Re: [PATCH v8 1/3] dt-bindings: mmc: mtk-sd: extend interrupts and pinctrls properties

[Rob Herring]

On Tue, Mar 22, 2022 at 05:33:55PM +0800, Axe Yang wrote:
> Hello AngeloGioacchino,
> 
> On Tue, 2022-03-22 at 09:42 +0100, AngeloGioacchino Del Regno wrote:
> > Il 22/03/22 02:35, Axe Yang ha scritto:
> > > On Mon, 2022-03-21 at 18:29 -0500, Rob Herring wrote:
> > > > On Mon, Mar 21, 2022 at 07:51:32PM +0800, Axe Yang wrote:
> > > > > Extend interrupts and pinctrls for SDIO wakeup interrupt
> > > > > feature.
> > > > > This feature allow SDIO devices alarm asynchronous interrupt to
> > > > > host
> > > > > even when host stop providing clock to SDIO card. An extra
> > > > > wakeup
> > > > > interrupt and pinctrl states for SDIO DAT1 pin state switching
> > > > > are
> > > > > required in this scenario.
> > > > > 
> > > > > Signed-off-by: Axe Yang <axe.yang@xxxxxxxxxxxx>
> > > > > ---
> > > > >   .../devicetree/bindings/mmc/mtk-sd.yaml       | 23
> > > > > ++++++++++++++++++-
> > > > >   1 file changed, 22 insertions(+), 1 deletion(-)
> > > > > 
> > > > > diff --git a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > > > b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > > > index 297ada03e3de..f57774535a1d 100644
> > > > > --- a/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > > > +++ b/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
> > > > > @@ -69,12 +69,23 @@ properties:
> > > > >         - const: ahb_cg
> > > > >   
> > > > >     interrupts:
> > > > > -    maxItems: 1
> > > > > +    description:
> > > > > +      Should at least contain MSDC GIC interrupt. To support
> > > > > SDIO
> > > > > in-band wakeup, an extended
> > > > > +      interrupt is required and be configured as wakeup source
> > > > > irq.
> > > > > +    minItems: 1
> > > > > +    maxItems: 2
> > > > >   
> > > > >     pinctrl-names:
> > > > > +    description:
> > > > > +      Should at least contain default and state_uhs. To
> > > > > support
> > > > > SDIO in-band wakeup, dat1 pin
> > > > > +      will be switched between GPIO mode and SDIO DAT1 mode,
> > > > > state_eint and state_dat1 are
> > > > > +      mandatory in this scenarios.
> > > > > +    minItems: 2
> > > > >       items:
> > > > >         - const: default
> > > > >         - const: state_uhs
> > > > > +      - const: state_eint
> > > > > +      - const: state_dat1
> > > > >   
> > > > >     pinctrl-0:
> > > > >       description:
> > > > > @@ -86,6 +97,16 @@ properties:
> > > > >         should contain uhs mode pin ctrl.
> > > > >       maxItems: 1
> > > > >   
> > > > > +  pinctrl-2:
> > > > > +    description:
> > > > > +      should switch dat1 pin to GPIO mode.
> > > > > +    maxItems: 1
> > > > > +
> > > > > +  pinctrl-3:
> > > > > +    description:
> > > > > +      should switch SDIO dat1 pin from GPIO mode back to SDIO
> > > > > mode.
> > > > 
> > > > How is this different than pinctrl-0?
> > > 
> > > pinctrl-0 contains default settings for all IO pins(CLK/CMD/DAT).
> > > pinctrl-1 contains settings for all IO pins(CLK/CMD/DAT) in UHS
> > > mode.
> > > pinctrl-3 is lightweight pinctrl-1, only keep SDIO DAT1 pin
> > > function
> > > switch part.
> > > 
> > 
> > Is there any particular reason why we cannot simply select pinctrl-1
> > again
> > instead of pinctrl-3, apart from the virtually not existent overhead
> > of one more mmio write?
> 
> No, there is no particular reason. 
> I just want to do the pin function switch quick and clean. 
> 
> The intention of pinctrl-1 is to set the most initial state of IO pins
> in UHS mode. If I don't need to adjust IO settings any longer, it is
> okay to select pinctrl-1 state instead of pinctrl-3. 
> But think about this scenarios: after initial SDIO IO pins to UHS mode,
> I want to adjust some IO related properties, such as driving strength.
> And I want to keep these settings because with new driving strength,
> the signal is better. I'd rather to choose pinctrl-3 but not pinctrl-1, 
> because I do not want the change be restored after next runtime resume.

The pinctrl-X properties set modes, they aren't supposed to be a state 
machine.

Rob