Re: [PATCH v7 04/16] dt-bindings: imx6q-pcie: Add linux,pci-domain as required for iMX8MQ

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sat, Jan 06, 2024 at 11:47:36PM -0500, Frank Li wrote:
> On Sun, Jan 07, 2024 at 08:45:06AM +0530, Manivannan Sadhasivam wrote:
> > On Wed, Dec 27, 2023 at 01:27:15PM -0500, Frank Li wrote:
> > > iMX8MQ have two pci controllers. Adds "linux,pci-domain" as required
> > > proptery for iMX8MQ to indicate pci controller index.
> > > 
> > 
> > property
> > 
> > > This adjustment paves the way for eliminating the hardcoded check on the
> > > base register for acquiring the controller_id.
> > > 
> > > 	...
> > > 	if (dbi_base->start == IMX8MQ_PCIE2_BASE_ADDR)
> > > 		imx6_pcie->controller_id = 1;
> > > 	...
> > > 
> > > The controller_id is crucial and utilized for certain register bit
> > > positions. It must align precisely with the controller index in the SoC.
> > > An auto-incremented ID don't fit this case. The DTS or fuse configurations
> > > may deactivate specific PCI controllers.
> > > 
> > 
> > You cannot change the binding for the sake of driver. But you can make this
> > change in other way. See below...
> > 
> > > Signed-off-by: Frank Li <Frank.Li@xxxxxxx>
> > > ---
> > > 
> > > Notes:
> > >     Change from v5 to v6
> > >     - rework commit message to explain why need required and why auto increase
> > >     id not work
> > >     
> > >     Change from v4 to v5
> > >     - new patch at v5
> > > 
> > >  .../bindings/pci/fsl,imx6q-pcie-common.yaml           | 11 +++++++++++
> > >  1 file changed, 11 insertions(+)
> > > 
> > > diff --git a/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie-common.yaml b/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie-common.yaml
> > > index d91b639ae7ae7..8f39b4e6e8491 100644
> > > --- a/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie-common.yaml
> > > +++ b/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie-common.yaml
> > > @@ -265,6 +265,17 @@ allOf:
> > >              - const: apps
> > >              - const: turnoff
> > >  
> > > +  - if:
> > > +      properties:
> > > +        compatible:
> > > +          contains:
> > > +            enum:
> > > +              - fsl,imx8mq-pcie
> > > +              - fsl,imx8mq-pcie-ep
> > 
> > "linux,pci-domain" is a generic property. So you cannot make it required only
> > for certain SoCs. 
> 
> Sorry, why not? there are many generic property.
> 

It doesn't make sense to make it required only for specific SoCs since it is not
specific to any SoC. You can make it required for all.

> > But you can make it so for all SoCs. This way, the drivers
> > can also rely on it.
> > 
> > Now, you should get rid of the commit message about driver internals:
> 
> Not all dts already added "linux,pci-domain" yet. If required for all SOCs,
> it will cause dtb check warnings.
> 

You can safely add this property to all DTS. Nothing will break.

- Mani

> Frank
> > 
> > > This adjustment paves the way for eliminating the hardcoded check on the
> > > base register for acquiring the controller_id.
> > > 
> > >       ...
> > >       if (dbi_base->start == IMX8MQ_PCIE2_BASE_ADDR)
> > >               imx6_pcie->controller_id = 1;
> > >       ...
> > > 
> > > The controller_id is crucial and utilized for certain register bit
> > > positions. It must align precisely with the controller index in the SoC.
> > > An auto-incremented ID don't fit this case. The DTS or fuse configurations
> > > may deactivate specific PCI controllers.
> > > 
> > 
> > - Mani
> > 
> > > +    then:
> > > +      required:
> > > +        - linux,pci-domain
> > > +
> > >  additionalProperties: true
> > >  
> > >  ...
> > > -- 
> > > 2.34.1
> > > 
> > 
> > -- 
> > மணிவண்ணன் சதாசிவம்

-- 
மணிவண்ணன் சதாசிவம்




[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]


  Powered by Linux