[PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices

Nicolin Chen posted 3 patches 4 weeks, 1 day ago
[PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Nicolin Chen 4 weeks, 1 day ago
Some NVIDIA GPU/NIC devices, although don't implement the CXL config space,
they have many CXL-like properties. Call this kind "pre-CXL".

Similar to CXL.cache capaiblity, these pre-CXL devices also require the ATS
function even when their RIDs are IOMMU bypassed, i.e. keep ATS "always on"
v.s. "on demand" when a non-zero PASID line gets enabled in SVA use cases.

Introduce pci_dev_specific_ats_always_on() quirk function to scan a list of
IDs for these device. Then, include it pci_ats_always_on().

Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
---
 drivers/pci/pci.h    |  9 +++++++++
 drivers/pci/ats.c    |  3 ++-
 drivers/pci/quirks.c | 26 ++++++++++++++++++++++++++
 3 files changed, 37 insertions(+), 1 deletion(-)

diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
index 13d998fbacce6..13fa71f965900 100644
--- a/drivers/pci/pci.h
+++ b/drivers/pci/pci.h
@@ -1150,6 +1150,15 @@ static inline int pci_dev_specific_reset(struct pci_dev *dev, bool probe)
 }
 #endif
 
+#if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_PCI_ATS)
+bool pci_dev_specific_ats_always_on(struct pci_dev *dev);
+#else
+static inline bool pci_dev_specific_ats_always_on(struct pci_dev *dev)
+{
+	return false;
+}
+#endif
+
 #if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_ARM64)
 int acpi_get_rc_resources(struct device *dev, const char *hid, u16 segment,
 			  struct resource *res);
diff --git a/drivers/pci/ats.c b/drivers/pci/ats.c
index cf262eb6e6890..e6995b536ad4c 100644
--- a/drivers/pci/ats.c
+++ b/drivers/pci/ats.c
@@ -243,7 +243,8 @@ bool pci_ats_always_on(struct pci_dev *pdev)
 	if (pdev->is_virtfn)
 		pdev = pci_physfn(pdev);
 
-	return pci_cxl_ats_always_on(pdev);
+	return pci_cxl_ats_always_on(pdev) ||
+	       pci_dev_specific_ats_always_on(pdev);
 }
 EXPORT_SYMBOL_GPL(pci_ats_always_on);
 
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 48946cca4be72..21451e62f284e 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -5714,6 +5714,32 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x1457, quirk_intel_e2000_no_ats);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x1459, quirk_intel_e2000_no_ats);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145a, quirk_intel_e2000_no_ats);
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145c, quirk_intel_e2000_no_ats);
+
+static const struct pci_dev_ats_always_on {
+	u16 vendor;
+	u16 device;
+} pci_dev_ats_always_on[] = {
+	/* NVIDIA GPUs */
+	{ PCI_VENDOR_ID_NVIDIA, 0x2e12, },
+	{ PCI_VENDOR_ID_NVIDIA, 0x2e2a, },
+	{ PCI_VENDOR_ID_NVIDIA, 0x2e2b, },
+	/* NVIDIA CX10 Family NVlink-C2C */
+	{ PCI_VENDOR_ID_MELLANOX, 0x2101, },
+	{ 0 }
+};
+
+/* Some pre-CXL devices require ATS on the RID when it is IOMMU-bypassed */
+bool pci_dev_specific_ats_always_on(struct pci_dev *pdev)
+{
+	const struct pci_dev_ats_always_on *i;
+
+	for (i = pci_dev_ats_always_on; i->vendor; i++) {
+		if (i->vendor == pdev->vendor && i->device == pdev->device)
+			return true;
+	}
+
+	return false;
+}
 #endif /* CONFIG_PCI_ATS */
 
