diff mbox series

[v2] PCI/DPC: Check host->native_dpc before enable dpc service

Message ID 1612356795-32505-1-git-send-email-yangyicong@hisilicon.com
State New
Headers show
Series [v2] PCI/DPC: Check host->native_dpc before enable dpc service | expand

Commit Message

Yicong Yang Feb. 3, 2021, 12:53 p.m. UTC
Per Downstream Port Containment Related Enhancements ECN[1]
Table 4-6, Interpretation of _OSC Control Field Returned Value,
for bit 7 of _OSC control return value:

  "Firmware sets this bit to 1 to grant the OS control over PCI Express
  Downstream Port Containment configuration."
  "If control of this feature was requested and denied,
  or was not requested, the firmware returns this bit set to 0."

We store bit 7 of _OSC control return value in host->native_dpc,
check it before enable the dpc service as the firmware may not
grant the control.

[1] Downstream Port Containment Related Enhancements ECN,
    Jan 28, 2019, affecting PCI Firmware Specification, Rev. 3.2
    https://members.pcisig.com/wg/PCI-SIG/document/12888

Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
---
Change since v1:
- use correct reference for _OSC control return value

 drivers/pci/pcie/portdrv_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

Yicong Yang Feb. 24, 2021, 9:47 a.m. UTC | #1
On 2021/2/3 20:53, Yicong Yang wrote:
> Per Downstream Port Containment Related Enhancements ECN[1]
> Table 4-6, Interpretation of _OSC Control Field Returned Value,
> for bit 7 of _OSC control return value:
> 
>   "Firmware sets this bit to 1 to grant the OS control over PCI Express
>   Downstream Port Containment configuration."
>   "If control of this feature was requested and denied,
>   or was not requested, the firmware returns this bit set to 0."
> 
> We store bit 7 of _OSC control return value in host->native_dpc,
> check it before enable the dpc service as the firmware may not
> grant the control.
> 
> [1] Downstream Port Containment Related Enhancements ECN,
>     Jan 28, 2019, affecting PCI Firmware Specification, Rev. 3.2
>     https://members.pcisig.com/wg/PCI-SIG/document/12888
> 
> Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
> ---
> Change since v1:
> - use correct reference for _OSC control return value
> 
>  drivers/pci/pcie/portdrv_core.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
> index e1fed664..7445d03 100644
> --- a/drivers/pci/pcie/portdrv_core.c
> +++ b/drivers/pci/pcie/portdrv_core.c
> @@ -253,7 +253,8 @@ static int get_port_device_capability(struct pci_dev *dev)
>  	 */
>  	if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
>  	    pci_aer_available() &&
> -	    (pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
> +	    (pcie_ports_dpc_native ||
> +	    ((services & PCIE_PORT_SERVICE_AER) && host->native_dpc)))
>  		services |= PCIE_PORT_SERVICE_DPC;
>  

the check here maybe problematic. the bit 7 of _OSC return value is reserved
previously and the change here may break the backward compatibility.
currently we make dpc enabled along with aer, which can ensure the native
dpc won't be enabled if the edr is enabled.

i feel a bit confused as the bit 7 is not used.
does it provide a way to enable native dpc regardless of aer ownership?
just as pcie_ports=dpc-native does when i checked the discussion in [1].

[1] https://lore.kernel.org/linux-pci/20191023192205.97024-1-olof@lixom.net/

Thanks,
Yicong

>  	if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM ||
>
Bjorn Helgaas July 26, 2021, 10:05 p.m. UTC | #2
On Wed, Feb 24, 2021 at 05:47:58PM +0800, Yicong Yang wrote:
> On 2021/2/3 20:53, Yicong Yang wrote:
> > Per Downstream Port Containment Related Enhancements ECN[1]
> > Table 4-6, Interpretation of _OSC Control Field Returned Value,
> > for bit 7 of _OSC control return value:
> > 
> >   "Firmware sets this bit to 1 to grant the OS control over PCI Express
> >   Downstream Port Containment configuration."
> >   "If control of this feature was requested and denied,
> >   or was not requested, the firmware returns this bit set to 0."
> > 
> > We store bit 7 of _OSC control return value in host->native_dpc,
> > check it before enable the dpc service as the firmware may not
> > grant the control.
> > 
> > [1] Downstream Port Containment Related Enhancements ECN,
> >     Jan 28, 2019, affecting PCI Firmware Specification, Rev. 3.2
> >     https://members.pcisig.com/wg/PCI-SIG/document/12888
> > 
> > Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
> > ---
> > Change since v1:
> > - use correct reference for _OSC control return value
> > 
> >  drivers/pci/pcie/portdrv_core.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
> > index e1fed664..7445d03 100644
> > --- a/drivers/pci/pcie/portdrv_core.c
> > +++ b/drivers/pci/pcie/portdrv_core.c
> > @@ -253,7 +253,8 @@ static int get_port_device_capability(struct pci_dev *dev)
> >  	 */
> >  	if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
> >  	    pci_aer_available() &&
> > -	    (pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
> > +	    (pcie_ports_dpc_native ||
> > +	    ((services & PCIE_PORT_SERVICE_AER) && host->native_dpc)))
> >  		services |= PCIE_PORT_SERVICE_DPC;
> >  
> 
> the check here maybe problematic. the bit 7 of _OSC return value is reserved
> previously and the change here may break the backward compatibility.
> currently we make dpc enabled along with aer, which can ensure the native
> dpc won't be enabled if the edr is enabled.

