[PATCH 5.15 0/1] pm, restore async device resume optimization

[PATCH 5.15 0/1] pm, restore async device resume optimization

[Yenchia Chen]

From: "yenchia.chen" <yenchia.chen@mediatek.com>

We have met a deadlock issue on our device when resuming.
After applying this patch which is picked from mainline, issue solved.
We'd like to backport to 5.15.y and could you help to review? thanks.

[ Upstream commit 3e999770ac1c7c31a70685dd5b88e89473509e9c ]

Rafael J. Wysocki (1):
  PM: sleep: Restore asynchronous device resume optimization

 drivers/base/power/main.c | 117 +++++++++++++++++++++-----------------
 include/linux/pm.h        |   1 +
 2 files changed, 65 insertions(+), 53 deletions(-)

-- 
2.18.0

[PATCH 5.15 1/1] PM: sleep: Restore asynchronous device resume optimization

[Yenchia Chen]

From: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>

Before commit 7839d0078e0d ("PM: sleep: Fix possible deadlocks in core
system-wide PM code"), the resume of devices that were allowed to resume
asynchronously was scheduled before starting the resume of the other
devices, so the former did not have to wait for the latter unless
functional dependencies were present.

Commit 7839d0078e0d removed that optimization in order to address a
correctness issue, but it can be restored with the help of a new device
power management flag, so do that now.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Signed-off-by: yenchia.chen <yenchia.chen@mediatek.com>
---
 drivers/base/power/main.c | 117 +++++++++++++++++++++-----------------
 include/linux/pm.h        |   1 +
 2 files changed, 65 insertions(+), 53 deletions(-)

diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 185ea0d93a5e..01591c105511 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -580,7 +580,7 @@ bool dev_pm_skip_resume(struct device *dev)
 }
 
 /**
- * __device_resume_noirq - Execute a "noirq resume" callback for given device.
+ * device_resume_noirq - Execute a "noirq resume" callback for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
@@ -588,7 +588,7 @@ bool dev_pm_skip_resume(struct device *dev)
  * The driver of @dev will not receive interrupts while this function is being
  * executed.
  */
-static void __device_resume_noirq(struct device *dev, pm_message_t state, bool async)
+static void device_resume_noirq(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -675,16 +675,22 @@ static bool dpm_async_fn(struct device *dev, async_func_t func)
 {
 	reinit_completion(&dev->power.completion);
 
-	if (!is_async(dev))
-		return false;
-
-	get_device(dev);
+	if (is_async(dev)) {
+		dev->power.async_in_progress = true;
 
-	if (async_schedule_dev_nocall(func, dev))
-		return true;
+		get_device(dev);
 
-	put_device(dev);
+		if (async_schedule_dev_nocall(func, dev))
+			return true;
 
+		put_device(dev);
+	}
+	/*
+	 * Because async_schedule_dev_nocall() above has returned false or it
+	 * has not been called at all, func() is not running and it is safe to
+	 * update the async_in_progress flag without extra synchronization.
+	 */
+	dev->power.async_in_progress = false;
 	return false;
 }
 
@@ -692,18 +698,10 @@ static void async_resume_noirq(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume_noirq(dev, pm_transition, true);
+	device_resume_noirq(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume_noirq(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume_noirq))
-		return;
-
-	__device_resume_noirq(dev, pm_transition, false);
-}
-
 static void dpm_noirq_resume_devices(pm_message_t state)
 {
 	struct device *dev;
@@ -713,18 +711,28 @@ static void dpm_noirq_resume_devices(pm_message_t state)
 	mutex_lock(&dpm_list_mtx);
 	pm_transition = state;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_noirq_list, power.entry)
+		dpm_async_fn(dev, async_resume_noirq);
+
 	while (!list_empty(&dpm_noirq_list)) {
 		dev = to_device(dpm_noirq_list.next);
-		get_device(dev);
 		list_move_tail(&dev->power.entry, &dpm_late_early_list);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			get_device(dev);
 
-		device_resume_noirq(dev);
+			mutex_unlock(&dpm_list_mtx);
 
-		put_device(dev);
+			device_resume_noirq(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			put_device(dev);
+
+			mutex_lock(&dpm_list_mtx);
+		}
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
@@ -750,14 +758,14 @@ void dpm_resume_noirq(pm_message_t state)
 }
 
