Re: [PATCH] arm64: dts: imx8mm-verdin: Link reg_nvcc_sd to usdhc2

[Philippe Schenker <philippe.schenker@impulsing.ch>]

[-- Attachment #1: Type: text/plain, Size: 1614 bytes --]

On Tue, 2025-04-22 at 10:41 +0200, Francesco Dolcini wrote:
> On Tue, Apr 22, 2025 at 07:57:32AM +0000, Philippe Schenker wrote:
> > Hi Francesco,
> Hey Philippe!
> 
> > Not sure this causes any side-effects maybe you guys want to
> > investigate further about this.
> 
> Yes, we did, the correct implementation would be the one I linked
> in the previous email.

Interesting, thanks for your further explanation. I do understand now.
The implementation with your SMARC module indeed looks much cleaner and
is better to understand what happens.

Please let me know if you wish to remove the vqmmc linkage on i.MX 8MP
-> I would need to test for side effects on my side. Please use the old
mail for it as I already hijacked this one enough :-).

https://lore.kernel.org/all/20240109121627.223017-1-dev@pschenker.ch/

> 
> > I needed it due to the strange requirements I had (described in
> > commit
> > message).  From my point of view it is correct to link the vqmmc-
> > supply so
> > the voltage can be set also to something different than the default
> > fusing
> > values.
> 
> It does not really work fine, because you have this IO driven by the
> SDHCI
> core that is going to affect the PMIC behavior at the same time as
> the I2C
> communication. And even if you remove it from the pinctrl, it's the
> default
> out-of-reset function, so you would have to override it and set this
> pin as
> GPIO even when not used (this would work, of course).
> 
> 
> My request is to fix it in a slighlty different way that matches with
> the way
> the HW was designed.
> 

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 717 bytes --]