Skip to contentSkip to conversation

driver core: Fix PM-runtime for links added during consumer probe

PATCH In review Rafael J. Wysocki 4 participants 4 messages Feb 18, 2019 – Feb 19, 2019

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 driver core: Fix PM-runtime for links added during consumer probe

From: Rafael J. Wysocki <[email protected]>

Commit 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance") introduced a regression that causes suppliers to be suspended prematurely for device links added during consumer driver probe if the initial PM-runtime status of the consumer is "suspended" and the consumer is resumed after adding the link and before pm_runtime_put_suppliers() is called. In that case, pm_runtime_put_suppliers() will drop the rpm_active refcount for the link by one and (since rpm_active is equal to two after the preceding consumer resume) the supplier's PM-runtime usage counter will be decremented, which may cause the supplier to suspend even though the consumer's PM-runtime status is "active".

For this reason, partially revert commit 4c06c4e6cf63 as the problem it tried to fix needs to be addressed somewhat differently, and change pm_runtime_get_suppliers() and pm_runtime_put_suppliers() so that the latter only drops rpm_active references acquired by the former. [This requires adding a new field to struct device_link, but I coulnd't find a cleaner way to address the issue that would work in all cases.]

Fixes 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance")

---

On top of driver-core/driver-core-next.

Jon, Ulf, please test!

---
 drivers/base/core.c          |    4 ++--
 drivers/base/power/runtime.c |   29 ++++++-----------------------
 include/linux/device.h       |    1 +
 include/linux/pm_runtime.h   |    4 ----
 4 files changed, 9 insertions(+), 29 deletions(-)
drivers/base/core.c +2 -2
Index: linux-pm/drivers/base/core.c
===================================================================
@@ -277,7 +277,7 @@ struct device_link *device_link_add(stru
 				link->flags |= DL_FLAG_PM_RUNTIME;
 			}
 			if (flags & DL_FLAG_RPM_ACTIVE)
-				pm_runtime_active_link(link, supplier);
+				refcount_inc(&link->rpm_active);
 		}
 
 		if (flags & DL_FLAG_STATELESS) {
@@ -310,7 +310,7 @@ struct device_link *device_link_add(stru
 
 	if (flags & DL_FLAG_PM_RUNTIME) {
 		if (flags & DL_FLAG_RPM_ACTIVE)
-			pm_runtime_active_link(link, supplier);
+			refcount_inc(&link->rpm_active);
 
 		pm_runtime_new_link(consumer);
 	}
drivers/base/power/runtime.c +6 -23
Index: linux-pm/drivers/base/power/runtime.c
===================================================================
@@ -1656,6 +1656,7 @@ void pm_runtime_get_suppliers(struct dev
 
 	list_for_each_entry_rcu(link, &dev->links.suppliers, c_node)
 		if (link->flags & DL_FLAG_PM_RUNTIME) {
+			link->supplier_preactivated = true;
 			refcount_inc(&link->rpm_active);
 			pm_runtime_get_sync(link->supplier);
 		}
@@ -1675,9 +1676,11 @@ void pm_runtime_put_suppliers(struct dev
 	idx = device_links_read_lock();
 
 	list_for_each_entry_rcu(link, &dev->links.suppliers, c_node)
-		if (link->flags & DL_FLAG_PM_RUNTIME &&
-		    refcount_dec_not_one(&link->rpm_active))
-			pm_runtime_put(link->supplier);
+		if (link->supplier_preactivated) {
+			link->supplier_preactivated = false;
+			if (refcount_dec_not_one(&link->rpm_active))
+				pm_runtime_put(link->supplier);
+		}
 
 	device_links_read_unlock(idx);
 }
@@ -1689,26 +1692,6 @@ void pm_runtime_new_link(struct device *
 	spin_unlock_irq(&dev->power.lock);
 }
 
