diff mbox series

[v2,06/13] dt-bindings: serial: atmel,at91-usart: Add SAM9260 compatibles to SAM9x60

Message ID 20220906135511.144725-7-sergiu.moga@microchip.com (mailing list archive)
State Superseded
Headers show
Series Make atmel serial driver aware of GCLK | expand

Commit Message

Sergiu Moga Sept. 6, 2022, 1:55 p.m. UTC
Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
in order to highlight the incremental characteristics of the SAM9X60
serial IP.

Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
---


v1 -> v2:
- Nothing, this patch was not here before


 Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
 1 file changed, 2 insertions(+)

Comments

Krzysztof Kozlowski Sept. 8, 2022, 12:30 p.m. UTC | #1
On 06/09/2022 15:55, Sergiu Moga wrote:
> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
> in order to highlight the incremental characteristics of the SAM9X60
> serial IP.
> 
> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
> ---
> 
> 
> v1 -> v2:
> - Nothing, this patch was not here before
> 
> 
>  Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> index b25535b7a4d2..4d80006963c7 100644
> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> @@ -26,6 +26,8 @@ properties:
>        - items:
>            - const: microchip,sam9x60-dbgu
>            - const: microchip,sam9x60-usart
> +          - const: atmel,at91sam9260-dbgu
> +          - const: atmel,at91sam9260-usart

This is weird. You say in commit msg to "highlight the incremental
characteristics" but you basically change here existing compatibles.
This is not enum, but a list.


Best regards,
Krzysztof
Sergiu Moga Sept. 8, 2022, 3:15 p.m. UTC | #2
On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
> On 06/09/2022 15:55, Sergiu Moga wrote:
>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>> in order to highlight the incremental characteristics of the SAM9X60
>> serial IP.
>>
>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>> ---
>>
>>
>> v1 -> v2:
>> - Nothing, this patch was not here before
>>
>>
>>   Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>> index b25535b7a4d2..4d80006963c7 100644
>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>> @@ -26,6 +26,8 @@ properties:
>>         - items:
>>             - const: microchip,sam9x60-dbgu
>>             - const: microchip,sam9x60-usart
>> +          - const: atmel,at91sam9260-dbgu
>> +          - const: atmel,at91sam9260-usart
> 
> This is weird. You say in commit msg to "highlight the incremental
> characteristics" but you basically change here existing compatibles.


Does "show that they are incremental IP's" sound better then?


> This is not enum, but a list.
> 


What do you mean by this? I know it is a list, I specified so in the 
commit message.


> 
> Best regards,
> Krzysztof


Regards,
	Sergiu
Rob Herring (Arm) Sept. 9, 2022, 1:36 a.m. UTC | #3
On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
> > On 06/09/2022 15:55, Sergiu Moga wrote:
> >> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
> >> in order to highlight the incremental characteristics of the SAM9X60
> >> serial IP.
> >>
> >> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
> >> ---
> >>
> >>
> >> v1 -> v2:
> >> - Nothing, this patch was not here before
> >>
> >>
> >>   Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
> >>   1 file changed, 2 insertions(+)
> >>
> >> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> >> index b25535b7a4d2..4d80006963c7 100644
> >> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> >> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
> >> @@ -26,6 +26,8 @@ properties:
> >>         - items:
> >>             - const: microchip,sam9x60-dbgu
> >>             - const: microchip,sam9x60-usart
> >> +          - const: atmel,at91sam9260-dbgu
> >> +          - const: atmel,at91sam9260-usart
> > 
> > This is weird. You say in commit msg to "highlight the incremental
> > characteristics" but you basically change here existing compatibles.
> 
> 
> Does "show that they are incremental IP's" sound better then?
> 
> 
> > This is not enum, but a list.
> > 
> 
> 
> What do you mean by this? I know it is a list, I specified so in the 
> commit message.

You are saying that compatible must be exactly the 4 strings above in 
the order listed. You need another entry with another 'items' list.

