-			dpm_save_failed_errno(error);
-		} else
-			suspend_stats.success++;
 	}
 #endif
 

kernel/power/power.h

@@ -177,13 +177,11 @@ extern const char *const pm_states[];
 
 extern bool valid_state(suspend_state_t state);
 extern int suspend_devices_and_enter(suspend_state_t state);
-extern int enter_state(suspend_state_t state);
 #else /* !CONFIG_SUSPEND */
 static inline int suspend_devices_and_enter(suspend_state_t state)
 {
 	return -ENOSYS;
 }
-static inline int enter_state(suspend_state_t state) { return -ENOSYS; }
 static inline bool valid_state(suspend_state_t state) { return false; }
 #endif /* !CONFIG_SUSPEND */
 
@@ -234,16 +232,14 @@ static inline int suspend_freeze_processes(void)
 	int error;
 
 	error = freeze_processes();
-
 	/*
 	 * freeze_processes() automatically thaws every task if freezing
 	 * fails. So we need not do anything extra upon error.
 	 */
 	if (error)
-		goto Finish;
+		return error;
 
 	error = freeze_kernel_threads();
-
 	/*
 	 * freeze_kernel_threads() thaws only kernel threads upon freezing
 	 * failure. So we have to thaw the userspace tasks ourselves.
@@ -251,7 +247,6 @@ static inline int suspend_freeze_processes(void)
 	if (error)
 		thaw_processes();
 
- Finish:
 	return error;
 }
 

kernel/power/process.c

@@ -53,11 +53,9 @@ static int try_to_freeze_tasks(bool user_only)
 			 * It is "frozen enough".  If the task does wake
 			 * up, it will immediately call try_to_freeze.
 			 *
-			 * Because freeze_task() goes through p's
+			 * Because freeze_task() goes through p's scheduler lock, it's
-			 * scheduler lock after setting TIF_FREEZE, it's
+			 * guaranteed that TASK_STOPPED/TRACED -> TASK_RUNNING
-			 * guaranteed that either we see TASK_RUNNING or
+			 * transition can't race with task state testing here.
-			 * try_to_stop() after schedule() in ptrace/signal
-			 * stop sees TIF_FREEZE.
 			 */
 			if (!task_is_stopped_or_traced(p) &&
 			    !freezer_should_skip(p))
@@ -98,13 +96,15 @@ static int try_to_freeze_tasks(bool user_only)
 		       elapsed_csecs / 100, elapsed_csecs % 100,
 		       todo - wq_busy, wq_busy);
 
-		read_lock(&tasklist_lock);
+		if (!wakeup) {
-		do_each_thread(g, p) {
+			read_lock(&tasklist_lock);
-			if (!wakeup && !freezer_should_skip(p) &&
+			do_each_thread(g, p) {
-			    p != current && freezing(p) && !frozen(p))
+				if (p != current && !freezer_should_skip(p)
-				sched_show_task(p);
+				    && freezing(p) && !frozen(p))
-		} while_each_thread(g, p);
+					sched_show_task(p);
-		read_unlock(&tasklist_lock);
+			} while_each_thread(g, p);
+			read_unlock(&tasklist_lock);
+		}
 	} else {
 		printk("(elapsed %d.%02d seconds) ", elapsed_csecs / 100,
 			elapsed_csecs % 100);

kernel/power/snapshot.c

@@ -711,9 +711,10 @@ static void mark_nosave_pages(struct memory_bitmap *bm)
 	list_for_each_entry(region, &nosave_regions, list) {
 		unsigned long pfn;
 
-		pr_debug("PM: Marking nosave pages: %016lx - %016lx\n",
+		pr_debug("PM: Marking nosave pages: [mem %#010llx-%#010llx]\n",
-				region->start_pfn << PAGE_SHIFT,
+			 (unsigned long long) region->start_pfn << PAGE_SHIFT,
-				region->end_pfn << PAGE_SHIFT);
+			 ((unsigned long long) region->end_pfn << PAGE_SHIFT)
+				- 1);
 
 		for (pfn = region->start_pfn; pfn < region->end_pfn; pfn++)
 			if (pfn_valid(pfn)) {

kernel/power/suspend.c

@@ -37,8 +37,8 @@ const char *const pm_states[PM_SUSPEND_MAX] = {
 static const struct platform_suspend_ops *suspend_ops;
 
 /**
- *	suspend_set_ops - Set the global suspend method table.
+ * suspend_set_ops - Set the global suspend method table.
- *	@ops:	Pointer to ops structure.
+ * @ops: Suspend operations to use.
  */
 void suspend_set_ops(const struct platform_suspend_ops *ops)
 {
@@ -58,11 +58,11 @@ bool valid_state(suspend_state_t state)
 }
 
 /**
- * suspend_valid_only_mem - generic memory-only valid callback
+ * suspend_valid_only_mem - Generic memory-only valid callback.
  *
- * Platform drivers that implement mem suspend only and only need
+ * Platform drivers that implement mem suspend only and only need to check for
- * to check for that in their .valid callback can use this instead
+ * that in their .valid() callback can use this instead of rolling their own
- * of rolling their own .valid callback.
+ * .valid() callback.
  */
 int suspend_valid_only_mem(suspend_state_t state)
 {
@@ -83,10 +83,11 @@ static int suspend_test(int level)
 }
 
