1. Patchew
  2. linux
  3. [PATCH v5 0/16] Enable CXL PCIe port protocol error handling and logging
  4. View patch

[PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices

Terry Bowman posted 16 patches 11 months, 2 weeks ago
There is a newer version of this series
[PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Terry Bowman 11 months, 2 weeks ago 
The AER service driver's aer_get_device_error_info() function doesn't read
uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
including CXL Upstream Switch Ports. As a result, fatal errors are not
logged or handled as needed for CXL PCIe Upstream Switch Port devices.

Update the aer_get_device_error_info() function to read the UCE fatal
status for all CXL PCIe devices. Make the change such that non-CXL devices
are not affected.

The fatal error status will be used in future patches implementing
CXL PCIe Port uncorrectable error handling and logging.

Signed-off-by: Terry Bowman <terry.bowman@amd.com>
---
 drivers/pci/pcie/aer.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
index 62be599e3bee..79c828bdcb6d 100644
--- a/drivers/pci/pcie/aer.c
+++ b/drivers/pci/pcie/aer.c
@@ -1253,7 +1253,8 @@ int aer_get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
 	} else if (type == PCI_EXP_TYPE_ROOT_PORT ||
 		   type == PCI_EXP_TYPE_RC_EC ||
 		   type == PCI_EXP_TYPE_DOWNSTREAM ||
-		   info->severity == AER_NONFATAL) {
+		   info->severity == AER_NONFATAL ||
+		   (pcie_is_cxl(dev) && type == PCI_EXP_TYPE_UPSTREAM)) {
 
 		/* Link is still healthy for IO reads */
 		pci_read_config_dword(dev, aer + PCI_ERR_UNCOR_STATUS,
-- 
2.34.1
Re: [PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Ira Weiny 11 months, 1 week ago 
Terry Bowman wrote:
> The AER service driver's aer_get_device_error_info() function doesn't read
> uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
> including CXL Upstream Switch Ports. As a result, fatal errors are not
> logged or handled as needed for CXL PCIe Upstream Switch Port devices.
> 
> Update the aer_get_device_error_info() function to read the UCE fatal
> status for all CXL PCIe devices. Make the change such that non-CXL devices
> are not affected.
> 
> The fatal error status will be used in future patches implementing
> CXL PCIe Port uncorrectable error handling and logging.
> 
> Signed-off-by: Terry Bowman <terry.bowman@amd.com>

Reviewed-by: Ira Weiny <ira.weiny@intel.com>

[snip]
Re: [PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Jonathan Cameron 11 months, 1 week ago 
On Tue, 7 Jan 2025 08:38:42 -0600
Terry Bowman <terry.bowman@amd.com> wrote:

> The AER service driver's aer_get_device_error_info() function doesn't read
> uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
> including CXL Upstream Switch Ports. As a result, fatal errors are not
> logged or handled as needed for CXL PCIe Upstream Switch Port devices.
> 
> Update the aer_get_device_error_info() function to read the UCE fatal
> status for all CXL PCIe devices. Make the change such that non-CXL devices
> are not affected.
> 
> The fatal error status will be used in future patches implementing
> CXL PCIe Port uncorrectable error handling and logging.
> 
> Signed-off-by: Terry Bowman <terry.bowman@amd.com>

This clashes with Shuai's series adding link healthy checks.
Maybe we can reuse that logic to incorporate the condition we
care about here?


