Re: [PATCH] PCI: cadence: Kconfig: change PCIE_CADENCE configs from tristate to bool

[Siddharth Vadapalli]

On Thu, 2025-11-13 at 11:13 +0100, Arnd Bergmann wrote:

Hello Arnd,

> On Thu, Nov 13, 2025, at 10:27, Siddharth Vadapalli wrote:
> > The drivers associated with the PCIE_CADENCE, PCIE_CADENCE_HOST AND
> > PCIE_CADENCE_EP configs are used by multiple vendor drivers and serve as a
> > library of helpers. Since the vendor drivers could individually be built
> > as built-in or as loadable modules, it is possible to select a build
> > configuration wherein a vendor driver is built-in while the library is
> > built as a loadable module. This will result in a build error as reported
> > in the 'Closes' link below.
> > 
> > Address the build error by changing the library configs to be 'bool'
> > instead of 'tristate'.
> > 
> > Reported-by: kernel test robot <lkp@xxxxxxxxx>
> > Closes: 
> > https://lore.kernel.org/oe-kbuild-all/202511111705.MZ7ls8Hm-lkp@xxxxxxxxx/
> > Fixes: 1c72774df028 ("PCI: sg2042: Add Sophgo SG2042 PCIe driver")
> > Cc: <stable@xxxxxxxxxxxxxxx>
> > Signed-off-by: Siddharth Vadapalli <s-vadapalli@xxxxxx>
> 
> I really think there has to be a better solution here, this is not
> an unusual problem.
> 
> > @@ -4,16 +4,16 @@ menu "Cadence-based PCIe controllers"
> >  	depends on PCI
> > 
> >  config PCIE_CADENCE
> > -	tristate
> > +	bool
> > 
> >  config PCIE_CADENCE_HOST
> > -	tristate
> > +	bool
> >  	depends on OF
> >  	select IRQ_DOMAIN
> >  	select PCIE_CADENCE
> > 
> >  config PCIE_CADENCE_EP
> > -	tristate
> > +	bool
> >  	depends on OF
> >  	depends on PCI_ENDPOINT
> >  	select PCIE_CADENCE
> 
> I think the easiest way would be to leave PCIE_CADENCE as
> a 'tristate' symbol but make the other two 'bool', and then
> adjust the Makefile logic to use CONFIG_PCIE_CADENCE as
> the thing that controls how the individual drivers are built.
> 
> That way, if any platform specific driver is built-in, both
> the EP and HOST support are built-in or disabled but never
> loadable modules. As long as all platform drivers are
> loadable modules, so would be the base support.

Thank you for the suggestion. I think that the following Makefile changes
will be sufficient and Kconfig doesn't need to be modified:

diff --git a/drivers/pci/controller/cadence/Makefile
b/drivers/pci/controller/cadence/Makefile
index 5e23f8539ecc..1a97c9b249b8 100644
--- a/drivers/pci/controller/cadence/Makefile
+++ b/drivers/pci/controller/cadence/Makefile
@@ -4,4 +4,6 @@ obj-$(CONFIG_PCIE_CADENCE_HOST) += pcie-cadence-host.o
 obj-$(CONFIG_PCIE_CADENCE_EP) += pcie-cadence-ep.o
 obj-$(CONFIG_PCIE_CADENCE_PLAT) += pcie-cadence-plat.o
 obj-$(CONFIG_PCI_J721E) += pci-j721e.o
+pci_j721e-y := pci-j721e.o pcie-cadence.o
 obj-$(CONFIG_PCIE_SG2042_HOST) += pcie-sg2042.o
+pci_sg2042_host-y := pci-sg2042.o pcie-cadence.o

If either of PCI_J721E or SG2042_HOST is selected as a built-in module,
then pcie-cadence-host.c, pcie-cadence-ep.c and pcie-cadence.c drivers will
be built-in. If both PCI_J721E and SG2042_HOST are selected as loadable
modules, only then the library drivers will be enabled as loadable modules.

Please let me know what you think.

Regards,
Siddharth.