diff mbox series

spi: atmel,quadspi: Define lan966x QSPI

Message ID 20220407105420.10765-1-kavyasree.kotagiri@microchip.com
State Changes Requested, archived
Headers show
Series spi: atmel,quadspi: Define lan966x QSPI | expand

Checks

Context Check Description
robh/checkpatch success
robh/patch-applied success
robh/dtbs-check warning build log
robh/dt-meta-schema success

Commit Message

Kavyasree Kotagiri April 7, 2022, 10:54 a.m. UTC
LAN966x SoC supports 3 QSPI controllers. Each of them support
data and clock frequency upto 100Mhz DDR and QUAD protocol.

Signed-off-by: Kavyasree Kotagiri <kavyasree.kotagiri@microchip.com>
---
 Documentation/devicetree/bindings/spi/atmel,quadspi.yaml | 1 +
 1 file changed, 1 insertion(+)

Comments

Mark Brown April 7, 2022, 11:02 a.m. UTC | #1
On Thu, Apr 07, 2022 at 04:24:20PM +0530, Kavyasree Kotagiri wrote:

> @@ -19,6 +19,7 @@ properties:
>        - microchip,sam9x60-qspi
>        - microchip,sama7g5-qspi
>        - microchip,sama7g5-ospi
> +      - microchip,lan966x-qspi

Generally DT compatibles should be for specific SoCs rather than having
wildcards in them, even if that means you have to list a lot of SoCs.
Having used wildcards in the past doesn't mean it's a good idea to
continue adding them!
Michael Walle April 7, 2022, 11:23 a.m. UTC | #2
> > @@ -19,6 +19,7 @@ properties:
> >        - microchip,sam9x60-qspi
> >        - microchip,sama7g5-qspi
> >        - microchip,sama7g5-ospi
> > +      - microchip,lan966x-qspi
> 
> Generally DT compatibles should be for specific SoCs rather than having
> wildcards in them, even if that means you have to list a lot of SoCs.
> Having used wildcards in the past doesn't mean it's a good idea to
> continue adding them!

The subject should also be prefixed with "dt-bindings: ".

Mark, I did a git log on
Documentation/devicetree/bindings/spi/atmel,quadspi.yaml and all the
subjects are without "dt-bindings:" although the original patch was with
that prefix [1]. Is that intended?

-michael

[1] https://lore.kernel.org/linux-devicetree/163962128492.2075495.3678727080606971257.b4-ty@kernel.org/
Mark Brown April 7, 2022, 11:31 a.m. UTC | #3
On Thu, Apr 07, 2022 at 01:23:45PM +0200, Michael Walle wrote:

> > > +      - microchip,lan966x-qspi

> > Generally DT compatibles should be for specific SoCs rather than having
> > wildcards in them, even if that means you have to list a lot of SoCs.
> > Having used wildcards in the past doesn't mean it's a good idea to
> > continue adding them!

> The subject should also be prefixed with "dt-bindings: ".

I tend to complain about people doing that.

> Mark, I did a git log on
> Documentation/devicetree/bindings/spi/atmel,quadspi.yaml and all the
> subjects are without "dt-bindings:" although the original patch was with
> that prefix [1]. Is that intended?

Yes.
Michael Walle April 7, 2022, 11:41 a.m. UTC | #4
Am 2022-04-07 13:31, schrieb Mark Brown:
> On Thu, Apr 07, 2022 at 01:23:45PM +0200, Michael Walle wrote:
>> The subject should also be prefixed with "dt-bindings: ".
> 
> I tend to complain about people doing that.

After all it is mentioned to use that prefix in
Documentation/devicetree/bindings/submitting-patches.rst. I try to
remember when submitting SPI related bindings.

-michael
Krzysztof Kozlowski April 7, 2022, 12:04 p.m. UTC | #5
On 07/04/2022 13:41, Michael Walle wrote:
> Am 2022-04-07 13:31, schrieb Mark Brown:
>> On Thu, Apr 07, 2022 at 01:23:45PM +0200, Michael Walle wrote:
>>> The subject should also be prefixed with "dt-bindings: ".
>>
>> I tend to complain about people doing that.
> 
> After all it is mentioned to use that prefix in
> Documentation/devicetree/bindings/submitting-patches.rst. I try to
> remember when submitting SPI related bindings.

From my point of view, the dt-bindings prefix is still expected, just
after "spi:" (and other Marks' subsystems), because that's I am
filtering the bindings.