 /**
- * __device_resume_early - Execute an "early resume" callback for given device.
+ * device_resume_early - Execute an "early resume" callback for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
  *
  * Runtime PM is disabled for @dev while this function is being executed.
  */
-static void __device_resume_early(struct device *dev, pm_message_t state, bool async)
+static void device_resume_early(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -823,18 +831,10 @@ static void async_resume_early(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume_early(dev, pm_transition, true);
+	device_resume_early(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume_early(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume_early))
-		return;
-
-	__device_resume_early(dev, pm_transition, false);
-}
-
 /**
  * dpm_resume_early - Execute "early resume" callbacks for all devices.
  * @state: PM transition of the system being carried out.
@@ -848,18 +848,28 @@ void dpm_resume_early(pm_message_t state)
 	mutex_lock(&dpm_list_mtx);
 	pm_transition = state;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_late_early_list, power.entry)
+		dpm_async_fn(dev, async_resume_early);
+
 	while (!list_empty(&dpm_late_early_list)) {
 		dev = to_device(dpm_late_early_list.next);
-		get_device(dev);
 		list_move_tail(&dev->power.entry, &dpm_suspended_list);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			get_device(dev);
 
-		device_resume_early(dev);
+			mutex_unlock(&dpm_list_mtx);
 
-		put_device(dev);
+			device_resume_early(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			put_device(dev);
+
+			mutex_lock(&dpm_list_mtx);
+		}
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
@@ -879,12 +889,12 @@ void dpm_resume_start(pm_message_t state)
 EXPORT_SYMBOL_GPL(dpm_resume_start);
 
 /**
- * __device_resume - Execute "resume" callbacks for given device.
+ * device_resume - Execute "resume" callbacks for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
  */
-static void __device_resume(struct device *dev, pm_message_t state, bool async)
+static void device_resume(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -978,18 +988,10 @@ static void async_resume(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume(dev, pm_transition, true);
+	device_resume(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume))
-		return;
-
-	__device_resume(dev, pm_transition, false);
-}
-
 /**
  * dpm_resume - Execute "resume" callbacks for non-sysdev devices.
  * @state: PM transition of the system being carried out.
@@ -1009,16 +1011,25 @@ void dpm_resume(pm_message_t state)
 	pm_transition = state;
 	async_error = 0;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_suspended_list, power.entry)
+		dpm_async_fn(dev, async_resume);
+
 	while (!list_empty(&dpm_suspended_list)) {
 		dev = to_device(dpm_suspended_list.next);
 
 		get_device(dev);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			mutex_unlock(&dpm_list_mtx);
 
-		device_resume(dev);
+			device_resume(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			mutex_lock(&dpm_list_mtx);
+		}
 
 		if (!list_empty(&dev->power.entry))
 			list_move_tail(&dev->power.entry, &dpm_prepared_list);
diff --git a/include/linux/pm.h b/include/linux/pm.h
index d1c19f5b1380..8f46e9f01929 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -606,6 +606,7 @@ struct dev_pm_info {
 	bool			wakeup_path:1;
 	bool			syscore:1;
 	bool			no_pm_callbacks:1;	/* Owned by the PM core */
+	bool			async_in_progress:1;	/* Owned by the PM core */
 	unsigned int		must_resume:1;	/* Owned by the PM core */
 	unsigned int		may_skip_resume:1;	/* Set by subsystems */
 #else
-- 
2.18.0

Re: [PATCH 5.15 1/1] PM: sleep: Restore asynchronous device resume optimization

[Greg Kroah-Hartman]

On Mon, Sep 02, 2024 at 11:10:45AM +0800, Yenchia Chen wrote:
> From: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
> 
> Before commit 7839d0078e0d ("PM: sleep: Fix possible deadlocks in core
> system-wide PM code"), the resume of devices that were allowed to resume
> asynchronously was scheduled before starting the resume of the other
> devices, so the former did not have to wait for the latter unless
> functional dependencies were present.
> 
> Commit 7839d0078e0d removed that optimization in order to address a
> correctness issue, but it can be restored with the help of a new device
> power management flag, so do that now.
> 
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
> Signed-off-by: yenchia.chen <yenchia.chen@mediatek.com>

Please sign off using your name, not your email alias.

Also, what git id is this?

thanks,

greg k-h

[PATCH 5.15 v2] PM: sleep: Restore asynchronous device resume optimization

[Yenchia Chen]

From: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>

commit 3e999770ac1c7c31a70685dd5b88e89473509e9c upstream.

