Mark Brown wrote: > On Wed, Nov 04, 2009 at 08:00:52PM +0530, Gadiyar, Anand wrote: > > Mark Brown wrote: > > > On Wed, Nov 04, 2009 at 06:02:48PM +0530, Gupta, Ajay Kumar wrote: > > > > > I can do regulator_get() and regulator_enable() but where ? Are you referring to board file? > > > > > I was thinking of doing it in a generic form within EHCI driver (ehci-omap.c). > > > > Do it in the EHCI driver. The EHCI driver should just unconditionally > > > use the regulators, the regulator API will be compiled out if support is > > > disabled in Kconfig. > > > NAK to do it unconditionally in the driver. This is board specific. > > It just so happens that this is the regulator used on the beagleboard > > and EVM, but it could very easily be another one (like on the SDP). > > Obviously the physical regulator hookup is board specific - please take > a look at how the regulator API operates here, none of this needs to be > handled by drivers using the API which was what the discussion above > (regarding regulator_get() and regulator_enable()) was about. > > regulator_get() takes the struct device for the consumer (the EHCI > controller in this case) and a name for the supply both of which depend > only on the consumer. The board drivers set up the mapping from these > to actual physical regulators during initialisation and the whole > process is completely transparent to the consumer drivers, they can just > ask for a fixed thing. > Okay. I see now. Thanks. So, if a board does not need a regulator at all, then the APIs do nothing? > > How about a hook from the platform code in mach-omap2/usb-ehci.c instead? > > If this is needed something is broken. > > -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html