Skip to contentSkip to conversation

pci: fix slot trylock error handling

PATCH In review Keith Busch 5 participants 16 messages Jan 16 – Jan 29 PCI

lore
Share & tools
lore
#0 lore raw

PATCH 1/2 pci: fix slot trylock error handling

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")
---
 drivers/pci/pci.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/pci/pci.c +1 -3
index 72d88ea95f3cc..3378221c5723a 100644
@@ -5494,10 +5494,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;
 	}
Signature · 1 line
2.47.3
#1 lore raw

PATCH 2/2 pci: fix slot reset device locking

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +11 -1
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
Signature · 1 line
2.47.3
#2 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

[+cc Dan, Jinhui]

Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

All these almost-but-not-quite identical paths make my head hurt ;)

Quoted 52 lines

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +11 -1
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
Signature · 1 line
2.47.3
#3 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

Quoted 11 lines · Bjorn Helgaas

[+cc Dan, Jinhui]

Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

Oh, this patch is changing both pci_slot_trylock and pci_slot_lock since they were both missing the equivalent locks that the "bus" version of those functions were doing.

Quoted 3 lines

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

Honestly I think the _slot versions should go away. Those don't handle resetting a bus with multiple device's on it: only some functions get saved and restored even though the bus reset hits all the devices. I'm working on a fix for that, but it's more difficult than these patches.

Quoted 3 lines

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

It should, and this patch is changing pci_slot_lock() to do that.

Quoted 1 line

All these almost-but-not-quite identical paths make my head hurt ;)

I agree! And the functions that sound almost the same but work quite different? Looking at "pci_bus_reset" vs "pci_reset_bus" :)

Quoted 52 lines
Quoted 52 lines

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +11 -1
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
Signature · 1 line
2.47.3
Reviewed-by Dan Williams Reviewed-by Dan Williams
#4 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

Quoted 7 lines · Keith Busch
Quoted 6 lines · Bjorn Helgaas
Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

Quoted 7 lines
Quoted 3 lines

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

Oh, this patch is changing both pci_slot_trylock and pci_slot_lock since they were both missing the equivalent locks that the "bus" version of those functions were doing.

Wow. The patch has hunks for:

  pci_slot_lock()               # lock bridge
  pci_slot_unlock()             # unlock bridge
  pci_slot_trylock()            # lock bridge

which makes perfect sense.

But when I apply it on either pci/main (v6.19-rc1) or pci/next, "git am" puts those hunks in:

  pci_slot_unlock()             # unlock bridge (as expected)
  pci_slot_trylock()            # lock bridge (as expected)
  pci_slot_restore_locked()     # lock bridge (completely wrong spot)

I have no idea why the pci_slot_lock() hunk got applied in pci_slot_restore_locked(). I do notice that the line offsets in your patch are much different than mine, so you must have other changes in pci.c; maybe that accounts for the confusion.

Can you regenerate this patch based on v6.19-rc1? And maybe incorporate Dan's "bridge = slot->bus->self" idea at the same time?

I'll attach my "git log -p" at the bottom if you want to see why I was so confused. What a day ;)

Quoted 69 lines
Quoted 3 lines

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

Honestly I think the _slot versions should go away. Those don't handle resetting a bus with multiple device's on it: only some functions get saved and restored even though the bus reset hits all the devices. I'm working on a fix for that, but it's more difficult than these patches.

Quoted 3 lines

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

It should, and this patch is changing pci_slot_lock() to do that.

Quoted 1 line

All these almost-but-not-quite identical paths make my head hurt ;)

I agree! And the functions that sound almost the same but work quite different? Looking at "pci_bus_reset" vs "pci_reset_bus" :)

Quoted 49 lines
Quoted 49 lines

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +11 -1
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Signature · 1 line
2.47.3
$ git log -p v6.19-rc1.. | cat
commit 44c651ea87f9 ("pci: fix slot reset device locking")
Author: Keith Busch <[email protected]>
Date:   Fri Jan 16 10:41:50 2026 -0800

pci: fix slot reset device locking

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

drivers/pci/pci.c +11 -1
index 59319e08fca6..73764e66cabd 100644
@@ -5335,12 +5335,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5363,6 +5368,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
@@ -5425,6 +5433,8 @@ static void pci_slot_restore_locked(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;

commit 3f2aea31058a ("pci: fix slot trylock error handling")

Author: Keith Busch <[email protected]>
Date:   Fri Jan 16 10:41:49 2026 -0800

pci: fix slot trylock error handling

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")
drivers/pci/pci.c +1 -3
index 13dbb405dc31..59319e08fca6 100644
@@ -5346,10 +5346,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;
 	}
#5 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

Quoted 5 lines · Bjorn Helgaas

I have no idea why the pci_slot_lock() hunk got applied in pci_slot_restore_locked(). I do notice that the line offsets in your patch are much different than mine, so you must have other changes in pci.c; maybe that accounts for the confusion.