> ---
>  drivers/pci/pcie/aer.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
> index 62be599e3bee..79c828bdcb6d 100644
> --- a/drivers/pci/pcie/aer.c
> +++ b/drivers/pci/pcie/aer.c
> @@ -1253,7 +1253,8 @@ int aer_get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
>  	} else if (type == PCI_EXP_TYPE_ROOT_PORT ||
>  		   type == PCI_EXP_TYPE_RC_EC ||
>  		   type == PCI_EXP_TYPE_DOWNSTREAM ||
> -		   info->severity == AER_NONFATAL) {
> +		   info->severity == AER_NONFATAL ||
> +		   (pcie_is_cxl(dev) && type == PCI_EXP_TYPE_UPSTREAM)) {
>  
>  		/* Link is still healthy for IO reads */
>  		pci_read_config_dword(dev, aer + PCI_ERR_UNCOR_STATUS,
Re: [PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Bowman, Terry 10 months, 3 weeks ago 


On 1/14/2025 5:32 AM, Jonathan Cameron wrote:
> On Tue, 7 Jan 2025 08:38:42 -0600
> Terry Bowman <terry.bowman@amd.com> wrote:
>
>> The AER service driver's aer_get_device_error_info() function doesn't read
>> uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
>> including CXL Upstream Switch Ports. As a result, fatal errors are not
>> logged or handled as needed for CXL PCIe Upstream Switch Port devices.
>>
>> Update the aer_get_device_error_info() function to read the UCE fatal
>> status for all CXL PCIe devices. Make the change such that non-CXL devices
>> are not affected.
>>
>> The fatal error status will be used in future patches implementing
>> CXL PCIe Port uncorrectable error handling and logging.
>>
>> Signed-off-by: Terry Bowman <terry.bowman@amd.com>
> This clashes with Shuai's series adding link healthy checks.
> Maybe we can reuse that logic to incorporate the condition we
> care about here?
>

Hi Jonathan, et. al,

After looking at this closer and considering the situation I believe
we should remove this patch from the patchset and defer adding these
changes to log USP AER and RAS UCE.

I propose we reintroduce this later as a RFC or RFT in a future patchset.
This will give more needed time for testing.

The only downside to adding later is in the case of CXL USP fatal UCE. AER and
RAS will not be logged but this was the AER driver's existing behavior and as a
result isn't a regression.

Your thoughts?

Regards,
Terry

>> ---
>>  drivers/pci/pcie/aer.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
>> index 62be599e3bee..79c828bdcb6d 100644
>> --- a/drivers/pci/pcie/aer.c
>> +++ b/drivers/pci/pcie/aer.c
>> @@ -1253,7 +1253,8 @@ int aer_get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
>>  	} else if (type == PCI_EXP_TYPE_ROOT_PORT ||
>>  		   type == PCI_EXP_TYPE_RC_EC ||
>>  		   type == PCI_EXP_TYPE_DOWNSTREAM ||
>> -		   info->severity == AER_NONFATAL) {
>> +		   info->severity == AER_NONFATAL ||
>> +		   (pcie_is_cxl(dev) && type == PCI_EXP_TYPE_UPSTREAM)) {
>>  
>>  		/* Link is still healthy for IO reads */
>>  		pci_read_config_dword(dev, aer + PCI_ERR_UNCOR_STATUS,
Re: [PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Jonathan Cameron 10 months, 3 weeks ago 
On Tue, 28 Jan 2025 14:25:54 -0600
"Bowman, Terry" <terry.bowman@amd.com> wrote:

> On 1/14/2025 5:32 AM, Jonathan Cameron wrote:
> > On Tue, 7 Jan 2025 08:38:42 -0600
> > Terry Bowman <terry.bowman@amd.com> wrote:
> >  
> >> The AER service driver's aer_get_device_error_info() function doesn't read
> >> uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
> >> including CXL Upstream Switch Ports. As a result, fatal errors are not
> >> logged or handled as needed for CXL PCIe Upstream Switch Port devices.
> >>
> >> Update the aer_get_device_error_info() function to read the UCE fatal
> >> status for all CXL PCIe devices. Make the change such that non-CXL devices
> >> are not affected.
> >>
> >> The fatal error status will be used in future patches implementing
> >> CXL PCIe Port uncorrectable error handling and logging.
> >>
> >> Signed-off-by: Terry Bowman <terry.bowman@amd.com>  
> > This clashes with Shuai's series adding link healthy checks.
> > Maybe we can reuse that logic to incorporate the condition we
> > care about here?
> >  
> 
> Hi Jonathan, et. al,
> 
> After looking at this closer and considering the situation I believe
> we should remove this patch from the patchset and defer adding these
> changes to log USP AER and RAS UCE.
> 
> I propose we reintroduce this later as a RFC or RFT in a future patchset.
> This will give more needed time for testing.
> 
> The only downside to adding later is in the case of CXL USP fatal UCE. AER and
> RAS will not be logged but this was the AER driver's existing behavior and as a
> result isn't a regression.

