Re: [PATCH v6 1/2] dt-bindings: dma: Add Loongson-1 DMA

[Conor Dooley <conor@kernel.org>]

[-- Attachment #1: Type: text/plain, Size: 2864 bytes --]

On Mon, Mar 18, 2024 at 02:18:27PM +0800, Keguang Zhang wrote:
> On Sun, Mar 17, 2024 at 10:40 PM Conor Dooley <conor@kernel.org> wrote:
> >
> > On Sat, Mar 16, 2024 at 07:33:53PM +0800, Keguang Zhang via B4 Relay wrote:
> > > From: Keguang Zhang <keguang.zhang@gmail.com>
> > >
> > > Add devicetree binding document for Loongson-1 DMA.
> > >
> > > Signed-off-by: Keguang Zhang <keguang.zhang@gmail.com>
> > > ---
> > > V5 -> V6:
> > >    Change the compatible to the fallback
> > >    Some minor fixes
> > > V4 -> V5:
> > >    A newly added patch
> > > ---
> > >  .../devicetree/bindings/dma/loongson,ls1x-dma.yaml | 66 ++++++++++++++++++++++
> > >  1 file changed, 66 insertions(+)
> > >
> > > diff --git a/Documentation/devicetree/bindings/dma/loongson,ls1x-dma.yaml b/Documentation/devicetree/bindings/dma/loongson,ls1x-dma.yaml
> > > new file mode 100644
> > > index 000000000000..06358df725c6
> > > --- /dev/null
> > > +++ b/Documentation/devicetree/bindings/dma/loongson,ls1x-dma.yaml
> > > @@ -0,0 +1,66 @@
> > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > > +%YAML 1.2
> > > +---
> > > +$id: http://devicetree.org/schemas/dma/loongson,ls1x-dma.yaml#
> > > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > > +
> > > +title: Loongson-1 DMA Controller
> > > +
> > > +maintainers:
> > > +  - Keguang Zhang <keguang.zhang@gmail.com>
> > > +
> > > +description:
> > > +  Loongson-1 DMA controller provides 3 independent channels for
> > > +  peripherals such as NAND and AC97.
> > > +
> > > +properties:
> > > +  compatible:
> > > +    oneOf:
> > > +      - const: loongson,ls1b-dma
> > > +      - items:
> > > +          - enum:
> > > +              - loongson,ls1c-dma
> > > +          - const: loongson,ls1b-dma
> >
> > Aren't there several more devices in this family? Do they not have DMA
> > controllers?
> >
> You are right. Loongson1 is a SoC family.
> However, only 1A/1B/1C have DMA controller.

You're missing the 1A then.

> 
> > > +
> > > +  reg:
> > > +    maxItems: 1
> > > +
> > > +  interrupts:
> > > +    description: Each channel has a dedicated interrupt line.
> > > +    minItems: 1
> > > +    maxItems: 3
> >
> > Is this number not fixed for each SoC?
> >
> Yes. Actually, it's fixed for the whole family.

Then why do you have minItems: 1? Are there multiple DMA controllers
on each SoC that only make use of a subset of the possible channels?

> > > +  interrupt-names:
> > > +    minItems: 1
> > > +    items:
> > > +      - pattern: ch0
> > > +      - pattern: ch1
> > > +      - pattern: ch2
> >
> > Why have you made these a pattern? There's no regex being used here at
> > all.
> >
> Will change items to the following regex.
>   interrupt-names:
>     minItems: 1
>     items:
>       - pattern: "^ch[0-2]$"


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]