Re: [PATCH] oxnas: select reset controller framework

From: Arnd Bergmann
Date: Wed Jul 06 2016 - 09:18:58 EST


On Wednesday, July 6, 2016 3:04:25 PM CEST Neil Armstrong wrote:
> 2016-07-06 14:48 GMT+02:00 Arnd Bergmann <arnd@xxxxxxxx>:
> > For unknown reasons, we have to enable three symbols for a platform
> > to use a reset controller driver, otherwise we get a Kconfig
> > warning:
> >
> > warning: (MACH_OX810SE) selects RESET_OXNAS which has unmet direct dependencies (RESET_CONTROLLER)
> >
> > This selects the other two symbols for oxnas.
> >
> > Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>
> > ---
> > arch/arm/mach-oxnas/Kconfig | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/arch/arm/mach-oxnas/Kconfig b/arch/arm/mach-oxnas/Kconfig
> > index 2dde0acf01fd..b6a36d660e6f 100644
> > --- a/arch/arm/mach-oxnas/Kconfig
> > +++ b/arch/arm/mach-oxnas/Kconfig
> > @@ -11,11 +11,13 @@ if ARCH_OXNAS
> >
> > config MACH_OX810SE
> > bool "Support OX810SE Based Products"
> > + select ARCH_HAS_RESET_CONTROLLER
>
> This is strange since this is selected by CONFIG_ARCH_OXNAS !
>

You are right, this one is not needed.

> > select ARM_TIMER_SP8
> > select COMMON_CLK_OXNAS
> > select CPU_ARM926T
> > select MFD_SYSCON
> > select PINCTRL_OXNAS
> > + select RESET_CONTROLLER
>
> Same here, it should be selected by ARCH_HAS_RESET_CONTROLLER

I actually did a patch at one point to kill off ARCH_HAS_RESET_CONTROLLER
completely, it's not really needed at all, but that is a larger rework
so (I think) I never submitted it.

Arnd