Oh weird. I'd hope that patch just wouldn't have applied in that case instead of appearing successful.

I was using a newer branch that also included some other fixes I'm working on, so not surprised the line numbers didn't match up. I'll rebase future patches to v6.19-rc1.

Quoted 2 lines

Can you regenerate this patch based on v6.19-rc1? And maybe incorporate Dan's "bridge = slot->bus->self" idea at the same time?

I've got a new proposal about ready that just drops the slot specific lock/unlock and save/restore instead. We kind of need to do that since slot reset breaks devices with more than 8 functions. I'll send the patch with the details in a few minutes.

#6 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

In reply to Bjorn Helgaas
Quoted 16 lines · Bjorn Helgaas

[+cc Dan, Jinhui]

Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

Quoted 5 lines

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

All these almost-but-not-quite identical paths make my head hurt ;)

My series does attempt to consolidation them (check the pending patches from me).

(It doesn't consider this fix patch though but the other one is taken into account, the series was build on top of it).

Signature · 1 line
 i.
Quoted 53 lines
Quoted 52 lines

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +11 -1
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;
 }
 
Signature · 1 line
2.47.3
#7 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

Quoted 22 lines · Ilpo Järvinen
Quoted 20 lines · Bjorn Helgaas

[+cc Dan, Jinhui]

Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

Quoted 7 lines
Quoted 1 line

All these almost-but-not-quite identical paths make my head hurt ;)

My series does attempt to consolidation them (check the pending patches from me).

(It doesn't consider this fix patch though but the other one is taken into account, the series was build on top of it).

Lore link?

#8 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

Quoted 22 lines · On Wed, Jan 28, 2026 at 01:07:37PM -0800, [email protected] wrote:
Quoted 21 lines · Ilpo Järvinen
Quoted 19 lines · Bjorn Helgaas
Quoted 4 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

I *think* this actually refers to pci_bus_trylock() and pci_slot_trylock() (not pci_bus_lock() and pci_slot_lock()), since that's what this patch changes?

It's unfortunate that pci_bus_trylock() and pci_slot_trylock() are so similar but separate. If there were combined, this kind of issue where one is fixed but the other isn't wouldn't happen.

But what about pci_bus_lock() and pci_slot_lock()? They are also almost identical, but pci_bus_lock() locks bus->self while pci_slot_lock() does not. Should it?

All these almost-but-not-quite identical paths make my head hurt ;)

Quoted 7 lines
Quoted 5 lines

My series does attempt to consolidation them (check the pending patches from me).

(It doesn't consider this fix patch though but the other one is taken into account, the series was build on top of it).

Lore link?

I think it's
https://lore.kernel.org/all/[email protected]/

Reviewed-by on v1 2/2
#9 lore raw

Re: [PATCH 2/2] pci: fix slot reset device locking

In reply to Keith Busch
Quoted 26 lines · Keith Busch

From: Keith Busch <[email protected]>

Like pci_bus_lock, pci_slot_lock needs to lock the bridge device to prevent the warning:

pcieport 0000:e2:05.0: unlocked secondary bus reset via: pciehp_reset_slot+0x55/0xa0

---
 drivers/pci/pci.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/pci/pci.c +2 -0
index 3378221c5723a..5f8b0d06a1459 100644
@@ -5460,6 +5460,8 @@ static void pci_slot_lock(struct pci_slot *slot)
 {
 	struct pci_dev *dev;
 
+	if (slot->bus->self)
+		pci_dev_lock(slot->bus->self);
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)
 			continue;

drivers/pci/pci.c static void pci_slot_unlock(struct pci_slot *slot) · line 5489

Quoted 5 lines
drivers/pci/pci.c +2 -0
@@ -5483,12 +5485,17 @@ static void pci_slot_unlock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+	if (slot->bus->self)
+		pci_dev_unlock(slot->bus->self);

Given how much context is needed to understand that this looks right, I would not say no to a follow-on patch that adds kdoc for these locking functions.