Since you think this is problematic, I'll drop it for now.  If you
decide it's the right thing to do, please post it again.

> i feel a bit confused as the bit 7 is not used.
> does it provide a way to enable native dpc regardless of aer ownership?
> just as pcie_ports=dpc-native does when i checked the discussion in [1].
> 
> [1] https://lore.kernel.org/linux-pci/20191023192205.97024-1-olof@lixom.net/
> 
> Thanks,
> Yicong
> 
> >  	if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM ||
> > 
>
Yicong Yang July 28, 2021, 9:22 a.m. UTC | #3
On 2021/7/27 6:05, Bjorn Helgaas wrote:
> On Wed, Feb 24, 2021 at 05:47:58PM +0800, Yicong Yang wrote:
>> On 2021/2/3 20:53, Yicong Yang wrote:
>>> Per Downstream Port Containment Related Enhancements ECN[1]
>>> Table 4-6, Interpretation of _OSC Control Field Returned Value,
>>> for bit 7 of _OSC control return value:
>>>
>>>   "Firmware sets this bit to 1 to grant the OS control over PCI Express
>>>   Downstream Port Containment configuration."
>>>   "If control of this feature was requested and denied,
>>>   or was not requested, the firmware returns this bit set to 0."
>>>
>>> We store bit 7 of _OSC control return value in host->native_dpc,
>>> check it before enable the dpc service as the firmware may not
>>> grant the control.
>>>
>>> [1] Downstream Port Containment Related Enhancements ECN,
>>>     Jan 28, 2019, affecting PCI Firmware Specification, Rev. 3.2
>>>     https://members.pcisig.com/wg/PCI-SIG/document/12888
>>>
>>> Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
>>> ---
>>> Change since v1:
>>> - use correct reference for _OSC control return value
>>>
>>>  drivers/pci/pcie/portdrv_core.c | 3 ++-
>>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
>>> index e1fed664..7445d03 100644
>>> --- a/drivers/pci/pcie/portdrv_core.c
>>> +++ b/drivers/pci/pcie/portdrv_core.c
>>> @@ -253,7 +253,8 @@ static int get_port_device_capability(struct pci_dev *dev)
>>>  	 */
>>>  	if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
>>>  	    pci_aer_available() &&
>>> -	    (pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
>>> +	    (pcie_ports_dpc_native ||
>>> +	    ((services & PCIE_PORT_SERVICE_AER) && host->native_dpc)))
>>>  		services |= PCIE_PORT_SERVICE_DPC;
>>>  
>>
>> the check here maybe problematic. the bit 7 of _OSC return value is reserved
>> previously and the change here may break the backward compatibility.
>> currently we make dpc enabled along with aer, which can ensure the native
>> dpc won't be enabled if the edr is enabled.
> 
> Since you think this is problematic, I'll drop it for now.  If you
> decide it's the right thing to do, please post it again.
> 

yes, I still think this has problem. sorry for the noise.

>> i feel a bit confused as the bit 7 is not used.
>> does it provide a way to enable native dpc regardless of aer ownership?
>> just as pcie_ports=dpc-native does when i checked the discussion in [1].
>>
>> [1] https://lore.kernel.org/linux-pci/20191023192205.97024-1-olof@lixom.net/
>>
>> Thanks,
>> Yicong
>>
>>>  	if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM ||
>>>
>>
> 
> .
>
diff mbox series

Patch

diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
index e1fed664..7445d03 100644
--- a/drivers/pci/pcie/portdrv_core.c
+++ b/drivers/pci/pcie/portdrv_core.c
@@ -253,7 +253,8 @@  static int get_port_device_capability(struct pci_dev *dev)
 	 */
 	if (pci_find_ext_capability(dev, PCI_EXT_CAP_ID_DPC) &&
 	    pci_aer_available() &&
-	    (pcie_ports_dpc_native || (services & PCIE_PORT_SERVICE_AER)))
+	    (pcie_ports_dpc_native ||
+	    ((services & PCIE_PORT_SERVICE_AER) && host->native_dpc)))
 		services |= PCIE_PORT_SERVICE_DPC;
 
 	if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM ||