Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <66cf260e-b8b2-b3d1-7e8e-d1cfd99989b8@linaro.org>
Date:   Fri, 9 Sep 2022 09:25:16 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.13.0
Subject: Re: [PATCH] dt-bindings: sound: ts3a227e: add control of debounce
 times
Content-Language: en-US
To:     Astrid Rost <astridr@axis.com>, Mark Brown <broonie@kernel.org>
Cc:     Astrid Rost <astrid.rost@axis.com>,
        Liam Girdwood <lgirdwood@gmail.com>,
        Rob Herring <robh+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        kernel@axis.com,
        alsa-devel-mejlinglistan <alsa-devel@alsa-project.org>,
        devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
References: <20220907135827.16209-1-astrid.rost@axis.com>
 <2b81d814-f47a-e548-83dc-b1e38857e8ce@linaro.org>
 <Yxn9o1MVMPnFO3PM@sirena.org.uk>
 <b42e7fe4-9e1b-fdda-44f2-aa87b06425c8@axis.com>
From:   Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
In-Reply-To: <b42e7fe4-9e1b-fdda-44f2-aa87b06425c8@axis.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 09/09/2022 09:16, Astrid Rost wrote:
> Hello,
> 
>  > Use standard property units "-ms".
> 
> I made it in exactly the same way as it is done for the "ti,micbias:".
> 
> - ti,micbias:   Intended MICBIAS voltage (datasheet section 9.6.7).
>        Select 0/1/2/3/4/5/6/7 to specify MICBIAS voltage
>        2.1V/2.2V/2.3V/2.4V/2.5V/2.6V/2.7V/2.8V
>        Default value is "1" (2.2V).

That's not correct way. Logical units should be encoded in logical
units, e.g. microvolt. This makes the binding re-usable, extendable
(imagine new device from the family where these values map to something
else!) and also the easiest to read.

> ?> Anyway new properties cannot be accepted. This has to be converted to DT
> ?> schema (YAML).
> 
>  > Doing a whole binding conversion feels like a bit of a steep requirement
>  > when people are just adding a simple property, it's a lot of stop energy
>  > to figure out the tooling, do the conversion and deal with all the
>  > bikeshedding that the tools don't catch.  It's definitely nice if people
>  > want to look at that, for more complex binding changes it gets more
>  > reasonable but for trivial properties it's disproportionate.
> 
> Thanks, I am not really sure how yaml works. But I can have a look.

Documentation/devicetree/bindings/writing-schema.rst
(and example-schema in the same place)

Best regards,
Krzysztof