Your submissions had the prefix in wrong place, this one patch does not
have it all. :(

Best regards,
Krzysztof
Krzysztof Kozlowski April 7, 2022, 12:05 p.m. UTC | #6
On 07/04/2022 12:54, Kavyasree Kotagiri wrote:
> LAN966x SoC supports 3 QSPI controllers. Each of them support
> data and clock frequency upto 100Mhz DDR and QUAD protocol.
> 
> Signed-off-by: Kavyasree Kotagiri <kavyasree.kotagiri@microchip.com>
> ---
>  Documentation/devicetree/bindings/spi/atmel,quadspi.yaml | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> index 1d493add4053..100d6e7f2748 100644
> --- a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> +++ b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> @@ -19,6 +19,7 @@ properties:
>        - microchip,sam9x60-qspi
>        - microchip,sama7g5-qspi
>        - microchip,sama7g5-ospi
> +      - microchip,lan966x-qspi

Expect the comment you got about wildcard, please also put it in
alphabetical order. As you can check, the other entries are ordered.

Best regards,
Krzysztof
Tudor Ambarus April 7, 2022, 12:54 p.m. UTC | #7
On 4/7/22 13:54, Kavyasree Kotagiri wrote:
> LAN966x SoC supports 3 QSPI controllers. Each of them support
> data and clock frequency upto 100Mhz DDR and QUAD protocol.

How is this IP different than microchip,sama7g5-qspi? Does this speed
limitation come from the IP itself or from the board that you're using?

Neither of these instances support octal mode?

Cheers,
ta

> 
> Signed-off-by: Kavyasree Kotagiri <kavyasree.kotagiri@microchip.com>
> ---
>  Documentation/devicetree/bindings/spi/atmel,quadspi.yaml | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> index 1d493add4053..100d6e7f2748 100644
> --- a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> +++ b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> @@ -19,6 +19,7 @@ properties:
>        - microchip,sam9x60-qspi
>        - microchip,sama7g5-qspi
>        - microchip,sama7g5-ospi
> +      - microchip,lan966x-qspi
>  
>    reg:
>      items:
Kavyasree Kotagiri April 8, 2022, 11:52 a.m. UTC | #8
> > LAN966x SoC supports 3 QSPI controllers. Each of them support
> > data and clock frequency upto 100Mhz DDR and QUAD protocol.
> 
> How is this IP different than microchip,sama7g5-qspi? Does this speed
> limitation come from the IP itself or from the board that you're using?
> 
> Neither of these instances support octal mode?
> 
Thanks for your comments. All the three instances support only QUAD protocol. 
You are correct. There is no difference from sama7g5-qspi. Please ignore this patch. I will send next version of dt patches where I will use "microchip,sama7g5-qspi" for all my qspi nodes.

> Cheers,
> ta
> 
> >
> > Signed-off-by: Kavyasree Kotagiri <kavyasree.kotagiri@microchip.com>
> > ---
> >  Documentation/devicetree/bindings/spi/atmel,quadspi.yaml | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> > index 1d493add4053..100d6e7f2748 100644
> > --- a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> > +++ b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
> > @@ -19,6 +19,7 @@ properties:
> >        - microchip,sam9x60-qspi
> >        - microchip,sama7g5-qspi
> >        - microchip,sama7g5-ospi
> > +      - microchip,lan966x-qspi
> >
> >    reg:
> >      items:
Michael Walle April 11, 2022, 2:46 p.m. UTC | #9
> > > LAN966x SoC supports 3 QSPI controllers. Each of them support
> > > data and clock frequency upto 100Mhz DDR and QUAD protocol.
> >
> > How is this IP different than microchip,sama7g5-qspi? Does this speed
> > limitation come from the IP itself or from the board that you're using?
> >
> > Neither of these instances support octal mode?
> >
> Thanks for your comments. All the three instances support only QUAD
> protocol.
> You are correct. There is no difference from sama7g5-qspi. Please ignore
> this patch. I will send next version of dt patches where I will use
> "microchip,sama7g5-qspi" for all my qspi nodes.

Are you sure? There is a max frequency property in Tudor's sama7g5-qspi
driver (200/133MHz) which doesn't match neither the LAN9668 manual (which
states 150MHz on QSPI0 and 100MHZ on QSPI1, funny enough there is no
mention of QSPI2) nor does it match the max frequency set in the downstream
linux driver (24 MHz).

-michael
diff mbox series

Patch

diff --git a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
index 1d493add4053..100d6e7f2748 100644
--- a/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
+++ b/Documentation/devicetree/bindings/spi/atmel,quadspi.yaml
@@ -19,6 +19,7 @@  properties:
       - microchip,sam9x60-qspi
       - microchip,sama7g5-qspi
       - microchip,sama7g5-ospi
+      - microchip,lan966x-qspi
 
   reg:
     items: