diff mbox series

[V2] PCI: pciehp: Disable ACS Source Validation during hot-remove

Message ID 20230111190533.29979-1-vidyas@nvidia.com (mailing list archive)
State Changes Requested
Delegated to: Bjorn Helgaas
Headers show
Series [V2] PCI: pciehp: Disable ACS Source Validation during hot-remove | expand

Commit Message

Vidya Sagar Jan. 11, 2023, 7:05 p.m. UTC
PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted to
send upstream messages before they have been assigned a bus number and
such messages have a Requester ID with Bus number set to 00h.
If the Downstrem port has ACS Source Validation enabled, these messages
will be detected as ACS violation error.
Hence, disable ACS Source Validation in the bridge device during
hot-remove operation and re-enable it after enumeration of the
downstream hierarchy but before binding the respective device drivers.

Signed-off-by: Vidya Sagar <vidyas@nvidia.com>
---
v2:
* Fixed build issues

 drivers/pci/hotplug/pciehp_pci.c | 13 ++++++++++++-
 drivers/pci/pci.c                | 22 ++++++++++++++++++++++
 include/linux/pci.h              |  6 ++++++
 3 files changed, 40 insertions(+), 1 deletion(-)

Comments

Bjorn Helgaas Feb. 14, 2023, 11:30 p.m. UTC | #1
[+cc Lukas, Alex for pciehp and ACS comments]

On Thu, Jan 12, 2023 at 12:35:33AM +0530, Vidya Sagar wrote:
> PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted to
> send upstream messages before they have been assigned a bus number and
> such messages have a Requester ID with Bus number set to 00h.
> If the Downstrem port has ACS Source Validation enabled, these messages
> will be detected as ACS violation error.
> Hence, disable ACS Source Validation in the bridge device during
> hot-remove operation and re-enable it after enumeration of the
> downstream hierarchy but before binding the respective device drivers.

s/Downstrem/Downstream/

Format as paragraphs (blank line between).

> Signed-off-by: Vidya Sagar <vidyas@nvidia.com>
> ---
> v2:
> * Fixed build issues
> 
>  drivers/pci/hotplug/pciehp_pci.c | 13 ++++++++++++-
>  drivers/pci/pci.c                | 22 ++++++++++++++++++++++
>  include/linux/pci.h              |  6 ++++++
>  3 files changed, 40 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/hotplug/pciehp_pci.c b/drivers/pci/hotplug/pciehp_pci.c
> index d17f3bf36f70..ad90bcf3f621 100644
> --- a/drivers/pci/hotplug/pciehp_pci.c
> +++ b/drivers/pci/hotplug/pciehp_pci.c
> @@ -63,6 +63,7 @@ int pciehp_configure_device(struct controller *ctrl)
>  
>  	pci_assign_unassigned_bridge_resources(bridge);
>  	pcie_bus_configure_settings(parent);
> +	pci_configure_acs_sv(bridge, true);
>  	pci_bus_add_devices(parent);
>  
>   out:
> @@ -117,6 +118,16 @@ void pciehp_unconfigure_device(struct controller *ctrl, bool presence)
>  		}
>  		pci_dev_put(dev);
>  	}
> -
> +	/*
> +	 * PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted
> +	 * to send upstream messages before they have been assigned a bus
> +	 * number and such messages have a Requester ID with Bus number
> +	 * set to 00h. If the Downstrem port has ACS Source Validation enabled,

s/Downstrem/Downstream/

> +	 * these messages will be detected as ACS violation error.
> +	 * Hence, disable ACS Source Validation here and re-enable it after
> +	 * enumeration of the downstream hierarchy and before binding the
> +	 * respective device drivers in pciehp_configure_device().
> +	 */
> +	pci_configure_acs_sv(ctrl->pcie->port, false);

What if we have a slot that's empty at boot and we add a device later?
It looks like we still might see ACS errors there because the add
happens before a remove?