 /**
- *	suspend_prepare - Do prep work before entering low-power state.
+ * suspend_prepare - Prepare for entering system sleep state.
  *
- *	This is common code that is called for each state that we're entering.
+ * Common code run for every system sleep state that can be entered (except for
- *	Run suspend notifiers, allocate a console and stop all processes.
+ * hibernation).  Run suspend notifiers, allocate the "suspend" console and
+ * freeze processes.
  */
 static int suspend_prepare(void)
 {
@@ -131,9 +132,9 @@ void __attribute__ ((weak)) arch_suspend_enable_irqs(void)
 }
 
 /**
- * suspend_enter - enter the desired system sleep state.
+ * suspend_enter - Make the system enter the given sleep state.
- * @state: State to enter
+ * @state: System sleep state to enter.
- * @wakeup: Returns information that suspend should not be entered again.
+ * @wakeup: Returns information that the sleep state should not be re-entered.
  *
  * This function should be called after devices have been suspended.
  */
@@ -147,7 +148,7 @@ static int suspend_enter(suspend_state_t state, bool *wakeup)
 			goto Platform_finish;
 	}
 
-	error = dpm_suspend_noirq(PMSG_SUSPEND);
+	error = dpm_suspend_end(PMSG_SUSPEND);
 	if (error) {
 		printk(KERN_ERR "PM: Some devices failed to power down\n");
 		goto Platform_finish;
@@ -189,7 +190,7 @@ static int suspend_enter(suspend_state_t state, bool *wakeup)
 	if (suspend_ops->wake)
 		suspend_ops->wake();
 
-	dpm_resume_noirq(PMSG_RESUME);
+	dpm_resume_start(PMSG_RESUME);
 
  Platform_finish:
 	if (suspend_ops->finish)
@@ -199,9 +200,8 @@ static int suspend_enter(suspend_state_t state, bool *wakeup)
 }
 
 /**
- *	suspend_devices_and_enter - suspend devices and enter the desired system
+ * suspend_devices_and_enter - Suspend devices and enter system sleep state.
- *				    sleep state.
+ * @state: System sleep state to enter.
- *	@state:		  state to enter
  */
 int suspend_devices_and_enter(suspend_state_t state)
 {
@@ -251,10 +251,10 @@ int suspend_devices_and_enter(suspend_state_t state)
 }
 
 /**
- *	suspend_finish - Do final work before exiting suspend sequence.
+ * suspend_finish - Clean up before finishing the suspend sequence.
  *
- *	Call platform code to clean up, restart processes, and free the
+ * Call platform code to clean up, restart processes, and free the console that
- *	console that we've allocated. This is not called for suspend-to-disk.
+ * we've allocated. This routine is not called for hibernation.
  */
 static void suspend_finish(void)
 {
@@ -265,16 +265,14 @@ static void suspend_finish(void)
 }
 
 /**
- *	enter_state - Do common work of entering low-power state.
+ * enter_state - Do common work needed to enter system sleep state.
- *	@state:		pm_state structure for state we're entering.
+ * @state: System sleep state to enter.
  *
- *	Make sure we're the only ones trying to enter a sleep state. Fail
+ * Make sure that no one else is trying to put the system into a sleep state.
- *	if someone has beat us to it, since we don't want anything weird to
+ * Fail if that's not the case.  Otherwise, prepare for system suspend, make the
- *	happen when we wake up.
+ * system enter the given sleep state and clean up after wakeup.
- *	Then, do the setup for suspend, enter the state, and cleaup (after
- *	we've woken up).
  */
-int enter_state(suspend_state_t state)
+static int enter_state(suspend_state_t state)
 {
 	int error;
 
@@ -310,24 +308,26 @@ int enter_state(suspend_state_t state)
 }
 
 /**
- *	pm_suspend - Externally visible function for suspending system.
+ * pm_suspend - Externally visible function for suspending the system.
- *	@state:		Enumerated value of state to enter.
+ * @state: System sleep state to enter.
  *
- *	Determine whether or not value is within range, get state
+ * Check if the value of @state represents one of the supported states,
- *	structure, and enter (above).
+ * execute enter_state() and update system suspend statistics.
  */
 int pm_suspend(suspend_state_t state)
 {
-	int ret;
+	int error;
-	if (state > PM_SUSPEND_ON && state < PM_SUSPEND_MAX) {
+
-		ret = enter_state(state);
+	if (state <= PM_SUSPEND_ON || state >= PM_SUSPEND_MAX)
-		if (ret) {
+		return -EINVAL;
-			suspend_stats.fail++;
+
-			dpm_save_failed_errno(ret);
+	error = enter_state(state);
-		} else
+	if (error) {
-			suspend_stats.success++;
+		suspend_stats.fail++;
-		return ret;
+		dpm_save_failed_errno(error);
+	} else {
+		suspend_stats.success++;
 	}
-	return -EINVAL;
+	return error;
 }
 EXPORT_SYMBOL(pm_suspend);

kernel/power/user.c

@@ -249,16 +249,10 @@ static long snapshot_ioctl(struct file *filp, unsigned int cmd,
 		}
 		pm_restore_gfp_mask();
 		error = hibernation_snapshot(data->platform_support);
-		if (error) {
+		if (!error) {
-			thaw_kernel_threads();
-		} else {
 			error = put_user(in_suspend, (int __user *)arg);
-			if (!error && !freezer_test_done)
+			data->ready = !freezer_test_done && !error;
-				data->ready = 1;
+			freezer_test_done = false;
-			if (freezer_test_done) {
-				freezer_test_done = false;
-				thaw_kernel_threads();
-			}
 		}
 		break;