Before commit 7839d0078e0d ("PM: sleep: Fix possible deadlocks in core
system-wide PM code"), the resume of devices that were allowed to resume
asynchronously was scheduled before starting the resume of the other
devices, so the former did not have to wait for the latter unless
functional dependencies were present.

Commit 7839d0078e0d removed that optimization in order to address a
correctness issue, but it can be restored with the help of a new device
power management flag, so do that now.

Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Signed-off-by: Yenchia Chen <yenchia.chen@mediatek.com>
---
 drivers/base/power/main.c | 117 +++++++++++++++++++++-----------------
 include/linux/pm.h        |   1 +
 2 files changed, 65 insertions(+), 53 deletions(-)

diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 185ea0d93a5e..01591c105511 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -580,7 +580,7 @@ bool dev_pm_skip_resume(struct device *dev)
 }
 
 /**
- * __device_resume_noirq - Execute a "noirq resume" callback for given device.
+ * device_resume_noirq - Execute a "noirq resume" callback for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
@@ -588,7 +588,7 @@ bool dev_pm_skip_resume(struct device *dev)
  * The driver of @dev will not receive interrupts while this function is being
  * executed.
  */
-static void __device_resume_noirq(struct device *dev, pm_message_t state, bool async)
+static void device_resume_noirq(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -675,16 +675,22 @@ static bool dpm_async_fn(struct device *dev, async_func_t func)
 {
 	reinit_completion(&dev->power.completion);
 
-	if (!is_async(dev))
-		return false;
-
-	get_device(dev);
+	if (is_async(dev)) {
+		dev->power.async_in_progress = true;
 
-	if (async_schedule_dev_nocall(func, dev))
-		return true;
+		get_device(dev);
 
-	put_device(dev);
+		if (async_schedule_dev_nocall(func, dev))
+			return true;
 
+		put_device(dev);
+	}
+	/*
+	 * Because async_schedule_dev_nocall() above has returned false or it
+	 * has not been called at all, func() is not running and it is safe to
+	 * update the async_in_progress flag without extra synchronization.
+	 */
+	dev->power.async_in_progress = false;
 	return false;
 }
 
@@ -692,18 +698,10 @@ static void async_resume_noirq(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume_noirq(dev, pm_transition, true);
+	device_resume_noirq(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume_noirq(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume_noirq))
-		return;
-
-	__device_resume_noirq(dev, pm_transition, false);
-}
-
 static void dpm_noirq_resume_devices(pm_message_t state)
 {
 	struct device *dev;
@@ -713,18 +711,28 @@ static void dpm_noirq_resume_devices(pm_message_t state)
 	mutex_lock(&dpm_list_mtx);
 	pm_transition = state;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_noirq_list, power.entry)
+		dpm_async_fn(dev, async_resume_noirq);
+
 	while (!list_empty(&dpm_noirq_list)) {
 		dev = to_device(dpm_noirq_list.next);
-		get_device(dev);
 		list_move_tail(&dev->power.entry, &dpm_late_early_list);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			get_device(dev);
 
-		device_resume_noirq(dev);
+			mutex_unlock(&dpm_list_mtx);
 
-		put_device(dev);
+			device_resume_noirq(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			put_device(dev);
+
+			mutex_lock(&dpm_list_mtx);
+		}
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
@@ -750,14 +758,14 @@ void dpm_resume_noirq(pm_message_t state)
 }
 
 /**
- * __device_resume_early - Execute an "early resume" callback for given device.
+ * device_resume_early - Execute an "early resume" callback for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
  *
  * Runtime PM is disabled for @dev while this function is being executed.
  */
-static void __device_resume_early(struct device *dev, pm_message_t state, bool async)
+static void device_resume_early(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -823,18 +831,10 @@ static void async_resume_early(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume_early(dev, pm_transition, true);
+	device_resume_early(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume_early(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume_early))
-		return;
-
-	__device_resume_early(dev, pm_transition, false);
-}
-
 /**
  * dpm_resume_early - Execute "early resume" callbacks for all devices.
  * @state: PM transition of the system being carried out.
@@ -848,18 +848,28 @@ void dpm_resume_early(pm_message_t state)
 	mutex_lock(&dpm_list_mtx);
 	pm_transition = state;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_late_early_list, power.entry)
+		dpm_async_fn(dev, async_resume_early);
+
 	while (!list_empty(&dpm_late_early_list)) {
 		dev = to_device(dpm_late_early_list.next);
-		get_device(dev);
 		list_move_tail(&dev->power.entry, &dpm_suspended_list);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			get_device(dev);
 