>  	pci_unlock_rescan_remove();
>  }
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index 95bc329e74c0..9cefaf814f49 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -991,6 +991,28 @@ static void pci_enable_acs(struct pci_dev *dev)
>  	pci_disable_acs_redir(dev);
>  }
>  
> +#ifdef CONFIG_HOTPLUG_PCI_PCIE
> +void pci_configure_acs_sv(struct pci_dev *dev, bool flag)
> +{
> +	u16 cap;
> +	u16 ctrl;
> +
> +	if (!pci_acs_enable || !dev->acs_cap)
> +		return;
> +
> +	pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CAP, &cap);
> +	pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, &ctrl);
> +
> +	if (flag)
> +		ctrl |= (cap & PCI_ACS_SV);
> +	else
> +		ctrl &= ~(cap & PCI_ACS_SV);
> +
> +	pci_write_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, ctrl);

I guess we don't have a way to do this for the non-standard ACS-like
devices, i.e., pci_dev_specific_enable_acs().  Not the end of the
world, just unfortunate that we'll have different behavior there.

> +}
> +EXPORT_SYMBOL_GPL(pci_configure_acs_sv);

Doesn't seem like this needs to be exported or exposed via
linux/pci.h.  pciehp cannot be built as a module.

> +#endif
> +
>  /**
>   * pci_restore_bars - restore a device's BAR values (e.g. after wake-up)
>   * @dev: PCI device to have its BARs restored
> diff --git a/include/linux/pci.h b/include/linux/pci.h
> index 060af91bafcd..edf516e39764 100644
> --- a/include/linux/pci.h
> +++ b/include/linux/pci.h
> @@ -2239,6 +2239,12 @@ void pci_hp_create_module_link(struct pci_slot *pci_slot);
>  void pci_hp_remove_module_link(struct pci_slot *pci_slot);
>  #endif
>  
> +#ifdef CONFIG_HOTPLUG_PCI_PCIE
> +void pci_configure_acs_sv(struct pci_dev *dev, bool flag);
> +#else
> +static inline void pci_configure_acs_sv(struct pci_dev *dev, bool flag) { }
> +#endif
> +
>  /**
>   * pci_pcie_cap - get the saved PCIe capability offset
>   * @dev: PCI device
> -- 
> 2.17.1
>
Vidya Sagar July 30, 2023, 7:14 p.m. UTC | #2
On 2/15/2023 5:00 AM, Bjorn Helgaas wrote:
> External email: Use caution opening links or attachments
> 
> 
> [+cc Lukas, Alex for pciehp and ACS comments]
> 
> On Thu, Jan 12, 2023 at 12:35:33AM +0530, Vidya Sagar wrote:
>> PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted to
>> send upstream messages before they have been assigned a bus number and
>> such messages have a Requester ID with Bus number set to 00h.
>> If the Downstrem port has ACS Source Validation enabled, these messages
>> will be detected as ACS violation error.
>> Hence, disable ACS Source Validation in the bridge device during
>> hot-remove operation and re-enable it after enumeration of the
>> downstream hierarchy but before binding the respective device drivers.
> 
> s/Downstrem/Downstream/
> 
> Format as paragraphs (blank line between).
> 
>> Signed-off-by: Vidya Sagar <vidyas@nvidia.com>
>> ---
>> v2:
>> * Fixed build issues
>>
>>   drivers/pci/hotplug/pciehp_pci.c | 13 ++++++++++++-
>>   drivers/pci/pci.c                | 22 ++++++++++++++++++++++
>>   include/linux/pci.h              |  6 ++++++
>>   3 files changed, 40 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/pci/hotplug/pciehp_pci.c b/drivers/pci/hotplug/pciehp_pci.c
>> index d17f3bf36f70..ad90bcf3f621 100644
>> --- a/drivers/pci/hotplug/pciehp_pci.c
>> +++ b/drivers/pci/hotplug/pciehp_pci.c
>> @@ -63,6 +63,7 @@ int pciehp_configure_device(struct controller *ctrl)
>>
>>        pci_assign_unassigned_bridge_resources(bridge);
>>        pcie_bus_configure_settings(parent);
>> +     pci_configure_acs_sv(bridge, true);
>>        pci_bus_add_devices(parent);
>>
>>    out:
>> @@ -117,6 +118,16 @@ void pciehp_unconfigure_device(struct controller *ctrl, bool presence)
>>                }
>>                pci_dev_put(dev);
>>        }
>> -
>> +     /*
>> +      * PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted
>> +      * to send upstream messages before they have been assigned a bus
>> +      * number and such messages have a Requester ID with Bus number
>> +      * set to 00h. If the Downstrem port has ACS Source Validation enabled,
> 
> s/Downstrem/Downstream/
> 
>> +      * these messages will be detected as ACS violation error.
>> +      * Hence, disable ACS Source Validation here and re-enable it after
>> +      * enumeration of the downstream hierarchy and before binding the
>> +      * respective device drivers in pciehp_configure_device().
>> +      */
>> +     pci_configure_acs_sv(ctrl->pcie->port, false);
> 
> What if we have a slot that's empty at boot and we add a device later?
> It looks like we still might see ACS errors there because the add
> happens before a remove?
Yes I'm afraid. But, looking at the implementation note in 6.22.1 and 
6.12.1.1, I think it is expected and spec thought through this scenario.
So, does it all come down to how the ASIC designers treat different 
messages (case-by-case) and see which of them need not have to be 
validated for source? For example, a DRS (Device Readiness Status) from 
an endpoint during enumeration (i.e. before the endpoint is being 
assigned with a valid bus number) can be ignored for ACS source validation?

