pci: remove slot specific lock/unlock and save/restore
PATCH 1/1 pci: remove slot specific lock/unlock and save/restore
From: Keith Busch <[email protected]>
The Linux pci driver resolves a "slot" to the "D" in the B:D.f (see PCI_SLOT()). A pcie "slot reset" is a secondary bus reset, which affects every function on every "D", not just the ones with a matching "slot". The slot lock/unlock and save/restore functions, however, are only handling a subset of the functions, breaking the rest.
ARI devices with more than 8 functions fail because their state is not properly handled, nor is the attached driver notified of the reset. In the best case, the device will appear unresponsive to the driver, resulting in unexpected errors. A worse possiblility may panic the kernel if in flight transactions trigger hardware reported errors like this real observation:
vfio-pci 0000:01:00.0: resetting
vfio-pci 0000:01:00.0: reset done
{1}[Hardware Error]: Error 1, type: fatal
{1}[Hardware Error]: section_type: PCIe error
{1}[Hardware Error]: port_type: 0, PCIe end point
{1}[Hardware Error]: version: 0.2
{1}[Hardware Error]: command: 0x0140, status: 0x0010
{1}[Hardware Error]: device_id: 0000:01:01.0
{1}[Hardware Error]: slot: 0
{1}[Hardware Error]: secondary_bus: 0x00
{1}[Hardware Error]: vendor_id: 0x1d9b, device_id: 0x0207
{1}[Hardware Error]: class_code: 020000
{1}[Hardware Error]: bridge: secondary_status: 0x0000, control: 0x0000
{1}[Hardware Error]: aer_cor_status: 0x00008000, aer_cor_mask: 0x00002000
{1}[Hardware Error]: aer_uncor_status: 0x00010000, aer_uncor_mask: 0x00100000
{1}[Hardware Error]: aer_uncor_severity: 0x006f6030
{1}[Hardware Error]: TLP Header: 0a412800 00192080 60000004 00000004
GHES: Fatal hardware error but panic disabled
Kernel panic - not syncing: GHES: Fatal hardware errorFix this by properly locking and notifying the entire affected bus topology, not just specific matching slots. For architectures that support "slot" specific resets, this patch potentially introduces an insignificant amount of overhead, but is otherwise harmless.
---
This patch is based on v6.19-rc1 as requested. Note, there will be a minor merge conflict if this other patch fixing the "trylock" handling is already included:
https://lore.kernel.org/linux-pci/20260128225344.GA422463@bhelgaas/
But I don't see it applied in any the pci git tree.
drivers/pci/pci.c | 113 ++++------------------------------------------ 1 file changed, 9 insertions(+), 104 deletions(-)
drivers/pci/pci.c +9 -104
@@ -5307,67 +5307,6 @@ static bool pci_slot_resettable(struct pci_slot *slot)return true; } -/* Lock devices from the top of the tree down */ -static void pci_slot_lock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_lock (dev->subordinate); -else -pci_dev_lock (dev); - } -} - -/* Unlock devices from the bottom of the tree up */ -static void pci_slot_unlock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_unlock (dev->subordinate); -else -pci_dev_unlock (dev); - } -} - -/* Return 1 on successful lock, 0 on contention */ -static int pci_slot_trylock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) { -if (!pci_bus_trylock (dev->subordinate)) { -pci_dev_unlock (dev); - goto unlock; - } - }else if (!pci_dev_trylock (dev)) - goto unlock; - } -return 1 ; - -unlock: -list_for_each_entry_continue_reverse (dev, - &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_unlock (dev->subordinate); -else -pci_dev_unlock (dev); - } -return 0 ; -} -/* * Save and disable devices from the top of the tree down while holding * the @dev mutex lock for the entire tree.
