Re: [PATCH] thermal: rockchip: shut up GRF warning

["Heiko Stübner" <heiko@sntech.de>]

Am Montag, 18. August 2025, 21:23:54 Mitteleuropäische Sommerzeit schrieb Sebastian Reichel:
> On Mon, Aug 18, 2025 at 08:44:15PM +0200, Heiko Stübner wrote:
> > Am Montag, 18. August 2025, 19:26:15 Mitteleuropäische Sommerzeit schrieb Sebastian Reichel:
> > > @@ -1621,12 +1656,16 @@ static int rockchip_configure_from_dt(struct device *dev,
> > >  		return -EINVAL;
> > >  	}
> > >  
> > > -	/* The tsadc wont to handle the error in here since some SoCs didn't
> > > -	 * need this property.
> > > -	 */
> > > -	thermal->grf = syscon_regmap_lookup_by_phandle(np, "rockchip,grf");
> > > -	if (IS_ERR(thermal->grf))
> > > -		dev_warn(dev, "Missing rockchip,grf property\n");
> > > +	if (thermal->chip->grf_mode != GRF_NONE) {
> > > +		thermal->grf = syscon_regmap_lookup_by_phandle(np, "rockchip,grf");
> > > +		if (IS_ERR(thermal->grf)) {
> > > +			ret = PTR_ERR(thermal->grf);
> > > +			if (thermal->chip->grf_mode == GRF_OPTIONAL)
> > > +				dev_warn(dev, "Missing rockchip,grf property\n");
> > 
> > I guess it might make it easier for people seeing the log, if we could
> > insert an "optional" into that message for the optional tier.
> 
> Sure, I can add an "optional". I'm not sure how "optional" they
> really are, though. Code like this looks quite fishy to me:
> 
> if (grf)
>     regmap_write(grf, ..., RK3568_GRF_TSADC_TSEN);
> 
> I marked these as optional, as the driver should probe without the
> GRF. But to me it looks like most platforms with optional GRF
> support should have been made mandatory in the first place.

okay ... then I take it back, and we should leave out the optional.

That way people may wonder and possibly verify their soc variant
or build setup.


Heiko