On 05-05-20, 13:04, Greg Kroah-Hartman wrote: > On Mon, May 04, 2020 at 08:04:38PM +0530, Vinod Koul wrote: > > > > --- a/drivers/usb/host/Makefile > > > > +++ b/drivers/usb/host/Makefile > > > > @@ -70,7 +70,7 @@ obj-$(CONFIG_USB_OHCI_HCD_DAVINCI) += ohci-da8xx.o > > > > obj-$(CONFIG_USB_UHCI_HCD) += uhci-hcd.o > > > > obj-$(CONFIG_USB_FHCI_HCD) += fhci.o > > > > obj-$(CONFIG_USB_XHCI_HCD) += xhci-hcd.o > > > > -obj-$(CONFIG_USB_XHCI_PCI) += xhci-pci.o > > > > +obj-$(CONFIG_USB_XHCI_PCI) += xhci-pci.o xhci-pci-renesas.o > > > > > > Hmm, now we end up with two modules, xhci-pci and xhci-pci-renesas, even if > > > xhci-pci-renesas just includes helper functions to load firmware for renesas. > > > > Right, these are two modules. Do you forsee an issue with two ko's > > Two kos should be fine, but as you aren't giving people the option to > not select this, it's a bit harsh to add it. > > Can this be a separate module/config option? Why force everyone to need > this additional code if they do not have this hardware? Since the code is moved out and is based on PCI ID of the device, this wont be invoked at all for folks not having this hardware. But adding a config option would work too and avoid renaming file. I think this looks as a better option to me atm. We no longer have load order issue with current approach so we dont care about that as well. Mathias, let me know if you are okay with approach, I can respin this, or if you have better idea do let us know Thanks -- ~Vinod