If we have doubts and it is complex then sure. Let's do this in stages.

Jonathan

> 
> Your thoughts?
> 
> Regards,
> Terry
> 
> >> ---
> >>  drivers/pci/pcie/aer.c | 3 ++-
> >>  1 file changed, 2 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
> >> index 62be599e3bee..79c828bdcb6d 100644
> >> --- a/drivers/pci/pcie/aer.c
> >> +++ b/drivers/pci/pcie/aer.c
> >> @@ -1253,7 +1253,8 @@ int aer_get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
> >>  	} else if (type == PCI_EXP_TYPE_ROOT_PORT ||
> >>  		   type == PCI_EXP_TYPE_RC_EC ||
> >>  		   type == PCI_EXP_TYPE_DOWNSTREAM ||
> >> -		   info->severity == AER_NONFATAL) {
> >> +		   info->severity == AER_NONFATAL ||
> >> +		   (pcie_is_cxl(dev) && type == PCI_EXP_TYPE_UPSTREAM)) {
> >>  
> >>  		/* Link is still healthy for IO reads */
> >>  		pci_read_config_dword(dev, aer + PCI_ERR_UNCOR_STATUS,  
> 
>
Re: [PATCH v5 06/16] PCI/AER: Change AER driver to read UCE fatal status for all CXL PCIe Port devices
Posted by Bowman, Terry 11 months, 1 week ago 


On 1/14/2025 5:32 AM, Jonathan Cameron wrote:
> On Tue, 7 Jan 2025 08:38:42 -0600
> Terry Bowman <terry.bowman@amd.com> wrote:
>
>> The AER service driver's aer_get_device_error_info() function doesn't read
>> uncorrectable (UCE) fatal error status from PCIe Upstream Port devices,
>> including CXL Upstream Switch Ports. As a result, fatal errors are not
>> logged or handled as needed for CXL PCIe Upstream Switch Port devices.
>>
>> Update the aer_get_device_error_info() function to read the UCE fatal
>> status for all CXL PCIe devices. Make the change such that non-CXL devices
>> are not affected.
>>
>> The fatal error status will be used in future patches implementing
>> CXL PCIe Port uncorrectable error handling and logging.
>>
>> Signed-off-by: Terry Bowman <terry.bowman@amd.com>
> This clashes with Shuai's series adding link healthy checks.
> Maybe we can reuse that logic to incorporate the condition we
> care about here?
>

I'll add changes to query Upstream Port link status. I'll borrow from Shuai's patch.

Regards,
Terry

>> ---
>>  drivers/pci/pcie/aer.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/pci/pcie/aer.c b/drivers/pci/pcie/aer.c
>> index 62be599e3bee..79c828bdcb6d 100644
>> --- a/drivers/pci/pcie/aer.c
>> +++ b/drivers/pci/pcie/aer.c
>> @@ -1253,7 +1253,8 @@ int aer_get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
>>  	} else if (type == PCI_EXP_TYPE_ROOT_PORT ||
>>  		   type == PCI_EXP_TYPE_RC_EC ||
>>  		   type == PCI_EXP_TYPE_DOWNSTREAM ||
>> -		   info->severity == AER_NONFATAL) {
>> +		   info->severity == AER_NONFATAL ||
>> +		   (pcie_is_cxl(dev) && type == PCI_EXP_TYPE_UPSTREAM)) {
>>  
>>  		/* Link is still healthy for IO reads */
>>  		pci_read_config_dword(dev, aer + PCI_ERR_UNCOR_STATUS,

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.