Rob
Sergiu Moga Sept. 12, 2022, 7:45 a.m. UTC | #4
On 09.09.2022 04:36, Rob Herring wrote:
> On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
>> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
>>> On 06/09/2022 15:55, Sergiu Moga wrote:
>>>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>>>> in order to highlight the incremental characteristics of the SAM9X60
>>>> serial IP.
>>>>
>>>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>>>> ---
>>>>
>>>>
>>>> v1 -> v2:
>>>> - Nothing, this patch was not here before
>>>>
>>>>
>>>>    Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>>>    1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>> index b25535b7a4d2..4d80006963c7 100644
>>>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>> @@ -26,6 +26,8 @@ properties:
>>>>          - items:
>>>>              - const: microchip,sam9x60-dbgu
>>>>              - const: microchip,sam9x60-usart
>>>> +          - const: atmel,at91sam9260-dbgu
>>>> +          - const: atmel,at91sam9260-usart
>>>
>>> This is weird. You say in commit msg to "highlight the incremental
>>> characteristics" but you basically change here existing compatibles.
>>
>>
>> Does "show that they are incremental IP's" sound better then?
>>
>>
>>> This is not enum, but a list.
>>>
>>
>>
>> What do you mean by this? I know it is a list, I specified so in the
>> commit message.
> 
> You are saying that compatible must be exactly the 4 strings above in
> the order listed. You need another entry with another 'items' list.
> 
> Rob


That is what was intended though: a list of the 4 compatibles in that 
exact order. The 4th patch of this series also ensures that all 9x60 
nodes have that exact list of 4 compatibles.

Regards,
	Sergiu
Krzysztof Kozlowski Sept. 12, 2022, 10:44 a.m. UTC | #5
On 12/09/2022 09:45, Sergiu.Moga@microchip.com wrote:
> On 09.09.2022 04:36, Rob Herring wrote:
>> On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
>>> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
>>>> On 06/09/2022 15:55, Sergiu Moga wrote:
>>>>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>>>>> in order to highlight the incremental characteristics of the SAM9X60
>>>>> serial IP.
>>>>>
>>>>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>>>>> ---
>>>>>
>>>>>
>>>>> v1 -> v2:
>>>>> - Nothing, this patch was not here before
>>>>>
>>>>>
>>>>>    Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>>>>    1 file changed, 2 insertions(+)
>>>>>
>>>>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>> index b25535b7a4d2..4d80006963c7 100644
>>>>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>> @@ -26,6 +26,8 @@ properties:
>>>>>          - items:
>>>>>              - const: microchip,sam9x60-dbgu
>>>>>              - const: microchip,sam9x60-usart
>>>>> +          - const: atmel,at91sam9260-dbgu
>>>>> +          - const: atmel,at91sam9260-usart
>>>>
>>>> This is weird. You say in commit msg to "highlight the incremental
>>>> characteristics" but you basically change here existing compatibles.
>>>
>>>
>>> Does "show that they are incremental IP's" sound better then?
>>>
>>>
>>>> This is not enum, but a list.
>>>>
>>>
>>>
>>> What do you mean by this? I know it is a list, I specified so in the
>>> commit message.
>>
>> You are saying that compatible must be exactly the 4 strings above in
>> the order listed. You need another entry with another 'items' list.
>>
>> Rob
> 
> 
> That is what was intended though: a list of the 4 compatibles in that 
> exact order. The 4th patch of this series also ensures that all 9x60 
> nodes have that exact list of 4 compatibles.

The commit msg suggest otherwise - two options, because it is
incremental... But this one is not really incremental - you require this
one, only one, configuration. It's in general fine, but commit msg
should reflect what you are really intend to do here and why you are
doing it.


Best regards,
Krzysztof
Sergiu Moga Sept. 12, 2022, 1:09 p.m. UTC | #6
On 12.09.2022 13:44, Krzysztof Kozlowski wrote:
> On 12/09/2022 09:45, Sergiu.Moga@microchip.com wrote:
>> On 09.09.2022 04:36, Rob Herring wrote:
>>> On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
>>>> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
>>>>> On 06/09/2022 15:55, Sergiu Moga wrote:
>>>>>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>>>>>> in order to highlight the incremental characteristics of the SAM9X60
>>>>>> serial IP.
>>>>>>
>>>>>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>>>>>> ---
>>>>>>
>>>>>>
>>>>>> v1 -> v2:
>>>>>> - Nothing, this patch was not here before
>>>>>>
>>>>>>
>>>>>>     Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>>>>>     1 file changed, 2 insertions(+)
>>>>>>
>>>>>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>> index b25535b7a4d2..4d80006963c7 100644
>>>>>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>> @@ -26,6 +26,8 @@ properties:
>>>>>>           - items:
>>>>>>               - const: microchip,sam9x60-dbgu
>>>>>>               - const: microchip,sam9x60-usart
>>>>>> +          - const: atmel,at91sam9260-dbgu
>>>>>> +          - const: atmel,at91sam9260-usart
>>>>>
>>>>> This is weird. You say in commit msg to "highlight the incremental
>>>>> characteristics" but you basically change here existing compatibles.
>>>>
>>>>
>>>> Does "show that they are incremental IP's" sound better then?
>>>>
>>>>
>>>>> This is not enum, but a list.
>>>>>
>>>>
>>>>
>>>> What do you mean by this? I know it is a list, I specified so in the
>>>> commit message.
>>>
>>> You are saying that compatible must be exactly the 4 strings above in
>>> the order listed. You need another entry with another 'items' list.
>>>
>>> Rob
>>
>>
>> That is what was intended though: a list of the 4 compatibles in that
>> exact order. The 4th patch of this series also ensures that all 9x60
>> nodes have that exact list of 4 compatibles.
> 
> The commit msg suggest otherwise - two options, because it is
> incremental... But this one is not really incremental - you require this
> one, only one, configuration. It's in general fine, but commit msg
> should reflect what you are really intend to do here and why you are
> doing it.
> 
> 
> Best regards,
> Krzysztof