At a minimum I think a breadcrumb of a local "bridge = slot->bus->self" variable would help.

Quoted 14 lines
drivers/pci/pci.c +4 -1
 }
 
 /* Return 1 on successful lock, 0 on contention */
 static int pci_slot_trylock(struct pci_slot *slot)
 {
-	struct pci_dev *dev;
+	struct pci_dev *dev, *bridge = slot->bus->self;
+
+	if (bridge && !pci_dev_trylock(bridge))
+		return 0;
 
 	list_for_each_entry(dev, &slot->bus->devices, bus_list) {
 		if (!dev->slot || dev->slot != slot)

drivers/pci/pci.c static int pci_slot_trylock(struct pci_slot *slot) · line 5524

Quoted 7 lines
drivers/pci/pci.c +3 -0
@@ -5511,6 +5518,9 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		else
 			pci_dev_unlock(dev);
 	}
+
+	if (bridge)
+		pci_dev_unlock(bridge);
 	return 0;

Even without the above cosmetic fixup, looks good.

#10 lore raw
In reply to Keith Busch
Quoted 7 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")

Ping?

Quoted 23 lines
---
 drivers/pci/pci.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/pci/pci.c +1 -3
index 72d88ea95f3cc..3378221c5723a 100644
@@ -5494,10 +5494,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;
 	}
Signature · 1 line
2.47.3
#11 lore raw
Quoted 10 lines · Keith Busch
Quoted 7 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")

Ping?

There's the older patch from Jinhui for the same thing:

https://patchwork.kernel.org/project/linux-pci/patch/[email protected]/

Quoted 24 lines
Quoted 23 lines
---
 drivers/pci/pci.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/pci/pci.c +1 -3
index 72d88ea95f3cc..3378221c5723a 100644
@@ -5494,10 +5494,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;
 	}
Signature · 1 line
2.47.3
Signature · 1 line
 i.
#12 lore raw
Quoted 8 lines · Ilpo Järvinen
Quoted 6 lines · Keith Busch
Quoted 5 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")

Ping?

There's the older patch from Jinhui for the same thing:

https://patchwork.kernel.org/project/linux-pci/patch/[email protected]/

Okay, let's go with that one then. But patch 2 from this series is still needed.

#13 lore raw
Quoted 12 lines · Keith Busch
Quoted 11 lines · Ilpo Järvinen
Quoted 9 lines · Keith Busch
Quoted 7 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")
Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Ping?

There's the older patch from Jinhui for the same thing:

https://patchwork.kernel.org/project/linux-pci/patch/[email protected]/

Okay, let's go with that one then. But patch 2 from this series is still needed.

Yes, agreed. I meant to mention that about the patch 2 specifically but forgot.

Signature · 1 line
 i.
#14 lore raw
In reply to Keith Busch

[+cc Dan, author of a4e772898f8bf2]

Quoted 30 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")
---
 drivers/pci/pci.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/pci/pci.c +1 -3
index 72d88ea95f3cc..3378221c5723a 100644
@@ -5494,10 +5494,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;
 	}
Signature · 1 line
2.47.3
Reviewed-by on v1 1/2
#15 lore raw
In reply to Keith Busch
Quoted 16 lines · Keith Busch

From: Keith Busch <[email protected]>

The device lock isn't held if pci_bus_trylock() fails, so the code had been attempting to improperly unlock it.

Fixes a4e772898f8bf2 ("PCI: Add missing bridge lock to pci_bus_lock()")
---
 drivers/pci/pci.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/pci/pci.c +0 -0
index 72d88ea95f3cc..3378221c5723a 100644

drivers/pci/pci.c static int pci_slot_trylock(struct pci_slot *slot) · line 5500

Quoted 10 lines
drivers/pci/pci.c +1 -3
@@ -5494,10 +5494,8 @@ static int pci_slot_trylock(struct pci_slot *slot)
 		if (!dev->slot || dev->slot != slot)
 			continue;
 		if (dev->subordinate) {
-			if (!pci_bus_trylock(dev->subordinate)) {
-				pci_dev_unlock(dev);
+			if (!pci_bus_trylock(dev->subordinate))
 				goto unlock;
-			}
 		} else if (!pci_dev_trylock(dev))
 			goto unlock;

This looks right and the original patch missed fixing up this inner unlock when it moved the @dev lock to be conditional for non-bridge devices.