Re: [PATCH v3 11/50] dt-bindings: crypto: add sam9x7 in Atmel TDES

[Tudor Ambarus <tudor.ambarus@linaro.org>]

On 8/21/23 09:41, Nicolas Ferre wrote:
> Hi Tudor, all,
> 
Hi,

> On 19/08/2023 at 16:34, Krzysztof Kozlowski wrote:
>> On 10/08/2023 09:22, Tudor Ambarus wrote:
>>>
>>> On 8/10/23 06:38, Varshini.Rajendran@microchip.com wrote:
>>>>> On 7/28/23 11:24, Varshini Rajendran wrote:
>>>>>> Add DT bindings for atmel TDES.
>>>>> NACK. The atmel crypto drivers check the version at runtime and
>>>>> fill a capabilities structure based on the version identified.
>>>>> There's a single compatible regardless of the version of the IP
>>>>> used until now, why do you want to change it?
>>>>>
>>>> Hi Tudor,
>>>
>>> Hi,
>>>
>>>> I am aware that there is no change in the crypto IP used. This patch is
> 
> Actually, recent history showed us that it's not only the IP itself but
> its integration into final product that could have an influence on the
> behavior.
> 
>>>> to add a SoC specific compatible as expected by writing-bindings
>>>> guideline. Maybe a bit more explanation in the commit description might
>>>> do the trick.
>>>>
>>>
>>> So you add a compatible that will never be used just to comply with
>>> the writing bindings guideline?
>>
>> How do you know that it is never going to be used? The guideline asks

See
https://git.kernel.org/pub/scm/linux/kernel/git/at91/linux.git/tree/drivers/crypto/atmel-tdes.c?h=at91-dt#n1120

>> for this on purpose, so any future quirks or incompatibilities can be
>> easily addressed.
> 
> In this recent case, having a an adapted compatibility string is an
> added value.
> 
> And yes, I changed my mind and would like to be systematic now with
> at91/microchip DT compatibility strings. Our long history and big legacy
> in arm-soc is sometimes difficult to handle, but we're moving little by
> little to comply with guidelines.
> 
> My conclusion is that Varshini's addition is the way to go.

Ok, fine by me. Then it would be good if one adds compatibles for the
previous SoCs as well and add a comment in the drivers that inform
readers that the atmel_*_get_cap() methods are used as backup where
"atmel,at91sam9g46-" compatibles are used. You'll then have all the
previous SoCs have their own dedicated compatibles which will have
"atmel,at91sam9g46-" compatible as backup, and "sam9x7" will be the
first that will not need the "atmel,at91sam9g46-" backup compatible.
In the drivers you'll have 2 flavors of identifying the IP caps, the
first one that backups to atmel_*_get_cap(), and a second one where
of_device_id data will suffice.

If the commit message described how the driver will handle the new
compatible, Varshini would have spared us of all these emails exchanged.
Varshini, please update the commit message in the next iteration and
describe how the driver will handle the new compatible.

Cheers,
ta