Re: [PATCH 3/3] dt-bindings: mfd: add "syscon-smc" YAML description

["Clément Léger" <clement.leger@bootlin.com>]

Le Thu, 29 Jul 2021 15:19:19 -0600,
Rob Herring <robh@kernel.org> a écrit :

> On Fri, Jul 23, 2021 at 03:52:39PM +0200, Clément Léger wrote:
> > This patch adds documentation for the "syscon-smc" compatible which
> > describes a syscon using a SMC regmap instead of a MMIO one. This
> > allows accessing system controllers that are set as secure by using
> > SMC handled by the secure monitor.
> > 
> > Signed-off-by: Clément Léger <clement.leger@bootlin.com>
> > ---
> >  .../devicetree/bindings/mfd/syscon-smc.yaml   | 57
> > +++++++++++++++++++ 1 file changed, 57 insertions(+)
> >  create mode 100644
> > Documentation/devicetree/bindings/mfd/syscon-smc.yaml
> > 
> > diff --git a/Documentation/devicetree/bindings/mfd/syscon-smc.yaml
> > b/Documentation/devicetree/bindings/mfd/syscon-smc.yaml new file
> > mode 100644 index 000000000000..6ce1392c5e7f
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/mfd/syscon-smc.yaml
> > @@ -0,0 +1,57 @@
> > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > +%YAML 1.2
> > +---
> > +$id: http://devicetree.org/schemas/mfd/syscon-smc.yaml#
> > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > +
> > +title: System Controller Registers R/W via SMC Device Tree Bindings
> > +
> > +description: |
> > +  System controller SMC node represents a register region
> > containing a set
> > +  of miscellaneous registers accessed through a secure monitor.
> > +  The typical use-case is the same as the syscon one but when
> > running with a
> > +  secure monitor.
> > +
> > +maintainers:
> > +  - Lee Jones <lee.jones@linaro.org>
> > +
> > +properties:
> > +  compatible:
> > +    anyOf:
> > +      - items:
> > +          - enum:
> > +              - atmel,sama5d2-sfr
> > +
> > +          - const: syscon-smc  
> 
> I regret having 'syscon' as a compatible, so nak on a 2nd flavor of
> it. It's only purpose is a hint to Linux to automagically create a
> regmap for you.

Indeed.

> 
> All you need is the specific compatible, atmel,sama5d2-sfr, and you
> can imply the rest of this from it. That's assuming the conclusion is
> a register read/write interface on SMC is a good idea, but I don't
> think it is.

Ok noted, I'll try to find something else to implement that.

Clément

> 
> Rob