-/**
- * pm_runtime_active_link - Set up new device link as active for PM-runtime.
- * @link: Device link to be set up as active.
- * @supplier: Supplier end of the link.
- *
- * Add 2 to the rpm_active refcount of @link and increment the PM-runtime
- * usage counter of @supplier once more in case the link is being added while
- * the consumer driver is probing and pm_runtime_put_suppliers() will be called
- * subsequently.
- *
- * Note that this doesn't prevent rpm_put_suppliers() from decreasing the link's
- * rpm_active refcount down to one, so runtime suspend of the consumer end of
- * @link is not affected.
- */
-void pm_runtime_active_link(struct device_link *link, struct device *supplier)
-{
-	refcount_add(2, &link->rpm_active);
-	pm_runtime_get_noresume(supplier);
-}
-
 void pm_runtime_drop_link(struct device *dev)
 {
 	spin_lock_irq(&dev->power.lock);
include/linux/device.h +1 -0
Index: linux-pm/include/linux/device.h
===================================================================
@@ -861,6 +861,7 @@ struct device_link {
 #ifdef CONFIG_SRCU
 	struct rcu_head rcu_head;
 #endif
+	bool supplier_preactivated; /* Owned by consumer probe. */
 };
 
 /**
include/linux/pm_runtime.h +0 -4
Index: linux-pm/include/linux/pm_runtime.h
===================================================================
@@ -59,8 +59,6 @@ extern void pm_runtime_clean_up_links(st
 extern void pm_runtime_get_suppliers(struct device *dev);
 extern void pm_runtime_put_suppliers(struct device *dev);
 extern void pm_runtime_new_link(struct device *dev);
-extern void pm_runtime_active_link(struct device_link *link,
-				   struct device *supplier);
 extern void pm_runtime_drop_link(struct device *dev);
 
 static inline void pm_suspend_ignore_children(struct device *dev, bool enable)
@@ -180,8 +178,6 @@ static inline void pm_runtime_clean_up_l
 static inline void pm_runtime_get_suppliers(struct device *dev) {}
 static inline void pm_runtime_put_suppliers(struct device *dev) {}
 static inline void pm_runtime_new_link(struct device *dev) {}
-static inline void pm_runtime_active_link(struct device_link *link,
-					  struct device *supplier) {}
 static inline void pm_runtime_drop_link(struct device *dev) {}
 
 #endif /* !CONFIG_PM */
maintainer Tested-by on v1 1/1
#1 lore raw
Quoted 22 lines · Rafael J. Wysocki

From: Rafael J. Wysocki <[email protected]>

Commit 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance") introduced a regression that causes suppliers to be suspended prematurely for device links added during consumer driver probe if the initial PM-runtime status of the consumer is "suspended" and the consumer is resumed after adding the link and before pm_runtime_put_suppliers() is called. In that case, pm_runtime_put_suppliers() will drop the rpm_active refcount for the link by one and (since rpm_active is equal to two after the preceding consumer resume) the supplier's PM-runtime usage counter will be decremented, which may cause the supplier to suspend even though the consumer's PM-runtime status is "active".

For this reason, partially revert commit 4c06c4e6cf63 as the problem it tried to fix needs to be addressed somewhat differently, and change pm_runtime_get_suppliers() and pm_runtime_put_suppliers() so that the latter only drops rpm_active references acquired by the former. [This requires adding a new field to struct device_link, but I coulnd't find a cleaner way to address the issue that would work in all cases.]

Quoted 8 lines
Fixes 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance")

---

On top of driver-core/driver-core-next.

Jon, Ulf, please test!

Thanks Rafael this works for me!

Cheers
Jon

Signature · 1 line
nvpublic
Tested-by Reviewed-by on v1 1/1
#2 lore raw
In reply to Rafael J. Wysocki
Quoted 14 lines · Rafael J. Wysocki

From: Rafael J. Wysocki <[email protected]>

Commit 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance") introduced a regression that causes suppliers to be suspended prematurely for device links added during consumer driver probe if the initial PM-runtime status of the consumer is "suspended" and the consumer is resumed after adding the link and before pm_runtime_put_suppliers() is called. In that case, pm_runtime_put_suppliers() will drop the rpm_active refcount for the link by one and (since rpm_active is equal to two after the preceding consumer resume) the supplier's PM-runtime usage counter will be decremented, which may cause the supplier to suspend even though the consumer's PM-runtime status is "active".

Just to confirm, this is the behavior that I see as well.

However, I am a bit a surprised on my side, that I did not run this particular test sequence earlier. Sorry about that!

Quoted 9 lines

For this reason, partially revert commit 4c06c4e6cf63 as the problem it tried to fix needs to be addressed somewhat differently, and change pm_runtime_get_suppliers() and pm_runtime_put_suppliers() so that the latter only drops rpm_active references acquired by the former. [This requires adding a new field to struct device_link, but I coulnd't find a cleaner way to address the issue that would work in all cases.]

Quoted 3 lines
Fixes 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance")

Besides validating the fix, I can confirm device link add/removal during probe works fine, as before.

I also tested the runtime PM autosuspend (several seconds timeout) together with DL_FLAG_RPM_ACTIVE - it's working fine! During this test, the rpm_active count reaches 3 after probe and the supplier's usage count reaches 2. When the consumer becomes runtime suspended, after the timeout, these "counts" becomes decremented to 1 and 0, so the supplier also becomes runtime suspended.

Thanks and kind regards
Uffe

Quoted 124 lines

---

On top of driver-core/driver-core-next.

Jon, Ulf, please test!

---
 drivers/base/core.c          |    4 ++--
 drivers/base/power/runtime.c |   29 ++++++-----------------------
 include/linux/device.h       |    1 +
 include/linux/pm_runtime.h   |    4 ----
 4 files changed, 9 insertions(+), 29 deletions(-)

Index: linux-pm/drivers/base/core.c
===================================================================

drivers/base/core.c +2 -2
@@ -277,7 +277,7 @@ struct device_link *device_link_add(stru
                                link->flags |= DL_FLAG_PM_RUNTIME;
                        }
                        if (flags & DL_FLAG_RPM_ACTIVE)
-                               pm_runtime_active_link(link, supplier);
+                               refcount_inc(&link->rpm_active);
                }
 
                if (flags & DL_FLAG_STATELESS) {
@@ -310,7 +310,7 @@ struct device_link *device_link_add(stru
 
        if (flags & DL_FLAG_PM_RUNTIME) {
                if (flags & DL_FLAG_RPM_ACTIVE)
-                       pm_runtime_active_link(link, supplier);
+                       refcount_inc(&link->rpm_active);
 
                pm_runtime_new_link(consumer);
        }

Index: linux-pm/drivers/base/power/runtime.c
===================================================================

drivers/base/power/runtime.c +6 -23
@@ -1656,6 +1656,7 @@ void pm_runtime_get_suppliers(struct dev
 
        list_for_each_entry_rcu(link, &dev->links.suppliers, c_node)
                if (link->flags & DL_FLAG_PM_RUNTIME) {
+                       link->supplier_preactivated = true;
                        refcount_inc(&link->rpm_active);
                        pm_runtime_get_sync(link->supplier);
                }
@@ -1675,9 +1676,11 @@ void pm_runtime_put_suppliers(struct dev
        idx = device_links_read_lock();
 
        list_for_each_entry_rcu(link, &dev->links.suppliers, c_node)
-               if (link->flags & DL_FLAG_PM_RUNTIME &&
-                   refcount_dec_not_one(&link->rpm_active))
-                       pm_runtime_put(link->supplier);
+               if (link->supplier_preactivated) {
+                       link->supplier_preactivated = false;
+                       if (refcount_dec_not_one(&link->rpm_active))
+                               pm_runtime_put(link->supplier);
+               }
 
        device_links_read_unlock(idx);
 }
@@ -1689,26 +1692,6 @@ void pm_runtime_new_link(struct device *
        spin_unlock_irq(&dev->power.lock);
 }
 
