Re: [PATCH v2] PCI: Add D3 support for PCI bridges in DT based platforms

[Manivannan Sadhasivam]

On Wed, Feb 14, 2024 at 11:28:37AM +0100, Lukas Wunner wrote:
> On Wed, Feb 14, 2024 at 02:18:31PM +0530, Manivannan Sadhasivam wrote:
> > +/**
> > + * of_pci_bridge_d3 - Check if the bridge is supporting D3 states or not
> > + *
> > + * @node: device tree node of the bridge
> > + *
> > + * Return: True if the bridge is supporting D3 states, False otherwise.
> 
> A lot of kernel-doc uses %true and %false.
> 

Ack.

> 
> > +bool of_pci_bridge_d3(struct device_node *node)
> > +{
> > +	return of_property_read_bool(node, "supports-d3");
> > +}
> 
> What's the difference between of_property_read_bool() and
> of_property_present()?  When should one use which?
> The former has 691 occurrences in the tree, the latter 120.
> The latter would seem more "literary" / readable here,
> but maybe that's just me.
> 

of_property_present() just calls of_property_read_bool() and it is fairly new.
But yeah, the API name itself indicates that it is better suited for the
purpose. Will change it.

> 
> > --- a/drivers/pci/pci.c
> > +++ b/drivers/pci/pci.c
> > @@ -1142,6 +1142,9 @@ static inline bool platform_pci_bridge_d3(struct pci_dev *dev)
> >  	if (pci_use_mid_pm())
> >  		return false;
> >  
> > +	if (dev->dev.of_node)
> > +		return of_pci_bridge_d3(dev->dev.of_node);
> > +
> >  	return acpi_pci_bridge_d3(dev);
> >  }
> 
> This will result in an unnecessary test on non-DT platforms (e.g. ACPI)
> whether dev->dev.of_node is set.
> 
> Please use dev_of_node() instead of "dev->dev.of_node" so that the
> code added here can be optimized away by the compiler on non-DT
> platforms (due to the IS_ENABLED(CONFIG_OF)).
> 

Sounds good.

- Mani

-- 
மணிவண்ணன் சதாசிவம்