fixed-regulator with enable from another regulator

fixed-regulator with enable from another regulator

[Francesco Dolcini]

Hello all,
I am trying to figure out if it is possible with the current
regulator-fixed or with any other solution already implemented in Linux
to have a regulator enabled from another one.

I am not talking about the regulator vin-supply, I am really talking
about the enable input signal of a LDO connected to the output of another
regulator, a sort of "chained" regulator.

As an example something like this

<reg_1v8>[out] ---> [vin-supply]<reg_1v0>[out] --->
<reg_3v3>[out] ---> [enable]

<reg_1v8>[out] ---> [vin-supply]<reg_1v2>[out] --->
<reg_3v3>[out] ---> [enable]

<gpio1> ---> [enable]<reg_3v3>[out]--->

in which toggling <gpio1> control 3 different outputs.

I am still not 100% sure if describing this specific hardware connection
would be required on my system, maybe I can just ignore all of that.

Any thought?

Francesco

Re: fixed-regulator with enable from another regulator

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 615 bytes --]

On Fri, Jan 27, 2023 at 10:33:14PM +0100, Francesco Dolcini wrote:

> I am trying to figure out if it is possible with the current
> regulator-fixed or with any other solution already implemented in Linux
> to have a regulator enabled from another one.

Not really.  You can model the regulators as sharing the same
enable if there's a GPIO enable for the parent regulator,
otherwise it'd need some framework work.

> I am still not 100% sure if describing this specific hardware connection
> would be required on my system, maybe I can just ignore all of that.

It seems like a fairly strange hardware design TBH.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: fixed-regulator with enable from another regulator

[Francesco Dolcini]

Il 28 gennaio 2023 14:16:00 CET, Mark Brown <broonie@kernel.org> ha scritto:
>On Fri, Jan 27, 2023 at 10:33:14PM +0100, Francesco Dolcini wrote:
>
>> I am trying to figure out if it is possible with the current
>> regulator-fixed or with any other solution already implemented in Linux
>> to have a regulator enabled from another one.
>
>Not really.  You can model the regulators as sharing the same
>enable if there's a GPIO enable for the parent regulator,
>otherwise it'd need some framework work.

Thanks, I'll propose a patch if needed.

>> I am still not 100% sure if describing this specific hardware connection
>> would be required on my system, maybe I can just ignore all of that.
>
>It seems like a fairly strange hardware design TBH.

Just to shred some light on this weird request, I am not the one who did the design, but there is a reason.

There is one device that requires 3 different voltages, one of those come from a dedicated output of a PMIC, the other 2 from dedicated LDOs.

Having the enable of the 2 LDO on the PMIC output allows to control the power of the device without additional GPIOs usage (that were not available, for instance, while the PMIC can enable/disable the output from I2C).

(not that is relevant for this discussion, but this is of course fine for the power sequencing of this specific device).

Francesco


Hello Mark,