Re: [PATCH 1/2] dt-bindings: media: i2c: Add OmniVision OV13858

[Bryan O'Donoghue <bod@kernel.org>]

On 17/10/2026 17:57, Krzysztof Kozlowski wrote:
> On Fri, Oct 16, 2026 at 09:49:16PM +0100, Bryan O'Donoghue wrote:
>> On 15/10/2026 22:48, George Neacsu via B4 Relay wrote:
>>> +  - clocks
>>> +  - avdd-supply
>>> +  - dovdd-supply
>>> +  - dvdd-supply
>>
>> The clocks, supplies and resets should all be optional as nothing says Linux
>> must control any of those on every system.
>>
>> i.e. you might have a co-processor or CPLD which owns the reset, power-rails
>> and the clock might not come from the SoC.
> 
> That would mean that every other binding/device would have them
> optional. What Linux controls or does not, is not really important here
> - these are required in the board design.
> 
> Best regards,
> Krzysztof

Yeah I take your point, they should be optional in the driver. Actually 
though we don't always know or represent a regulator for example or the 
regulator can be on a daughter board - as is common on Laptops.

What is the DT policy for that ? My own understanding was that it is 
better to not require all the regulators from the DT description and 
leave it up to the individual DTs to describe what can be described.

Should we have a fake regulator - or reuse one ?

OK the LLM tells me that I should have 'regulator-fixed' for that case 
always-on, no GPIO, so describing a regulator we know exists but can't 
actually control, hmm.

Fair enough, that's the answer.

---
bod