My apologies, I still do not understand what is wrong with the commit 
message. My intention was to ensure that every 9x60 usart compatible is 
followed by the 9260 compatibles because the 9x60 serial IP is an 
improvement over the 9260 one. Would you prefer it to be just the first 
part of the commit message: `Add the AT91SAM9260 serial compatibles to 
the list of SAM9X60 compatibles`? That way it would really only be what 
the commit does.

Thank you,
	Sergiu
Krzysztof Kozlowski Sept. 13, 2022, 8:58 a.m. UTC | #7
On 12/09/2022 15:09, Sergiu.Moga@microchip.com wrote:
> On 12.09.2022 13:44, Krzysztof Kozlowski wrote:
>> On 12/09/2022 09:45, Sergiu.Moga@microchip.com wrote:
>>> On 09.09.2022 04:36, Rob Herring wrote:
>>>> On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
>>>>> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
>>>>>> On 06/09/2022 15:55, Sergiu Moga wrote:
>>>>>>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>>>>>>> in order to highlight the incremental characteristics of the SAM9X60
>>>>>>> serial IP.
>>>>>>>
>>>>>>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>>>>>>> ---
>>>>>>>
>>>>>>>
>>>>>>> v1 -> v2:
>>>>>>> - Nothing, this patch was not here before
>>>>>>>
>>>>>>>
>>>>>>>     Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>>>>>>     1 file changed, 2 insertions(+)
>>>>>>>
>>>>>>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>> index b25535b7a4d2..4d80006963c7 100644
>>>>>>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>> @@ -26,6 +26,8 @@ properties:
>>>>>>>           - items:
>>>>>>>               - const: microchip,sam9x60-dbgu
>>>>>>>               - const: microchip,sam9x60-usart
>>>>>>> +          - const: atmel,at91sam9260-dbgu
>>>>>>> +          - const: atmel,at91sam9260-usart
>>>>>>
>>>>>> This is weird. You say in commit msg to "highlight the incremental
>>>>>> characteristics" but you basically change here existing compatibles.
>>>>>
>>>>>
>>>>> Does "show that they are incremental IP's" sound better then?
>>>>>
>>>>>
>>>>>> This is not enum, but a list.
>>>>>>
>>>>>
>>>>>
>>>>> What do you mean by this? I know it is a list, I specified so in the
>>>>> commit message.
>>>>
>>>> You are saying that compatible must be exactly the 4 strings above in
>>>> the order listed. You need another entry with another 'items' list.
>>>>
>>>> Rob
>>>
>>>
>>> That is what was intended though: a list of the 4 compatibles in that
>>> exact order. The 4th patch of this series also ensures that all 9x60
>>> nodes have that exact list of 4 compatibles.
>>
>> The commit msg suggest otherwise - two options, because it is
>> incremental... But this one is not really incremental - you require this
>> one, only one, configuration. It's in general fine, but commit msg
>> should reflect what you are really intend to do here and why you are
>> doing it.
>>
>>
>> Best regards,
>> Krzysztof
> 
> 
> My apologies, I still do not understand what is wrong with the commit 
> message. My intention was to ensure that every 9x60 usart compatible is 
> followed by the 9260 compatibles because the 9x60 serial IP is an 
> improvement over the 9260 one. Would you prefer it to be just the first 
> part of the commit message: `Add the AT91SAM9260 serial compatibles to 
> the list of SAM9X60 compatibles`? That way it would really only be what 
> the commit does.

Let me rephrase it:

What your commit is doing is requiring additional fallback compatibles.
Therefore the commit msg should answer - why do you require additional
fallback compatibles?

Incremental characteristics sound to me optional. I can increment
sam9x60 with something or I can skip it. But you are not doing it...
sam9x60 was already there and now you require a fallback.