@@ -5401,59 +5340,24 @@ static void pci_bus_restore_locked(struct pci_bus *bus) } } -/* -* Save and disable devices from the top of the tree down while holding -* the @dev mutex lock for the entire tree. -*/ -static void pci_slot_save_and_disable_locked (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -pci_dev_save_and_disable (dev); -if (dev->subordinate) -pci_bus_save_and_disable_locked (dev->subordinate); - } -} - -/* -* Restore devices from top of the tree down while holding @dev mutex lock -* for the entire tree. Parent bridges need to be restored before we can -* get to subordinate devices. -*/ -static void pci_slot_restore_locked (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -pci_dev_restore (dev); -if (dev->subordinate) { -pci_bridge_wait_for_secondary_bus (dev,"slot reset" ); -pci_bus_restore_locked (dev->subordinate); - } - } -} -static int pci_slot_reset (struct pci_slot *slot,bool probe) { +struct pci_bus *bus;int rc;if (!slot || !pci_slot_resettable (slot))return -ENOTTY ; + bus = slot->bus;if (!probe) -pci_slot_lock (slot); +pci_bus_lock (bus);might_sleep (); rc =pci_reset_hotplug_slot (slot->hotplug, probe);if (!probe) -pci_slot_unlock (slot); +pci_bus_unlock (bus);return rc; }
@@ -5487,18 +5391,19 @@ EXPORT_SYMBOL_GPL(pci_probe_reset_slot);*/ static int __pci_reset_slot (struct pci_slot *slot) { +struct pci_bus *bus = slot->bus;int rc; rc =pci_slot_reset (slot,PCI_RESET_PROBE );if (rc)return rc; -if (pci_slot_trylock (slot)) { -pci_slot_save_and_disable_locked (slot); +if (pci_bus_trylock (bus)) { +pci_bus_save_and_disable_locked (bus);might_sleep (); rc =pci_reset_hotplug_slot (slot->hotplug,PCI_RESET_DO_RESET ); -pci_slot_restore_locked (slot); -pci_slot_unlock (slot); +pci_bus_restore_locked (bus); +pci_bus_unlock (bus); }else rc = -EAGAIN ;
Signature · 1 line
2.47.3
Quoted 35 lines · Keith Busch
From: Keith Busch <[email protected]>
The Linux pci driver resolves a "slot" to the "D" in the B:D.f (see
PCI_SLOT()). A pcie "slot reset" is a secondary bus reset, which affects every function on every "D", not just the ones with a matching "slot". The slot lock/unlock and save/restore functions, however, are only handling a subset of the functions, breaking the rest.ARI devices with more than 8 functions fail because their state is not properly handled, nor is the attached driver notified of the reset. In the best case, the device will appear unresponsive to the driver, resulting in unexpected errors. A worse possiblility may panic the kernel if in flight transactions trigger hardware reported errors like this real observation:
vfio-pci 0000:01:00.0: resetting
vfio-pci 0000:01:00.0: reset done
{1}[Hardware Error]: Error 1, type: fatal{1}[Hardware Error]: section_type: PCIe error {1}[Hardware Error]: port_type: 0, PCIe end point {1}[Hardware Error]: version: 0.2 {1}[Hardware Error]: command: 0x0140, status: 0x0010 {1}[Hardware Error]: device_id: 0000:01:01.0 {1}[Hardware Error]: slot: 0 {1}[Hardware Error]: secondary_bus: 0x00 {1}[Hardware Error]: vendor_id: 0x1d9b, device_id: 0x0207 {1}[Hardware Error]: class_code: 020000 {1}[Hardware Error]: bridge: secondary_status: 0x0000, control: 0x0000 {1}[Hardware Error]: aer_cor_status: 0x00008000, aer_cor_mask: 0x00002000 {1}[Hardware Error]: aer_uncor_status: 0x00010000, aer_uncor_mask: 0x00100000 {1}[Hardware Error]: aer_uncor_severity: 0x006f6030 {1}[Hardware Error]: TLP Header: 0a412800 00192080 60000004 00000004 GHES: Fatal hardware error but panic disabled Kernel panic - not syncing: GHES: Fatal hardware error
Quoted 7 lines
Fix this by properly locking and notifying the entire affected bus topology, not just specific matching slots. For architectures that support "slot" specific resets, this patch potentially introduces an insignificant amount of overhead, but is otherwise harmless.
Signed-off-by Keith Busch <[email protected]>---
This seems reasonable, but is a much larger change than just fixing the slot locking, so it would really be nice to get some review besides just me.
Quoted 5 lines
This patch is based on v6.19-rc1 as requested. Note, there will be a minor merge conflict if this other patch fixing the "trylock" handling is already included:
https://lore.kernel.org/linux-pci/20260128225344.GA422463@bhelgaas/
The patch below removes pci_slot_trylock() completely, which renders Jinhui's fix to that function moot, so if we apply this one, there would be no reason to keep Jinhui's patch.
An alternate path would be to:
- keep Jinhui's patch that fixes the pci_slot_trylock() error path (same as your 1/1 patch)
- apply your 2/2 patch [1] to fix pci_slot_lock(), pci_slot_unlock(), and pci_slot_trylock() to lock the bridge (after accounting for the mistaken merge)
- finally, rebase *this* patch on top to remove the pci_slot_lock(), pci_slot_unlock(), and pci_slot_trylock() that have just been fixed
The advantage would be that if there's any issue or objection to completely removing pci_slot_lock() et al, we could drop or defer that while still fixing the locking issues.
