Re: [PATCH 2/4] remoteproc: qcom_q6v5_pas: Use resource with CX PD for MSM8974

[Luca Weiss]

On 06-07-2025 12:08 p.m., Dmitry Baryshkov wrote:
> On Sun, Jul 06, 2025 at 10:31:38AM +0200, Luca Weiss wrote:
> > On 05-07-2025 10:57 p.m., Dmitry Baryshkov wrote:
> > > On Sat, Jun 21, 2025 at 03:19:57PM +0200, Luca Weiss wrote:
> > > > MSM8974 requires the CX power domain, so use the msm8996_adsp_resource
> > > > which has cx under proxy_pd_names and is otherwise equivalent.
> > > >
> > > > Signed-off-by: Luca Weiss <luca@xxxxxxxxxxxx>
> > > > ---
> > > >    drivers/remoteproc/qcom_q6v5_pas.c | 2 +-
> > > >    1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > Hmm. You are modifying the ADSP configuration in the driver, but at the
> > > same time you've dropped CX supply from the MSS remoteproc.
> >
> > The qcom_q6v5_mss driver has this support for .fallback_proxy_supply, which
> > are used in case the power domain is not specified.
> >
> > So no driver change is necessary in the mss driver for both old and new
> > devicetrees, but the adsp driver does not have this fallback, so that's why
> > the adsp config is updated.
> >
> > Does that make it clear?
>
> Yes. Would it make sense to implement fallback_proxy_supply for ADSP
> too?

Definitely would be possible, but I don't see the point in doing the 
work to implement this, to then carry around a bunch of legacy 
compatibility code that (very likely) won't really be used in practice.
I don't think any platform apart from msm8974 are going to be affected 
by this anyways.

Still same argument from my side, I think breaking compatibility here 
for this one driver outweighs the effort/code of implementing compatibility.

Regards
Luca

> > Regards
> > Luca