 /* Freescale PCIe doesn't support MSI in RC mode */
-- 
2.43.0
RE: [PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Tian, Kevin 5 days, 11 hours ago
> From: Nicolin Chen <nicolinc@nvidia.com>
> Sent: Saturday, March 7, 2026 7:41 AM
> 
> Some NVIDIA GPU/NIC devices, although don't implement the CXL config
> space,
> they have many CXL-like properties. Call this kind "pre-CXL".
> 
> Similar to CXL.cache capaiblity, these pre-CXL devices also require the ATS

s/capaiblity/capability/

> function even when their RIDs are IOMMU bypassed, i.e. keep ATS "always
> on"
> v.s. "on demand" when a non-zero PASID line gets enabled in SVA use cases.
> 
> Introduce pci_dev_specific_ats_always_on() quirk function to scan a list of
> IDs for these device. Then, include it pci_ats_always_on().

"include it *in* pci_ats_always_on()"

> +
> +/* Some pre-CXL devices require ATS on the RID when it is IOMMU-
> bypassed */
> +bool pci_dev_specific_ats_always_on(struct pci_dev *pdev)

clearer to remove "on the RID ...". 

"always on" implies no condition required. and adding IOMMU bypass
info there is confusing.

Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Re: [PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Jason Gunthorpe 6 days, 6 hours ago
On Fri, Mar 06, 2026 at 03:41:16PM -0800, Nicolin Chen wrote:
> Some NVIDIA GPU/NIC devices, although don't implement the CXL config space,
> they have many CXL-like properties. Call this kind "pre-CXL".
> 
> Similar to CXL.cache capaiblity, these pre-CXL devices also require the ATS
> function even when their RIDs are IOMMU bypassed, i.e. keep ATS "always on"
> v.s. "on demand" when a non-zero PASID line gets enabled in SVA use cases.
> 
> Introduce pci_dev_specific_ats_always_on() quirk function to scan a list of
> IDs for these device. Then, include it pci_ats_always_on().
> 
> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
> ---
>  drivers/pci/pci.h    |  9 +++++++++
>  drivers/pci/ats.c    |  3 ++-
>  drivers/pci/quirks.c | 26 ++++++++++++++++++++++++++
>  3 files changed, 37 insertions(+), 1 deletion(-)

Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>

Jason
Re: [PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Jonathan Cameron 3 weeks, 6 days ago
On Fri, 6 Mar 2026 15:41:16 -0800
Nicolin Chen <nicolinc@nvidia.com> wrote:

> Some NVIDIA GPU/NIC devices, although don't implement the CXL config space,
> they have many CXL-like properties. Call this kind "pre-CXL".
> 
> Similar to CXL.cache capaiblity, these pre-CXL devices also require the ATS
> function even when their RIDs are IOMMU bypassed, i.e. keep ATS "always on"
> v.s. "on demand" when a non-zero PASID line gets enabled in SVA use cases.
> 
> Introduce pci_dev_specific_ats_always_on() quirk function to scan a list of
> IDs for these device. Then, include it pci_ats_always_on().
> 
> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Seems like a reasonably tidy quirk.
Reviewed-by: Jonathan Cameron <jonathan.cameron@huawei.com>
Re: [PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Nirmoy Das 3 weeks, 6 days ago
On 07.03.26 00:41, Nicolin Chen wrote:
> Some NVIDIA GPU/NIC devices, although don't implement the CXL config space,
> they have many CXL-like properties. Call this kind "pre-CXL".
>
> Similar to CXL.cache capaiblity, these pre-CXL devices also require the ATS
> function even when their RIDs are IOMMU bypassed, i.e. keep ATS "always on"
> v.s. "on demand" when a non-zero PASID line gets enabled in SVA use cases.
>
> Introduce pci_dev_specific_ats_always_on() quirk function to scan a list of
> IDs for these device. Then, include it pci_ats_always_on().
>
> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>