[1] https://lore.kernel.org/r/[email protected]
Quoted 168 lines
But I don't see it applied in any the pci git tree.
drivers/pci/pci.c | 113 ++++------------------------------------------ 1 file changed, 9 insertions(+), 104 deletions(-)drivers/pci/pci.c +9
-104@@ -5307,67 +5307,6 @@ static bool pci_slot_resettable(struct pci_slot *slot)return true; } -/* Lock devices from the top of the tree down */ -static void pci_slot_lock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_lock (dev->subordinate); -else -pci_dev_lock (dev); - } -} - -/* Unlock devices from the bottom of the tree up */ -static void pci_slot_unlock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_unlock (dev->subordinate); -else -pci_dev_unlock (dev); - } -} - -/* Return 1 on successful lock, 0 on contention */ -static int pci_slot_trylock (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) { -if (!pci_bus_trylock (dev->subordinate)) { -pci_dev_unlock (dev); - goto unlock; - } - }else if (!pci_dev_trylock (dev)) - goto unlock; - } -return 1 ; - -unlock: -list_for_each_entry_continue_reverse (dev, - &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -if (dev->subordinate) -pci_bus_unlock (dev->subordinate); -else -pci_dev_unlock (dev); - } -return 0 ; -} -/* * Save and disable devices from the top of the tree down while holding * the @dev mutex lock for the entire tree. @@ -5401,59 +5340,24 @@ static void pci_bus_restore_locked(struct pci_bus *bus) } } -/* -* Save and disable devices from the top of the tree down while holding -* the @dev mutex lock for the entire tree. -*/ -static void pci_slot_save_and_disable_locked (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -pci_dev_save_and_disable (dev); -if (dev->subordinate) -pci_bus_save_and_disable_locked (dev->subordinate); - } -} - -/* -* Restore devices from top of the tree down while holding @dev mutex lock -* for the entire tree. Parent bridges need to be restored before we can -* get to subordinate devices. -*/ -static void pci_slot_restore_locked (struct pci_slot *slot) -{ -struct pci_dev *dev; - -list_for_each_entry (dev, &slot->bus->devices, bus_list) { -if (!dev->slot || dev->slot != slot) -continue ; -pci_dev_restore (dev); -if (dev->subordinate) { -pci_bridge_wait_for_secondary_bus (dev,"slot reset" ); -pci_bus_restore_locked (dev->subordinate); - } - } -} -static int pci_slot_reset (struct pci_slot *slot,bool probe) { +struct pci_bus *bus;int rc;if (!slot || !pci_slot_resettable (slot))return -ENOTTY ; + bus = slot->bus;if (!probe) -pci_slot_lock (slot); +pci_bus_lock (bus);might_sleep (); rc =pci_reset_hotplug_slot (slot->hotplug, probe);if (!probe) -pci_slot_unlock (slot); +pci_bus_unlock (bus);return rc; }@@ -5487,18 +5391,19 @@ EXPORT_SYMBOL_GPL(pci_probe_reset_slot);*/ static int __pci_reset_slot (struct pci_slot *slot) { +struct pci_bus *bus = slot->bus;int rc; rc =pci_slot_reset (slot,PCI_RESET_PROBE );if (rc)return rc; -if (pci_slot_trylock (slot)) { -pci_slot_save_and_disable_locked (slot); +if (pci_bus_trylock (bus)) { +pci_bus_save_and_disable_locked (bus);might_sleep (); rc =pci_reset_hotplug_slot (slot->hotplug,PCI_RESET_DO_RESET ); -pci_slot_restore_locked (slot); -pci_slot_unlock (slot); +pci_bus_restore_locked (bus); +pci_bus_unlock (bus); }else rc = -EAGAIN ;Signature · 1 line
2.47.3
Quoted 9 lines · Bjorn Helgaas
An alternate path would be to:
- keep Jinhui's patch that fixes the
pci_slot_trylock()error path (same as your 1/1 patch)- apply your 2/2 patch [1] to fix
pci_slot_lock(),pci_slot_unlock(), andpci_slot_trylock()to lock the bridge (after accounting for the mistaken merge)
Quoted 7 lines
- finally, rebase *this* patch on top to remove the
pci_slot_lock(),pci_slot_unlock(), andpci_slot_trylock()that have just been fixedThe advantage would be that if there's any issue or objection to completely removing
pci_slot_lock()et al, we could drop or defer that while still fixing the locking issues.
Sounds good. Another bonus is we can be more cautious for stable backports since the first two are clear isolated bug fixes. I'll prepare a rebased version with the proposed sequence.