-		device_resume_early(dev);
+			mutex_unlock(&dpm_list_mtx);
 
-		put_device(dev);
+			device_resume_early(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			put_device(dev);
+
+			mutex_lock(&dpm_list_mtx);
+		}
 	}
 	mutex_unlock(&dpm_list_mtx);
 	async_synchronize_full();
@@ -879,12 +889,12 @@ void dpm_resume_start(pm_message_t state)
 EXPORT_SYMBOL_GPL(dpm_resume_start);
 
 /**
- * __device_resume - Execute "resume" callbacks for given device.
+ * device_resume - Execute "resume" callbacks for given device.
  * @dev: Device to handle.
  * @state: PM transition of the system being carried out.
  * @async: If true, the device is being resumed asynchronously.
  */
-static void __device_resume(struct device *dev, pm_message_t state, bool async)
+static void device_resume(struct device *dev, pm_message_t state, bool async)
 {
 	pm_callback_t callback = NULL;
 	const char *info = NULL;
@@ -978,18 +988,10 @@ static void async_resume(void *data, async_cookie_t cookie)
 {
 	struct device *dev = data;
 
-	__device_resume(dev, pm_transition, true);
+	device_resume(dev, pm_transition, true);
 	put_device(dev);
 }
 
-static void device_resume(struct device *dev)
-{
-	if (dpm_async_fn(dev, async_resume))
-		return;
-
-	__device_resume(dev, pm_transition, false);
-}
-
 /**
  * dpm_resume - Execute "resume" callbacks for non-sysdev devices.
  * @state: PM transition of the system being carried out.
@@ -1009,16 +1011,25 @@ void dpm_resume(pm_message_t state)
 	pm_transition = state;
 	async_error = 0;
 
+	/*
+	 * Trigger the resume of "async" devices upfront so they don't have to
+	 * wait for the "non-async" ones they don't depend on.
+	 */
+	list_for_each_entry(dev, &dpm_suspended_list, power.entry)
+		dpm_async_fn(dev, async_resume);
+
 	while (!list_empty(&dpm_suspended_list)) {
 		dev = to_device(dpm_suspended_list.next);
 
 		get_device(dev);
 
-		mutex_unlock(&dpm_list_mtx);
+		if (!dev->power.async_in_progress) {
+			mutex_unlock(&dpm_list_mtx);
 
-		device_resume(dev);
+			device_resume(dev, state, false);
 
-		mutex_lock(&dpm_list_mtx);
+			mutex_lock(&dpm_list_mtx);
+		}
 
 		if (!list_empty(&dev->power.entry))
 			list_move_tail(&dev->power.entry, &dpm_prepared_list);
diff --git a/include/linux/pm.h b/include/linux/pm.h
index d1c19f5b1380..8f46e9f01929 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -606,6 +606,7 @@ struct dev_pm_info {
 	bool			wakeup_path:1;
 	bool			syscore:1;
 	bool			no_pm_callbacks:1;	/* Owned by the PM core */
+	bool			async_in_progress:1;	/* Owned by the PM core */
 	unsigned int		must_resume:1;	/* Owned by the PM core */
 	unsigned int		may_skip_resume:1;	/* Set by subsystems */
 #else
-- 
2.18.0

Re: [PATCH 5.15 0/1] pm, restore async device resume optimization

[Greg Kroah-Hartman]

On Mon, Sep 02, 2024 at 11:10:44AM +0800, Yenchia Chen wrote:
> From: "yenchia.chen" <yenchia.chen@mediatek.com>
> 
> We have met a deadlock issue on our device when resuming.
> After applying this patch which is picked from mainline, issue solved.
> We'd like to backport to 5.15.y and could you help to review? thanks.
> 
> [ Upstream commit 3e999770ac1c7c31a70685dd5b88e89473509e9c ]

For obvious reasons (and as per the documentation), we can't take
commits that are not also in all newer kernel releases, otherwise you
would have a regression if you moved to a newer kernel, right?

Please submit a series for all affected releases if you wish for a patch
to be applied to the trees.

thanks,

greg k-h