linux-kernel - Re: [PATCH v3 1/4] dt-bindings: iio: adc: adi,ad4000: Add PulSAR

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <a311de1b-cd59-4f67-9bd1-61596a54c8cd@kernel.org>
Date: Fri, 22 Nov 2024 19:29:06 +0100
From: Krzysztof Kozlowski <krzk@...nel.org>
To: Marcelo Schmitt <marcelo.schmitt1@...il.com>
Cc: Marcelo Schmitt <marcelo.schmitt@...log.com>, lars@...afoo.de,
 Michael.Hennerich@...log.com, jic23@...nel.org, robh@...nel.org,
 krzk+dt@...nel.org, conor+dt@...nel.org, linux-iio@...r.kernel.org,
 devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 1/4] dt-bindings: iio: adc: adi,ad4000: Add PulSAR

On 22/11/2024 16:33, Marcelo Schmitt wrote:
>>
>>> +      - items:
>>> +          - enum:
>>> +              - adi,ad7942
>>> +          - const: adi,ad7946
>>> +
>>> +      - const: adi,ad7983
>>> +      - items:
>>> +          - enum:
>>> +              - adi,ad7980
>>> +              - adi,ad7988-5
>>> +              - adi,ad7686
>>> +              - adi,ad7685
>>
>> Keep alphabetical order.
> 
> Do the fallbacks declared here have any impact on the match try order or on how
> the compatible list should be ordered?

I don't understand, we do not talk about fallbacks. I also do not
understand at all how this relates to my comment.

> The only significant difference between each group of devices is the sample rate.
> A faster device can read at slower sample rates so if somebody knows to have
> a 16-bit pseudo-differential PulSAR but doesn't know about the exact model they
> could have a compatible like
>       compatible = "adi,ad7980", "adi,ad7988-5", "adi,ad7686", "adi,ad7685",
>                    "adi,ad7988-1", "adi,ad7983";

Can't you autodetect this?

> 
> to try from fastest to slowest device.
> The dt doc would indicate that order in the fallback list?
>       - items:
>           - enum:
>               - adi,ad7980    # Fastest 16-bit pseudo-differential ADC
>               - adi,ad7988-5  # 2nd fastest 16-bit pseudo-differential ADC
>               - adi,ad7686    # 3rd fastest 16-bit pseudo-differential ADC
>               - adi,ad7685    # 4th fastest 16-bit pseudo-differential ADC
>               - adi,ad7988-1  # 5th fastest 16-bit pseudo-differential ADC
>           - const: adi,ad7983 # Slowest 16-bit pseudo-differential ADC

Again, only one fallback here, not sure what are you asking about. BTW,
DT spec explains compatibles...

> 
> https://www.analog.com/media/en/technical-documentation/data-sheets/ad7691.pdf
> has a nice table with the different devices and sample rates.
> 
> writing-bindings.rst says "DO use fallback compatibles when devices are the same
> as or a subset of prior implementations."
> But, how can we use fallbacks properly?

How DT spec and tutorials like elinux ask... What is exactly the problem
or question?

> From Documentation/devicetree/bindings/display/bridge/lvds-codec.yaml I'm

How LVDS bridge is related to this one here?

> inferring only one fallback should be provided per group of devices.
> 
>>
>>> +              - adi,ad7988-1
>>> +          - const: adi,ad7983
>>> +
>>> +      - const: adi,ad7688
>>> +      - items:
>>> +          - enum:
>>> +              - adi,ad7693
>>> +              - adi,ad7687
>>> +          - const: adi,ad7688
>>> +
>>> +      - const: adi,ad7984
>>> +      - items:
>>> +          - enum:
>>> +              - adi,ad7982
>>> +              - adi,ad7690
>>> +              - adi,ad7691
>>> +          - const: adi,ad7984
>>> +
>>>    reg:
>>>      maxItems: 1
>>>  
>>> @@ -133,6 +178,32 @@ required:
>>>    - ref-supply
>>>  
>>>  allOf:
>>> +  # Single-channel PulSAR devices have SDI either tied to VIO, GND, or host CS.
>>> +  - if:
>>> +      properties:
>>> +        compatible:
>>> +          contains:
>>> +            enum:
>>> +              - adi,ad7685
>>
>> Why do you need this? It's fallback is already here.
> 
> So dtbs_check can provide an error message if for example compatible = "adi,ad7687";
> and adi,sdi-pin = "sdi";


I mean this compatible, not if clause.




Best regards,
Krzysztof