Best regards,
Krzysztof
Sergiu Moga Sept. 13, 2022, 9:19 a.m. UTC | #8
On 13.09.2022 11:58, Krzysztof Kozlowski wrote:
> On 12/09/2022 15:09, Sergiu.Moga@microchip.com wrote:
>> On 12.09.2022 13:44, Krzysztof Kozlowski wrote:
>>> On 12/09/2022 09:45, Sergiu.Moga@microchip.com wrote:
>>>> On 09.09.2022 04:36, Rob Herring wrote:
>>>>> On Thu, Sep 08, 2022 at 03:15:44PM +0000, Sergiu.Moga@microchip.com wrote:
>>>>>> On 08.09.2022 15:30, Krzysztof Kozlowski wrote:
>>>>>>> On 06/09/2022 15:55, Sergiu Moga wrote:
>>>>>>>> Add the AT91SAM9260 serial compatibles to the list of SAM9X60 compatibles
>>>>>>>> in order to highlight the incremental characteristics of the SAM9X60
>>>>>>>> serial IP.
>>>>>>>>
>>>>>>>> Signed-off-by: Sergiu Moga <sergiu.moga@microchip.com>
>>>>>>>> ---
>>>>>>>>
>>>>>>>>
>>>>>>>> v1 -> v2:
>>>>>>>> - Nothing, this patch was not here before
>>>>>>>>
>>>>>>>>
>>>>>>>>      Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml | 2 ++
>>>>>>>>      1 file changed, 2 insertions(+)
>>>>>>>>
>>>>>>>> diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>>> index b25535b7a4d2..4d80006963c7 100644
>>>>>>>> --- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>>> +++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
>>>>>>>> @@ -26,6 +26,8 @@ properties:
>>>>>>>>            - items:
>>>>>>>>                - const: microchip,sam9x60-dbgu
>>>>>>>>                - const: microchip,sam9x60-usart
>>>>>>>> +          - const: atmel,at91sam9260-dbgu
>>>>>>>> +          - const: atmel,at91sam9260-usart
>>>>>>>
>>>>>>> This is weird. You say in commit msg to "highlight the incremental
>>>>>>> characteristics" but you basically change here existing compatibles.
>>>>>>
>>>>>>
>>>>>> Does "show that they are incremental IP's" sound better then?
>>>>>>
>>>>>>
>>>>>>> This is not enum, but a list.
>>>>>>>
>>>>>>
>>>>>>
>>>>>> What do you mean by this? I know it is a list, I specified so in the
>>>>>> commit message.
>>>>>
>>>>> You are saying that compatible must be exactly the 4 strings above in
>>>>> the order listed. You need another entry with another 'items' list.
>>>>>
>>>>> Rob
>>>>
>>>>
>>>> That is what was intended though: a list of the 4 compatibles in that
>>>> exact order. The 4th patch of this series also ensures that all 9x60
>>>> nodes have that exact list of 4 compatibles.
>>>
>>> The commit msg suggest otherwise - two options, because it is
>>> incremental... But this one is not really incremental - you require this
>>> one, only one, configuration. It's in general fine, but commit msg
>>> should reflect what you are really intend to do here and why you are
>>> doing it.
>>>
>>>
>>> Best regards,
>>> Krzysztof
>>
>>
>> My apologies, I still do not understand what is wrong with the commit
>> message. My intention was to ensure that every 9x60 usart compatible is
>> followed by the 9260 compatibles because the 9x60 serial IP is an
>> improvement over the 9260 one. Would you prefer it to be just the first
>> part of the commit message: `Add the AT91SAM9260 serial compatibles to
>> the list of SAM9X60 compatibles`? That way it would really only be what
>> the commit does.
> 
> Let me rephrase it:
> 
> What your commit is doing is requiring additional fallback compatibles.
> Therefore the commit msg should answer - why do you require additional
> fallback compatibles?
> 


The additional fallback compatibles are required because the driver in 
question only knows about the atmel,at91sam9260-usart compatible. 
Furthermore, it is also a better representation of the fact that the 
serial IP of 9x60 is an improvement over the serial IP of 9260 (it 
contains more hardware features not yet implemented in the driver).


> Incremental characteristics sound to me optional. I can increment
> sam9x60 with something or I can skip it. But you are not doing it...
> sam9x60 was already there and now you require a fallback.
> 
> Best regards,
> Krzysztof

So, what is your opinion on the following commit message:

"Fix sam9x60 compatible list by adding the sam9260 compatibles as 
fallback, since the atmel_serial driver only knows of the latter's 
compatible. The atmel_serial driver only has knowledge of the sam9260 
compatible because it does not have the sam9x60's serial IP specific 
features implemented yet and adding an empty compatible without adding 
support specific to that compatible would be misleading. Thus prefer the 
fallback mechanism in the detriment of adding an empty compatible in the 
driver."