> 
>>        pci_unlock_rescan_remove();
>>   }
>> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
>> index 95bc329e74c0..9cefaf814f49 100644
>> --- a/drivers/pci/pci.c
>> +++ b/drivers/pci/pci.c
>> @@ -991,6 +991,28 @@ static void pci_enable_acs(struct pci_dev *dev)
>>        pci_disable_acs_redir(dev);
>>   }
>>
>> +#ifdef CONFIG_HOTPLUG_PCI_PCIE
>> +void pci_configure_acs_sv(struct pci_dev *dev, bool flag)
>> +{
>> +     u16 cap;
>> +     u16 ctrl;
>> +
>> +     if (!pci_acs_enable || !dev->acs_cap)
>> +             return;
>> +
>> +     pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CAP, &cap);
>> +     pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, &ctrl);
>> +
>> +     if (flag)
>> +             ctrl |= (cap & PCI_ACS_SV);
>> +     else
>> +             ctrl &= ~(cap & PCI_ACS_SV);
>> +
>> +     pci_write_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, ctrl);
> 
> I guess we don't have a way to do this for the non-standard ACS-like
> devices, i.e., pci_dev_specific_enable_acs().  Not the end of the
> world, just unfortunate that we'll have different behavior there.
I'll remove the exporting part in the next patch. As far as exposing it 
in the pci.h is concerned, since the actual usage of this function is in 
pciehp_pci.c file, I thought the best place to keep the declaration is 
pci.h file. Let me know if you have any other suggestion.

