Re: [PATCH 1/4] arm64: defconfig: Enable rv3028 i2c rtc driver

[Krzysztof Kozlowski <krzk@kernel.org>]

On Mon, Dec 07, 2020 at 02:38:51PM +0100, Teresa Remmet wrote:
> Am Montag, den 07.12.2020, 13:10 +0100 schrieb Krzysztof Kozlowski:
> > On Fri, Dec 04, 2020 at 09:32:59PM +0100, Teresa Remmet wrote:
> > > Enable rv3028 i2c rtc driver populated on phyBOARD-Pollux-i.MX8M
> > > Plus.
> > > 
> > > Signed-off-by: Teresa Remmet <t.remmet@phytec.de>
> > > ---
> > >  arch/arm64/configs/defconfig | 1 +
> > >  1 file changed, 1 insertion(+)
> > > 
> > > diff --git a/arch/arm64/configs/defconfig
> > > b/arch/arm64/configs/defconfig
> > > index 5cfe3cf6f2ac..2034fefb3f44 100644
> > > --- a/arch/arm64/configs/defconfig
> > > +++ b/arch/arm64/configs/defconfig
> > > @@ -818,6 +818,7 @@ CONFIG_RTC_DRV_MAX77686=y
> > >  CONFIG_RTC_DRV_RK808=m
> > >  CONFIG_RTC_DRV_PCF85363=m
> > >  CONFIG_RTC_DRV_RX8581=m
> > > +CONFIG_RTC_DRV_RV3028=y
> > 
> > This should be a module.
> 
> with having this build in the clock out of this rtc will be disabled as
> not using during boot up. The clock out is enabled as default and has
> to be disabled explicit. But I guess the correct way would be to find a
> solution in the rtc driver itself?

Regardless whether it is built-in or module, the driver will behave the
same. So either the driver disables the clock, or not. The only
difference will be *when* the action happens.

Choosing it as built-in for the purpose of disabling some clock is not a
proper approach.

The defconfig here serves only a development/debugging/reference purpose.
It's not for production so anyway the effect on disabling some parts is
not that important.

Best regards,
Krzysztof