Thanks,
	Sergiu
Krzysztof Kozlowski Sept. 13, 2022, 9:30 a.m. UTC | #9
On 13/09/2022 11:19, Sergiu.Moga@microchip.com wrote:
>>
>> Let me rephrase it:
>>
>> What your commit is doing is requiring additional fallback compatibles.
>> Therefore the commit msg should answer - why do you require additional
>> fallback compatibles?
>>
> 
> 
> The additional fallback compatibles are required because the driver in 
> question only knows about the atmel,at91sam9260-usart compatible. 
> Furthermore, it is also a better representation of the fact that the 
> serial IP of 9x60 is an improvement over the serial IP of 9260 (it 
> contains more hardware features not yet implemented in the driver).
> 
> 
>> Incremental characteristics sound to me optional. I can increment
>> sam9x60 with something or I can skip it. But you are not doing it...
>> sam9x60 was already there and now you require a fallback.
>>
>> Best regards,
>> Krzysztof
> 
> So, what is your opinion on the following commit message:
> 
> "Fix sam9x60 compatible list by adding the sam9260 compatibles as 
> fallback, since the atmel_serial driver only knows of the latter's 
> compatible. The atmel_serial driver only has knowledge of the sam9260 
> compatible because it does not have the sam9x60's serial IP specific 
> features implemented yet and adding an empty compatible without adding 
> support specific to that compatible would be misleading. Thus prefer the 
> fallback mechanism in the detriment of adding an empty compatible in the 
> driver."

It's fine. Also could work:

"Require sam9260 fallback compatible for sam9x60, because sam9x60 is
fully compatible with sam9260 and Linux driver requires the latter."

If it fixes any observable issue like lack of driver binding to DTS, you
can also mention that.

Best regards,
Krzysztof
Sergiu Moga Sept. 13, 2022, 9:35 a.m. UTC | #10
On 13.09.2022 12:30, Krzysztof Kozlowski wrote:
> On 13/09/2022 11:19, Sergiu.Moga@microchip.com wrote:
>>>
>>> Let me rephrase it:
>>>
>>> What your commit is doing is requiring additional fallback compatibles.
>>> Therefore the commit msg should answer - why do you require additional
>>> fallback compatibles?
>>>
>>
>>
>> The additional fallback compatibles are required because the driver in
>> question only knows about the atmel,at91sam9260-usart compatible.
>> Furthermore, it is also a better representation of the fact that the
>> serial IP of 9x60 is an improvement over the serial IP of 9260 (it
>> contains more hardware features not yet implemented in the driver).
>>
>>
>>> Incremental characteristics sound to me optional. I can increment
>>> sam9x60 with something or I can skip it. But you are not doing it...
>>> sam9x60 was already there and now you require a fallback.
>>>
>>> Best regards,
>>> Krzysztof
>>
>> So, what is your opinion on the following commit message:
>>
>> "Fix sam9x60 compatible list by adding the sam9260 compatibles as
>> fallback, since the atmel_serial driver only knows of the latter's
>> compatible. The atmel_serial driver only has knowledge of the sam9260
>> compatible because it does not have the sam9x60's serial IP specific
>> features implemented yet and adding an empty compatible without adding
>> support specific to that compatible would be misleading. Thus prefer the
>> fallback mechanism in the detriment of adding an empty compatible in the
>> driver."
> 
> It's fine. Also could work:
> 
> "Require sam9260 fallback compatible for sam9x60, because sam9x60 is
> fully compatible with sam9260 and Linux driver requires the latter."
> 


This version looks better indeed. Sums it all up and is only 2 lines :). 
Thank you very much for the suggestion it is greatly appeciated.


> If it fixes any observable issue like lack of driver binding to DTS, you
> can also mention that.
> 
> Best regards,
> Krzysztof


Thanks,
	Sergiu
diff mbox series

Patch

diff --git a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
index b25535b7a4d2..4d80006963c7 100644
--- a/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
+++ b/Documentation/devicetree/bindings/serial/atmel,at91-usart.yaml
@@ -26,6 +26,8 @@  properties:
       - items:
           - const: microchip,sam9x60-dbgu
           - const: microchip,sam9x60-usart
+          - const: atmel,at91sam9260-dbgu
+          - const: atmel,at91sam9260-usart
 
   reg:
     maxItems: 1