Re: [RFC PATCH v2 1/4] dt-bindings: net: can: add STM32 bxcan DT bindings

[Dario Binacchi]

Hi Krzysztof,

On Tue, Aug 23, 2022 at 3:41 PM Krzysztof Kozlowski
<krzysztof.kozlowski@xxxxxxxxxx> wrote:
>
> On 20/08/2022 11:29, Dario Binacchi wrote:
> > Add documentation of device tree bindings for the STM32 basic extended
> > CAN (bxcan) controller.
> >
> > Signed-off-by: Dario Binacchi <dariobin@xxxxxxxxx>
> > Signed-off-by: Dario Binacchi <dario.binacchi@xxxxxxxxxxxxxxxxxxxx>
> >
> > ---
> >
> > Changes in v2:
> > - Change the file name into 'st,stm32-bxcan-core.yaml'.
> > - Rename compatibles:
> >   - st,stm32-bxcan-core -> st,stm32f4-bxcan-core
> >   - st,stm32-bxcan -> st,stm32f4-bxcan
> > - Rename master property to st,can-master.
> > - Remove the status property from the example.
> > - Put the node child properties as required.
> >
> >  .../bindings/net/can/st,stm32-bxcan.yaml      | 136 ++++++++++++++++++
> >  1 file changed, 136 insertions(+)
> >  create mode 100644 Documentation/devicetree/bindings/net/can/st,stm32-bxcan.yaml
> >
> > diff --git a/Documentation/devicetree/bindings/net/can/st,stm32-bxcan.yaml b/Documentation/devicetree/bindings/net/can/st,stm32-bxcan.yaml
> > new file mode 100644
> > index 000000000000..288631b5556d
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/net/can/st,stm32-bxcan.yaml
> > @@ -0,0 +1,136 @@
> > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > +%YAML 1.2
> > +---
> > +$id: http://devicetree.org/schemas/net/can/st,stm32-bxcan.yaml#
> > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > +
> > +title: STMicroelectronics bxCAN controller
> > +
> > +description: STMicroelectronics BxCAN controller for CAN bus
> > +
> > +maintainers:
> > +  - Dario Binacchi <dario.binacchi@xxxxxxxxxxxxxxxxxxxx>
> > +
> > +allOf:
> > +  - $ref: can-controller.yaml#
> > +
> > +properties:
> > +  compatible:
> > +    enum:
> > +      - st,stm32f4-bxcan-core
> > +
> > +  reg:
> > +    maxItems: 1
> > +
> > +  resets:
> > +    maxItems: 1
> > +
> > +  clocks:
> > +    description:
> > +      Input clock for registers access
> > +    maxItems: 1
> > +
> > +  '#address-cells':
> > +    const: 1
> > +
> > +  '#size-cells':
> > +    const: 0
> > +
> > +additionalProperties: false
> > +
> > +required:
> > +  - compatible
> > +  - reg
> > +  - resets
> > +  - clocks
> > +  - '#address-cells'
> > +  - '#size-cells'
> > +
> > +patternProperties:
>
> No improvements here, so my comment stay. Please fix it.

Sorry, I'ff fix it in version 3.

>
>
> > +  "^can@[0-9]+$":
> > +    type: object
> > +    description:
> > +      A CAN block node contains two subnodes, representing each one a CAN
> > +      instance available on the machine.
>
> I still do not understand why you need children. You did not CC me on
> driver change, so difficult to say. You did not describe the parent

On the next submissions I'll send you all the series patches.

> device - there is no description.

Ok, I'll do it.

> Why do you need parent device at all?
> This looks like some driver-driven-bindings instead of just real
> hardware description.

The two devices are not independent.
As described in the reference manual RM0386 (STM32F469xx and STM32F479xx
advanced Arm®-based 32-bit MCUs) in paragraph 34.2, the bxCAN controller is a
dual CAN peripheral configuration:

• CAN1: Master bxCAN for managing the communication between a Slave bxCAN and
the 512-byte SRAM memory
• CAN2: Slave bxCAN, with no direct access to the SRAM memory.

So, if I want to use CAN2 only (and not CAN1), I need to be able to use shared
resources with CAN1 without having to probe the CAN1 driver. IMHO here is the
justification of the parent node.

Thanks and regards,
Dario

>
> Best regards,
> Krzysztof



-- 

Dario Binacchi

Embedded Linux Developer

dario.binacchi@xxxxxxxxxxxxxxxxxxxx

__________________________________


Amarula Solutions SRL

Via Le Canevare 30, 31100 Treviso, Veneto, IT

T. +39 042 243 5310
info@xxxxxxxxxxxxxxxxxxxx

www.amarulasolutions.com