Re: [RFC PATCH 6/7] dt-bindings: riscv: Remove "riscv,svXX" property from device-tree

[Palmer Dabbelt]

On Sun, 22 Mar 2020 04:00:27 PDT (-0700), alex@xxxxxxxx wrote:
> This property can not be used before virtual memory is set up
> and then the  distinction between sv39 and sv48 is done at runtime
> using SATP csr property: this property is now useless, so remove it.
>
> Signed-off-by: Alexandre Ghiti <alex@xxxxxxxx>
> ---
>  Documentation/devicetree/bindings/riscv/cpus.yaml | 13 -------------
>  1 file changed, 13 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/riscv/cpus.yaml b/Documentation/devicetree/bindings/riscv/cpus.yaml
> index 04819ad379c2..12baabbac213 100644
> --- a/Documentation/devicetree/bindings/riscv/cpus.yaml
> +++ b/Documentation/devicetree/bindings/riscv/cpus.yaml
> @@ -39,19 +39,6 @@ properties:
>        Identifies that the hart uses the RISC-V instruction set
>        and identifies the type of the hart.
>
> -  mmu-type:
> -    allOf:
> -      - $ref: "/schemas/types.yaml#/definitions/string"
> -      - enum:
> -          - riscv,sv32
> -          - riscv,sv39
> -          - riscv,sv48
> -    description:
> -      Identifies the MMU address translation mode used on this
> -      hart.  These values originate from the RISC-V Privileged
> -      Specification document, available from
> -      https://riscv.org/specifications/
> -
>    riscv,isa:
>      allOf:
>        - $ref: "/schemas/types.yaml#/definitions/string"

I'd prefer if we continue to define this in the schema: while Linux won't use
it, it's still useful for other programs that want to statically determine the
available VA widths.