Re: [Question] directory for SoC-related DT binding

[Stefan Wahren]

Hi,

Am 10.10.2018 um 13:19 schrieb Rob Herring:
> On Wed, Oct 10, 2018 at 6:08 AM Masahiro Yamada
> <yamada.masahiro@xxxxxxxxxxxxx> wrote:
>> Hi,
>>
>>
>> I see a bunch of vendor (or SoC) names in
>> Documentation/device/bindings/arm/
>>
>> ./Documentation/devicetree/bindings/arm/altera
>> ./Documentation/devicetree/bindings/arm/amlogic
> Yeah, it's kind of a mixture of board/soc bindings mostly with some
> ARM architecture, ARM, Ltd. IP, and SoC system reg bindings.
>
> Eventually, I'd like to not split board bindings by arch and maybe we
> should move all the system/misc reg bindings out.
>
> [,,,]
>
>> I also see some vendor names in
>> Documentation/device/bindings/soc/
>>
>> ./Documentation/devicetree/bindings/soc/bcm
>> ./Documentation/devicetree/bindings/soc/dove
>> ./Documentation/devicetree/bindings/soc/fsl
>> ./Documentation/devicetree/bindings/soc/mediatek
>> ./Documentation/devicetree/bindings/soc/qcom
>> ./Documentation/devicetree/bindings/soc/rockchip
>> ./Documentation/devicetree/bindings/soc/ti
>> ./Documentation/devicetree/bindings/soc/xilinx
>> ./Documentation/devicetree/bindings/soc/zte
> This I believe is mostly SoC system reg bindings though there's
> probably a few other things.
>
>> Confusingly, I see bcm, mediatek, rockchip
>> in both locations.
>>
>> Is there any rule to choose one than the other?
> Top-level SoC/board bindings in arm/ and anything else elsewhere ideally.

in case of Documentation/devicetree/bindings/soc/bcm the directory
contains SoC / board bindings, cpu-enable and a firmware binding.

Is there any action required?

Btw the Broadcom SoC / boards from this directory has been left out for
the yaml conversion [1] was this intended?

[1] - https://lwn.net/Articles/767723/

>
> Rob
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel