Hi Moritz, Michal,
Am Dienstag, den 28.07.2015, 06:57 -0700 schrieb Moritz Fischer:
[...]
> >>> +Example:
> >>> + rstc: rstc@240 {
> >>> + #reset-cells = <1>;
> >>> + compatible = "xlnx,zynq-reset-pl";
> >>> + syscon = <&slcr>;
> >>
> >> Why the syscon phandle if rstc always is the child of slcr? Why not just
> >> request the syscon for the rstc's parent node.
> >
> > We are using this description for pincntrl which was properly reviewed
> > that's why I expect Moritz just use the same style.
> > But yes also referencing parent should work.
>
> Michal is right, I tried to be consistent with the pinctrl. Either one
> is fine for me.
> We'll just have to make a decision :-)
Do you have a pointer to the pinctrl review? I'd like to know if
somebody had a good reason to use the phandle over the parent-child
relationship. I'd rather not add DT properties if they are not
necessary.
Regarding consistency, since the pinctrl node is also a child of the
slcr, you could just as well make the syscon phandle optional there and
remove it from the DT without breaking backwards compatibility.
> > TBH I don't have strong preference but having unified style is something
> > what I would prefer.
> >
> > Also I see that using parent is used by others and it looks like that
> > having something like syscon_regmap_lookup_parent will be worth to have.
That would be useful, yes.
regards
Philipp
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
