diff mbox series

[v2,1/6] dt-bindings: dma: uniphier-xdmac: Remove extension register region description

Message ID 1584604252-13172-2-git-send-email-hayashi.kunihiko@socionext.com (mailing list archive)
State New, archived
Headers show
Series Add devicetree features and fixes for UniPhier SoCs | expand

Commit Message

Kunihiko Hayashi March 19, 2020, 7:50 a.m. UTC
The address of the extension register region in example is incorrect,
however, this extension register region is optional and isn't currently
referred from the driver, so the description of the region should be
suppressed until referred by the driver.

Fixes: b9fb56b6ba8a ("dt-bindings: dmaengine: Add UniPhier external DMA controller bindings")
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
---
 Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

Comments

Masahiro Yamada March 19, 2020, 2:18 p.m. UTC | #1
On Thu, Mar 19, 2020 at 4:51 PM Kunihiko Hayashi
<hayashi.kunihiko@socionext.com> wrote:
>
> The address of the extension register region in example is incorrect,
> however, this extension register region is optional


On which SoC is it optional?

In your previous DT submission, every reg was,
like this:

reg = <0x5fc10000 0x1000>, <0x5fc20000 0x800>;



and you meant

reg = <0x5fc10000 0x1000>, <0x5fc12000 0x800>;

?

> and isn't currently
> referred from the driver, so the description of the region should be
> suppressed until referred by the driver.

This sounds like you plan to get it back
as you extend the driver.

I checked the datasheet. This controller has more registers,
so you split the reg into small chunks, the final form will look scary:

reg = <0x5fc10000 0x1000>, <0x5fc12000 0x800>,
      <0x5fc14000 0x100>,  <0x5fc15000 0x100>;


My question is why you did not use a single reg tuple
that covers the whole registers.

Is any other hardware reg interleaved in between?






>
> Fixes: b9fb56b6ba8a ("dt-bindings: dmaengine: Add UniPhier external DMA controller bindings")
> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
> ---
>  Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml b/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
> index 86cfb59..f4c3f49 100644
> --- a/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
> +++ b/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
> @@ -24,7 +24,6 @@ properties:
>    reg:
>      items:
>        - description: XDMAC base register region (offset and length)
> -      - description: XDMAC extension register region (offset and length)
>
>    interrupts:
>      maxItems: 1
> @@ -54,7 +53,7 @@ examples:
>    - |
>      xdmac: dma-controller@5fc10000 {
>          compatible = "socionext,uniphier-xdmac";
> -        reg = <0x5fc10000 0x1000>, <0x5fc20000 0x800>;
> +        reg = <0x5fc10000 0x1000>;
>          interrupts = <0 188 4>;
>          #dma-cells = <2>;
>          dma-channels = <16>;
> --
> 2.7.4
>
Kunihiko Hayashi March 23, 2020, 2:26 a.m. UTC | #2
On Thu, 19 Mar 2020 23:18:29 +0900 <masahiroy@kernel.org> wrote:

> On Thu, Mar 19, 2020 at 4:51 PM Kunihiko Hayashi
> <hayashi.kunihiko@socionext.com> wrote:
> >
> > The address of the extension register region in example is incorrect,
> > however, this extension register region is optional
> 
> 
> On which SoC is it optional?
> 
> In your previous DT submission, every reg was,
> like this:
> 
> reg = <0x5fc10000 0x1000>, <0x5fc20000 0x800>;
> 
> 
> 
> and you meant
> 
> reg = <0x5fc10000 0x1000>, <0x5fc12000 0x800>;
> 
> ?

Yes. 'Optional' might not be appropriate because all SoCs has the region.

> > and isn't currently
> > referred from the driver, so the description of the region should be
> > suppressed until referred by the driver.
> 
> This sounds like you plan to get it back
> as you extend the driver.

Right, however, it isn't desiable that the description of the region is
changed by extending the driver.

> I checked the datasheet. This controller has more registers,
> so you split the reg into small chunks, the final form will look scary:
> 
> reg = <0x5fc10000 0x1000>, <0x5fc12000 0x800>,
>       <0x5fc14000 0x100>,  <0x5fc15000 0x100>;
> 
> 
> My question is why you did not use a single reg tuple
> that covers the whole registers.
> 
> Is any other hardware reg interleaved in between?

No, there is no other hardware reg between each region.

Surely it seems pointless to divide all tuples individually.
I'll rewrite it to cover entire xdmac reg region.

Thank you,

---
Best Regards,
Kunihiko Hayashi
diff mbox series

Patch

diff --git a/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml b/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
index 86cfb59..f4c3f49 100644
--- a/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
+++ b/Documentation/devicetree/bindings/dma/socionext,uniphier-xdmac.yaml
@@ -24,7 +24,6 @@  properties:
   reg:
     items:
       - description: XDMAC base register region (offset and length)
-      - description: XDMAC extension register region (offset and length)
 
   interrupts:
     maxItems: 1
@@ -54,7 +53,7 @@  examples:
   - |
     xdmac: dma-controller@5fc10000 {
         compatible = "socionext,uniphier-xdmac";
-        reg = <0x5fc10000 0x1000>, <0x5fc20000 0x800>;
+        reg = <0x5fc10000 0x1000>;
         interrupts = <0 188 4>;
         #dma-cells = <2>;
         dma-channels = <16>;