[2/2] dt-bindings: phy: Fix UniPhier SATA controller node names in example

  The word "glue" is ambiguous for the controller node name. Should put
"sata-controller" instead of "ahci-glue" on the controller node.
And rename a phy node to "sata-phy".

Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
---
 .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
  

On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
> The word "glue" is ambiguous for the controller node name. Should put
> "sata-controller" instead of "ahci-glue" on the controller node.
> And rename a phy node to "sata-phy".
> 
> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
> ---
>  .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> index a3cd45acea28..89b8b133777a 100644
> --- a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> +++ b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> @@ -117,14 +117,14 @@ additionalProperties: false
>  
>  examples:
>    - |
> -    ahci-glue@65700000 {
> +    sata-controller@65700000 {

But is this really the SATA controller aka AHCI?

In cases where we don't have a standardized name, I don't think it makes 
sense changing node names from one non-standard name to another.

>          compatible = "socionext,uniphier-pxs3-ahci-glue",
>                       "simple-mfd";
>          #address-cells = <1>;
>          #size-cells = <1>;
>          ranges = <0 0x65700000 0x100>;
>  
> -        ahci_phy: phy@10 {
> +        ahci_phy: sata-phy@10 {
>              compatible = "socionext,uniphier-pxs3-ahci-phy";
>              reg = <0x10 0x10>;
>              #phy-cells = <0>;
> -- 
> 2.25.1
> 
>
  

Hi Rob,

On 2022/10/29 4:38, Rob Herring wrote:
> On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
>> The word "glue" is ambiguous for the controller node name. Should put
>> "sata-controller" instead of "ahci-glue" on the controller node.
>> And rename a phy node to "sata-phy".
>>
>> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
>> ---
>>   .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4 ++--
>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git
> a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
>> index a3cd45acea28..89b8b133777a 100644
>> --- a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> yaml
>> +++ b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> yaml
>> @@ -117,14 +117,14 @@ additionalProperties: false
>>   
>>   examples:
>>     - |
>> -    ahci-glue@65700000 {
>> +    sata-controller@65700000 {
> 
> But is this really the SATA controller aka AHCI?
> 
> In cases where we don't have a standardized name, I don't think it makes
> sense changing node names from one non-standard name to another.

I see. This can't be named generic, so this change will drop.
And "sata-phy" in the same way.

Thank you,

---
Best Regards
Kunihiko Hayashi
  

On Tue, Nov 1, 2022 at 4:10 AM Kunihiko Hayashi
<hayashi.kunihiko@socionext.com> wrote:
>
> Hi Rob,
>
> On 2022/10/29 4:38, Rob Herring wrote:
> > On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
> >> The word "glue" is ambiguous for the controller node name. Should put
> >> "sata-controller" instead of "ahci-glue" on the controller node.
> >> And rename a phy node to "sata-phy".
> >>
> >> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
> >> ---
> >>   .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4 ++--
> >>   1 file changed, 2 insertions(+), 2 deletions(-)
> >>
> >> diff --git
> > a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> > b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> >> index a3cd45acea28..89b8b133777a 100644
> >> --- a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> > yaml
> >> +++ b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> > yaml
> >> @@ -117,14 +117,14 @@ additionalProperties: false
> >>
> >>   examples:
> >>     - |
> >> -    ahci-glue@65700000 {
> >> +    sata-controller@65700000 {
> >
> > But is this really the SATA controller aka AHCI?
> >
> > In cases where we don't have a standardized name, I don't think it makes
> > sense changing node names from one non-standard name to another.
>
> I see. This can't be named generic, so this change will drop.
> And "sata-phy" in the same way.

I think sata-phy is fine. It may not be added, but usb3-phy or
pcie-phy are common already.

Rob
  

On 2022/11/02 4:31, Rob Herring wrote:
> On Tue, Nov 1, 2022 at 4:10 AM Kunihiko Hayashi
> <hayashi.kunihiko@socionext.com> wrote:
>>
>> Hi Rob,
>>
>> On 2022/10/29 4:38, Rob Herring wrote:
>>> On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
>>>> The word "glue" is ambiguous for the controller node name. Should put
>>>> "sata-controller" instead of "ahci-glue" on the controller node.
>>>> And rename a phy node to "sata-phy".
>>>>
>>>> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
>>>> ---
>>>>    .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4
>>>> ++--
>>>>    1 file changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git
>>> a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
>>> b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
>>>> index a3cd45acea28..89b8b133777a 100644
>>>> ---
>>>> a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
>>> yaml
>>>> +++
>>>> b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
>>> yaml
>>>> @@ -117,14 +117,14 @@ additionalProperties: false
>>>>
>>>>    examples:
>>>>      - |
>>>> -    ahci-glue@65700000 {
>>>> +    sata-controller@65700000 {
>>>
>>> But is this really the SATA controller aka AHCI?
>>>
>>> In cases where we don't have a standardized name, I don't think it makes
>>> sense changing node names from one non-standard name to another.
>>
>> I see. This can't be named generic, so this change will drop.
>> And "sata-phy" in the same way.
> 
> I think sata-phy is fine. It may not be added, but usb3-phy or
> pcie-phy are common already.
I understand. But of cource "phy" is also common, so I defer the change
from "phy" to "sata-phy".

Thank you,

---
Best Regards
Kunihiko Hayashi
  

On 02-11-22, 08:14, Kunihiko Hayashi wrote:
> On 2022/11/02 4:31, Rob Herring wrote:
> > On Tue, Nov 1, 2022 at 4:10 AM Kunihiko Hayashi
> > <hayashi.kunihiko@socionext.com> wrote:
> > > 
> > > Hi Rob,
> > > 
> > > On 2022/10/29 4:38, Rob Herring wrote:
> > > > On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
> > > > > The word "glue" is ambiguous for the controller node name. Should put
> > > > > "sata-controller" instead of "ahci-glue" on the controller node.
> > > > > And rename a phy node to "sata-phy".
> > > > > 
> > > > > Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
> > > > > ---
> > > > >    .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml  | 4
> > > > > ++--
> > > > >    1 file changed, 2 insertions(+), 2 deletions(-)
> > > > > 
> > > > > diff --git
> > > > a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> > > > b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> > > > > index a3cd45acea28..89b8b133777a 100644
> > > > > ---
> > > > > a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> > > > yaml
> > > > > +++
> > > > > b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
> > > > yaml
> > > > > @@ -117,14 +117,14 @@ additionalProperties: false
> > > > > 
> > > > >    examples:
> > > > >      - |
> > > > > -    ahci-glue@65700000 {
> > > > > +    sata-controller@65700000 {
> > > > 
> > > > But is this really the SATA controller aka AHCI?
> > > > 
> > > > In cases where we don't have a standardized name, I don't think it makes
> > > > sense changing node names from one non-standard name to another.
> > > 
> > > I see. This can't be named generic, so this change will drop.
> > > And "sata-phy" in the same way.
> > 
> > I think sata-phy is fine. It may not be added, but usb3-phy or
> > pcie-phy are common already.
> I understand. But of cource "phy" is also common, so I defer the change
> from "phy" to "sata-phy".

Should it not be just phy@xxx does it really need to be foo-phy?
  

Hi Vinod,

On 2022/11/05 16:34, Vinod Koul wrote:
> On 02-11-22, 08:14, Kunihiko Hayashi wrote:
>> On 2022/11/02 4:31, Rob Herring wrote:
>>> On Tue, Nov 1, 2022 at 4:10 AM Kunihiko Hayashi
>>> <hayashi.kunihiko@socionext.com> wrote:
>>>>
>>>> Hi Rob,
>>>>
>>>> On 2022/10/29 4:38, Rob Herring wrote:
>>>>> On Thu, Oct 27, 2022 at 01:49:20PM +0900, Kunihiko Hayashi wrote:
>>>>>> The word "glue" is ambiguous for the controller node name.
> Should put
>>>>>> "sata-controller" instead of "ahci-glue" on the controller node.
>>>>>> And rename a phy node to "sata-phy".
>>>>>>
>>>>>> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
>>>>>> ---
>>>>>>     .../devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
> | 4
>>>>>> ++--
>>>>>>     1 file changed, 2 insertions(+), 2 deletions(-)
>>>>>>
>>>>>> diff --git
>>>>>
> a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
>>>>>
> b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.yaml
>>>>>> index a3cd45acea28..89b8b133777a 100644
>>>>>> ---
>>>>>>
> a/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
>>>>> yaml
>>>>>> +++
>>>>>>
> b/Documentation/devicetree/bindings/phy/socionext,uniphier-ahci-phy.
>>>>> yaml
>>>>>> @@ -117,14 +117,14 @@ additionalProperties: false
>>>>>>
>>>>>>     examples:
>>>>>>       - |
>>>>>> -    ahci-glue@65700000 {
>>>>>> +    sata-controller@65700000 {
>>>>>
>>>>> But is this really the SATA controller aka AHCI?
>>>>>
>>>>> In cases where we don't have a standardized name, I don't think it
> makes
>>>>> sense changing node names from one non-standard name to another.
>>>>
>>>> I see. This can't be named generic, so this change will drop.
>>>> And "sata-phy" in the same way.
>>>
>>> I think sata-phy is fine. It may not be added, but usb3-phy or
>>> pcie-phy are common already.
>> I understand. But of cource "phy" is also common, so I defer the change
>> from "phy" to "sata-phy".
> 
> Should it not be just phy@xxx does it really need to be foo-phy?

This change was pointed out in the previous review for our devicetree,
so I made this change in the devicetree, however, I think I can leave it
as "phy" until the properties that depend on "sata-phy" are defined.

Thank you,

---
Best Regards
Kunihiko Hayashi