Tested-by: Nirmoy Das <nirmoyd@nvidia.com>

Reviewed-by: Nirmoy Das <nirmoy@nvidia.com>

> ---
>   drivers/pci/pci.h    |  9 +++++++++
>   drivers/pci/ats.c    |  3 ++-
>   drivers/pci/quirks.c | 26 ++++++++++++++++++++++++++
>   3 files changed, 37 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
> index 13d998fbacce6..13fa71f965900 100644
> --- a/drivers/pci/pci.h
> +++ b/drivers/pci/pci.h
> @@ -1150,6 +1150,15 @@ static inline int pci_dev_specific_reset(struct pci_dev *dev, bool probe)
>   }
>   #endif
>   
> +#if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_PCI_ATS)
> +bool pci_dev_specific_ats_always_on(struct pci_dev *dev);
> +#else
> +static inline bool pci_dev_specific_ats_always_on(struct pci_dev *dev)
> +{
> +	return false;
> +}
> +#endif
> +
>   #if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_ARM64)
>   int acpi_get_rc_resources(struct device *dev, const char *hid, u16 segment,
>   			  struct resource *res);
> diff --git a/drivers/pci/ats.c b/drivers/pci/ats.c
> index cf262eb6e6890..e6995b536ad4c 100644
> --- a/drivers/pci/ats.c
> +++ b/drivers/pci/ats.c
> @@ -243,7 +243,8 @@ bool pci_ats_always_on(struct pci_dev *pdev)
>   	if (pdev->is_virtfn)
>   		pdev = pci_physfn(pdev);
>   
> -	return pci_cxl_ats_always_on(pdev);
> +	return pci_cxl_ats_always_on(pdev) ||
> +	       pci_dev_specific_ats_always_on(pdev);
>   }
>   EXPORT_SYMBOL_GPL(pci_ats_always_on);
>   
> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> index 48946cca4be72..21451e62f284e 100644
> --- a/drivers/pci/quirks.c
> +++ b/drivers/pci/quirks.c
> @@ -5714,6 +5714,32 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x1457, quirk_intel_e2000_no_ats);
>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x1459, quirk_intel_e2000_no_ats);
>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145a, quirk_intel_e2000_no_ats);
>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145c, quirk_intel_e2000_no_ats);
> +
> +static const struct pci_dev_ats_always_on {
> +	u16 vendor;
> +	u16 device;
> +} pci_dev_ats_always_on[] = {
> +	/* NVIDIA GPUs */
> +	{ PCI_VENDOR_ID_NVIDIA, 0x2e12, },
> +	{ PCI_VENDOR_ID_NVIDIA, 0x2e2a, },
> +	{ PCI_VENDOR_ID_NVIDIA, 0x2e2b, },
> +	/* NVIDIA CX10 Family NVlink-C2C */
> +	{ PCI_VENDOR_ID_MELLANOX, 0x2101, },
> +	{ 0 }
> +};
> +
> +/* Some pre-CXL devices require ATS on the RID when it is IOMMU-bypassed */
> +bool pci_dev_specific_ats_always_on(struct pci_dev *pdev)
> +{
> +	const struct pci_dev_ats_always_on *i;
> +
> +	for (i = pci_dev_ats_always_on; i->vendor; i++) {
> +		if (i->vendor == pdev->vendor && i->device == pdev->device)
> +			return true;
> +	}
> +
> +	return false;
> +}
>   #endif /* CONFIG_PCI_ATS */
>   
>   /* Freescale PCIe doesn't support MSI in RC mode */
Re: [PATCH v3 2/3] PCI: Allow ATS to be always on for pre-CXL devices
Posted by Nirmoy Das 3 weeks, 6 days ago
On 08.03.26 21:50, Nirmoy Das wrote:
>
> On 07.03.26 00:41, Nicolin Chen wrote:
>> Some NVIDIA GPU/NIC devices, although don't implement the CXL config 
>> space,
>> they have many CXL-like properties. Call this kind "pre-CXL".
>>
>> Similar to CXL.cache capaiblity, these pre-CXL devices also require 
>> the ATS
>> function even when their RIDs are IOMMU bypassed, i.e. keep ATS 
>> "always on"
>> v.s. "on demand" when a non-zero PASID line gets enabled in SVA use 
>> cases.
>>
>> Introduce pci_dev_specific_ats_always_on() quirk function to scan a 
>> list of
>> IDs for these device. Then, include it pci_ats_always_on().
>>
>> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
>> Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
>
> Tested-by: Nirmoy Das <nirmoyd@nvidia.com>
>
> Reviewed-by: Nirmoy Das <nirmoy@nvidia.com>


Sent with wrong email address

Reviewed-by: Nirmoy Das <nirmoyd@nvidia.com>
>
>> ---
>>   drivers/pci/pci.h    |  9 +++++++++
>>   drivers/pci/ats.c    |  3 ++-
>>   drivers/pci/quirks.c | 26 ++++++++++++++++++++++++++
>>   3 files changed, 37 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
>> index 13d998fbacce6..13fa71f965900 100644
>> --- a/drivers/pci/pci.h
>> +++ b/drivers/pci/pci.h
>> @@ -1150,6 +1150,15 @@ static inline int 
>> pci_dev_specific_reset(struct pci_dev *dev, bool probe)
>>   }
>>   #endif
>>   +#if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_PCI_ATS)
>> +bool pci_dev_specific_ats_always_on(struct pci_dev *dev);
>> +#else
>> +static inline bool pci_dev_specific_ats_always_on(struct pci_dev *dev)
>> +{
>> +    return false;
>> +}
>> +#endif
>> +
>>   #if defined(CONFIG_PCI_QUIRKS) && defined(CONFIG_ARM64)
>>   int acpi_get_rc_resources(struct device *dev, const char *hid, u16 
>> segment,
>>                 struct resource *res);
>> diff --git a/drivers/pci/ats.c b/drivers/pci/ats.c
>> index cf262eb6e6890..e6995b536ad4c 100644
>> --- a/drivers/pci/ats.c
>> +++ b/drivers/pci/ats.c
>> @@ -243,7 +243,8 @@ bool pci_ats_always_on(struct pci_dev *pdev)
>>       if (pdev->is_virtfn)
>>           pdev = pci_physfn(pdev);
>>   -    return pci_cxl_ats_always_on(pdev);
>> +    return pci_cxl_ats_always_on(pdev) ||
>> +           pci_dev_specific_ats_always_on(pdev);
>>   }
>>   EXPORT_SYMBOL_GPL(pci_ats_always_on);
>>   diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
>> index 48946cca4be72..21451e62f284e 100644
>> --- a/drivers/pci/quirks.c
>> +++ b/drivers/pci/quirks.c
>> @@ -5714,6 +5714,32 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 
>> 0x1457, quirk_intel_e2000_no_ats);
>>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x1459, 
>> quirk_intel_e2000_no_ats);
>>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145a, 
>> quirk_intel_e2000_no_ats);
>>   DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x145c, 
>> quirk_intel_e2000_no_ats);
>> +
>> +static const struct pci_dev_ats_always_on {
>> +    u16 vendor;
>> +    u16 device;
>> +} pci_dev_ats_always_on[] = {
>> +    /* NVIDIA GPUs */
>> +    { PCI_VENDOR_ID_NVIDIA, 0x2e12, },
>> +    { PCI_VENDOR_ID_NVIDIA, 0x2e2a, },
>> +    { PCI_VENDOR_ID_NVIDIA, 0x2e2b, },
>> +    /* NVIDIA CX10 Family NVlink-C2C */
>> +    { PCI_VENDOR_ID_MELLANOX, 0x2101, },
>> +    { 0 }
>> +};
>> +
>> +/* Some pre-CXL devices require ATS on the RID when it is 
>> IOMMU-bypassed */
>> +bool pci_dev_specific_ats_always_on(struct pci_dev *pdev)
>> +{
>> +    const struct pci_dev_ats_always_on *i;
>> +
>> +    for (i = pci_dev_ats_always_on; i->vendor; i++) {
>> +        if (i->vendor == pdev->vendor && i->device == pdev->device)
>> +            return true;
>> +    }
>> +
>> +    return false;
>> +}
>>   #endif /* CONFIG_PCI_ATS */
>>     /* Freescale PCIe doesn't support MSI in RC mode */
>