-/**
- * pm_runtime_active_link - Set up new device link as active for PM-runtime.
- * @link: Device link to be set up as active.
- * @supplier: Supplier end of the link.
- *
- * Add 2 to the rpm_active refcount of @link and increment the PM-runtime
- * usage counter of @supplier once more in case the link is being added while
- * the consumer driver is probing and pm_runtime_put_suppliers() will be called
- * subsequently.
- *
- * Note that this doesn't prevent rpm_put_suppliers() from decreasing the link's
- * rpm_active refcount down to one, so runtime suspend of the consumer end of
- * @link is not affected.
- */
-void pm_runtime_active_link(struct device_link *link, struct device *supplier)
-{
-       refcount_add(2, &link->rpm_active);
-       pm_runtime_get_noresume(supplier);
-}
-
 void pm_runtime_drop_link(struct device *dev)
 {
        spin_lock_irq(&dev->power.lock);

Index: linux-pm/include/linux/device.h
===================================================================

include/linux/device.h +1 -0
@@ -861,6 +861,7 @@ struct device_link {
 #ifdef CONFIG_SRCU
        struct rcu_head rcu_head;
 #endif
+       bool supplier_preactivated; /* Owned by consumer probe. */
 };
 
 /**

Index: linux-pm/include/linux/pm_runtime.h
===================================================================

include/linux/pm_runtime.h +0 -4
@@ -59,8 +59,6 @@ extern void pm_runtime_clean_up_links(st
 extern void pm_runtime_get_suppliers(struct device *dev);
 extern void pm_runtime_put_suppliers(struct device *dev);
 extern void pm_runtime_new_link(struct device *dev);
-extern void pm_runtime_active_link(struct device_link *link,
-                                  struct device *supplier);
 extern void pm_runtime_drop_link(struct device *dev);
 
 static inline void pm_suspend_ignore_children(struct device *dev, bool enable)
@@ -180,8 +178,6 @@ static inline void pm_runtime_clean_up_l
 static inline void pm_runtime_get_suppliers(struct device *dev) {}
 static inline void pm_runtime_put_suppliers(struct device *dev) {}
 static inline void pm_runtime_new_link(struct device *dev) {}
-static inline void pm_runtime_active_link(struct device_link *link,
-                                         struct device *supplier) {}
 static inline void pm_runtime_drop_link(struct device *dev) {}
 
 #endif /* !CONFIG_PM */
#3 lore raw
Quoted 45 lines · Ulf Hansson
Quoted 14 lines · Rafael J. Wysocki

From: Rafael J. Wysocki <[email protected]>

Commit 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance") introduced a regression that causes suppliers to be suspended prematurely for device links added during consumer driver probe if the initial PM-runtime status of the consumer is "suspended" and the consumer is resumed after adding the link and before pm_runtime_put_suppliers() is called. In that case, pm_runtime_put_suppliers() will drop the rpm_active refcount for the link by one and (since rpm_active is equal to two after the preceding consumer resume) the supplier's PM-runtime usage counter will be decremented, which may cause the supplier to suspend even though the consumer's PM-runtime status is "active".

Just to confirm, this is the behavior that I see as well.

However, I am a bit a surprised on my side, that I did not run this particular test sequence earlier. Sorry about that!

Quoted 9 lines

For this reason, partially revert commit 4c06c4e6cf63 as the problem it tried to fix needs to be addressed somewhat differently, and change pm_runtime_get_suppliers() and pm_runtime_put_suppliers() so that the latter only drops rpm_active references acquired by the former. [This requires adding a new field to struct device_link, but I coulnd't find a cleaner way to address the issue that would work in all cases.]

Quoted 3 lines
Fixes 4c06c4e6cf63 ("driver core: Fix possible supplier PM-usage counter imbalance")

Besides validating the fix, I can confirm device link add/removal during probe works fine, as before.

I also tested the runtime PM autosuspend (several seconds timeout) together with DL_FLAG_RPM_ACTIVE - it's working fine! During this test, the rpm_active count reaches 3 after probe and the supplier's usage count reaches 2. When the consumer becomes runtime suspended, after the timeout, these "counts" becomes decremented to 1 and 0, so the supplier also becomes runtime suspended.

Quoted 2 lines

Many thanks to you and Jon for verifying it!