[v1,1/2] dt-bindings: input: atmel,maxtouch: add poweroff-in-suspend property

From: Stefan Eichenberger <stefan.eichenberger@toradex.com>

Add a new property to indicate that the device should be powered off in
suspend mode.

Signed-off-by: Stefan Eichenberger <stefan.eichenberger@toradex.com>
---
 Documentation/devicetree/bindings/input/atmel,maxtouch.yaml | 6 ++++++
 1 file changed, 6 insertions(+)

On Thu, Dec 7, 2023 at 12:13 PM Stefan Eichenberger <eichest@gmail.com> wrote:

> From: Stefan Eichenberger <stefan.eichenberger@toradex.com>
>
> Add a new property to indicate that the device should be powered off in
> suspend mode.
>
> Signed-off-by: Stefan Eichenberger <stefan.eichenberger@toradex.com>
(...)
> +  atmel,poweroff-in-suspend:
> +    description: |
> +      When this property is set, all supplies are turned off when the system is
> +      going to suspend.
> +    type: boolean
   wakeup-source:
     type: boolean

As Krzysztof says it seems you are describing an operating system feature.

I can't help but wonder: shouldn't that pretty much be the default behaviour
if wakeup-source is *not* specified?

I.e. the property kind of describes !wakeup-source.

Yours,
Linus Walleij

Hi Linus, Krzysztof,

On Fri, Dec 08, 2023 at 01:54:21PM +0100, Linus Walleij wrote:
> On Thu, Dec 7, 2023 at 12:13 PM Stefan Eichenberger <eichest@gmail.com> wrote:
> 
> > From: Stefan Eichenberger <stefan.eichenberger@toradex.com>
> >
> > Add a new property to indicate that the device should be powered off in
> > suspend mode.
> >
> > Signed-off-by: Stefan Eichenberger <stefan.eichenberger@toradex.com>
> (...)
> > +  atmel,poweroff-in-suspend:
> > +    description: |
> > +      When this property is set, all supplies are turned off when the system is
> > +      going to suspend.
> > +    type: boolean
>    wakeup-source:
>      type: boolean
> 
> As Krzysztof says it seems you are describing an operating system feature.

It appears to be an OS feature, but I would argue that it is also a
property of a board. It is tempting to say that if DTS defines supplies
for the controller we should use them to power off the controller in
suspend, otherwise we should use the deep sleep functionality of the
controller. But a mere presence of regulators does not indicate if they
can actually be powered off in suspend (i.e. if controllers shares power
rails with another device that can be a wakeup source), so we need to
have additional hints on how OS should behave on a given device.

On top of that we have regulator framework supplying dummy regulators...

Thanks.

On Fri, Dec 08, 2023 at 11:37:47PM +0000, Dmitry Torokhov wrote:
> Hi Linus, Krzysztof,
> 
> On Fri, Dec 08, 2023 at 01:54:21PM +0100, Linus Walleij wrote:
> > On Thu, Dec 7, 2023 at 12:13 PM Stefan Eichenberger <eichest@gmail.com> wrote:
> > 
> > > From: Stefan Eichenberger <stefan.eichenberger@toradex.com>
> > >
> > > Add a new property to indicate that the device should be powered off in
> > > suspend mode.
> > >
> > > Signed-off-by: Stefan Eichenberger <stefan.eichenberger@toradex.com>
> > (...)
> > > +  atmel,poweroff-in-suspend:
> > > +    description: |
> > > +      When this property is set, all supplies are turned off when the system is
> > > +      going to suspend.
> > > +    type: boolean
> >    wakeup-source:
> >      type: boolean
> > 
> > As Krzysztof says it seems you are describing an operating system feature.
> 
> It appears to be an OS feature, but I would argue that it is also a
> property of a board. It is tempting to say that if DTS defines supplies
> for the controller we should use them to power off the controller in
> suspend, otherwise we should use the deep sleep functionality of the
> controller. But a mere presence of regulators does not indicate if they
> can actually be powered off in suspend (i.e. if controllers shares power
> rails with another device that can be a wakeup source), so we need to
> have additional hints on how OS should behave on a given device.
> 
> On top of that we have regulator framework supplying dummy regulators...

Simply rephrasing the property might be sufficient? The current one
sounds making policy decisions with the "should be". Reframing the
commit message and property description etc in terms of what aspect of
the hardware the ability to turn off all supplies in suspend comes from
would make it more acceptable. Pretty much answering the question "why
can't we try and turn off all supplies on all systems with this device"
should get things rolling.

Cheers,
Conor.