Peter Chen <peter.chen@xxxxxxxxxxxxx> writes: > On Tue, Jan 27, 2015 at 11:21:56AM -0500, Alan Stern wrote: >> On Tue, 27 Jan 2015, Peter Chen wrote: >> >> > > > > Please fix this properly. >> > > > >> > > > I don't know the right way to fix this. Alan, has you any suggestion? >> > > >> > > It depends. How did the code before the adfa79d1c06a commit avoid this >> > > problem? By simply not enabling CONFIG_USB_EHCI_HCD? >> > > >> > >> > Hi Andy, Would you define pci_id at ci_hdrc_pci.c (instead of ehci-pci.c) >> > at your platform code/table to fix this problem? >> > I am not familiar with pci, but it works at other platforms, like ARM. >> >> You can't remove the definition of pci_id in ehci-pci.c; if you did >> then ehci-pci wouldn't bind to PCI controllers. >> >> It may turn out that the best way to fix this is to include a quirk in >> ehci-pci.c to prevent it from binding to the Intel MID on-board EHCI >> controller. >> >> Alan Stern >> > > Then, it is strange. Do we need even two glue layer drivers for pci > device? Look at usb/chipidea/ci_hdrc_pci.c it has pci_register_driver, > and its host driver will call ehci_init_driver, it is definitely > duplicated with usb/host/ehci-pci.c. No, what ci_hdrc_pci does is a superset of what ehci_pci does, being a dual role controller and all that. One register window and one interrupt is shared between a device controller, a host controller and an otg controller. Regards, -- Alex -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