> 
>> +}
>> +EXPORT_SYMBOL_GPL(pci_configure_acs_sv);
> 
> Doesn't seem like this needs to be exported or exposed via
> linux/pci.h.  pciehp cannot be built as a module.
> 
>> +#endif
>> +
>>   /**
>>    * pci_restore_bars - restore a device's BAR values (e.g. after wake-up)
>>    * @dev: PCI device to have its BARs restored
>> diff --git a/include/linux/pci.h b/include/linux/pci.h
>> index 060af91bafcd..edf516e39764 100644
>> --- a/include/linux/pci.h
>> +++ b/include/linux/pci.h
>> @@ -2239,6 +2239,12 @@ void pci_hp_create_module_link(struct pci_slot *pci_slot);
>>   void pci_hp_remove_module_link(struct pci_slot *pci_slot);
>>   #endif
>>
>> +#ifdef CONFIG_HOTPLUG_PCI_PCIE
>> +void pci_configure_acs_sv(struct pci_dev *dev, bool flag);
>> +#else
>> +static inline void pci_configure_acs_sv(struct pci_dev *dev, bool flag) { }
>> +#endif
>> +
>>   /**
>>    * pci_pcie_cap - get the saved PCIe capability offset
>>    * @dev: PCI device
>> --
>> 2.17.1
>>
diff mbox series

Patch

diff --git a/drivers/pci/hotplug/pciehp_pci.c b/drivers/pci/hotplug/pciehp_pci.c
index d17f3bf36f70..ad90bcf3f621 100644
--- a/drivers/pci/hotplug/pciehp_pci.c
+++ b/drivers/pci/hotplug/pciehp_pci.c
@@ -63,6 +63,7 @@  int pciehp_configure_device(struct controller *ctrl)
 
 	pci_assign_unassigned_bridge_resources(bridge);
 	pcie_bus_configure_settings(parent);
+	pci_configure_acs_sv(bridge, true);
 	pci_bus_add_devices(parent);
 
  out:
@@ -117,6 +118,16 @@  void pciehp_unconfigure_device(struct controller *ctrl, bool presence)
 		}
 		pci_dev_put(dev);
 	}
-
+	/*
+	 * PCIe 6.0, 6.12.1.1 specifies that downstream devices are permitted
+	 * to send upstream messages before they have been assigned a bus
+	 * number and such messages have a Requester ID with Bus number
+	 * set to 00h. If the Downstrem port has ACS Source Validation enabled,
+	 * these messages will be detected as ACS violation error.
+	 * Hence, disable ACS Source Validation here and re-enable it after
+	 * enumeration of the downstream hierarchy and before binding the
+	 * respective device drivers in pciehp_configure_device().
+	 */
+	pci_configure_acs_sv(ctrl->pcie->port, false);
 	pci_unlock_rescan_remove();
 }
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index 95bc329e74c0..9cefaf814f49 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -991,6 +991,28 @@  static void pci_enable_acs(struct pci_dev *dev)
 	pci_disable_acs_redir(dev);
 }
 
+#ifdef CONFIG_HOTPLUG_PCI_PCIE
+void pci_configure_acs_sv(struct pci_dev *dev, bool flag)
+{
+	u16 cap;
+	u16 ctrl;
+
+	if (!pci_acs_enable || !dev->acs_cap)
+		return;
+
+	pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CAP, &cap);
+	pci_read_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, &ctrl);
+
+	if (flag)
+		ctrl |= (cap & PCI_ACS_SV);
+	else
+		ctrl &= ~(cap & PCI_ACS_SV);
+
+	pci_write_config_word(dev, dev->acs_cap + PCI_ACS_CTRL, ctrl);
+}
+EXPORT_SYMBOL_GPL(pci_configure_acs_sv);
+#endif
+
 /**
  * pci_restore_bars - restore a device's BAR values (e.g. after wake-up)
  * @dev: PCI device to have its BARs restored
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 060af91bafcd..edf516e39764 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -2239,6 +2239,12 @@  void pci_hp_create_module_link(struct pci_slot *pci_slot);
 void pci_hp_remove_module_link(struct pci_slot *pci_slot);
 #endif
 
+#ifdef CONFIG_HOTPLUG_PCI_PCIE
+void pci_configure_acs_sv(struct pci_dev *dev, bool flag);
+#else
+static inline void pci_configure_acs_sv(struct pci_dev *dev, bool flag) { }
+#endif
+
 /**
  * pci_pcie_cap - get the saved PCIe capability offset
  * @dev: PCI device