Re: [PATCH v2 01/12] dt-bindings: clock: google,gs101-clock: add PERIC0 clock management unit

[Rob Herring <robh@kernel.org>]

On Mon, Jan 08, 2024 at 02:18:21PM +0000, Peter Griffin wrote:
> Hi Tudor,
> 
> On Thu, 28 Dec 2023 at 12:58, Tudor Ambarus <tudor.ambarus@linaro.org> wrote:
> >
> > Add dt-schema documentation for the Connectivity Peripheral 0 (PERIC0)
> > clock management unit.
> >
> > Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
> > Signed-off-by: Tudor Ambarus <tudor.ambarus@linaro.org>
> > ---
> > v2:
> > - fix comments as per Sam's suggestion and collect his R-b tag
> > - Rob's suggestion of renaming the clock-names to just "bus" and "ip"
> >   was not implemented as I felt it affects readability in the driver
> >   and consistency with other exynos clock drivers. I will happily update
> >   the names in the -rc phase if someone else has a stronger opinion than
> >   mine.
> >
> 
> It would be good to get Krzysztof and Robs view on whether they agree
> with the above rationale or whether they would still like to see the
> names updated.
> 
> Personally I like the consistency, grepability and the fact the
> current name encodes whether it is a gate, divider into the name.
> Seeing 'sss' or 'ip' as a clock name in the driver code doesn't tell
> you a lot without having to then cross reference with the dts.
> 
> Is there some rationale and/or benefit behind having the shorter
> names? The only thing I could think of is trying to partially re-use
> this file on future SoCs like gs201 which might be clocked
> differently, but then these exynos clock drivers seem to be SoC
> specific anyway.

The point of -names is to identify one entry from another in the list. 
Having the name of the block is just redundant.

I like consistency, but not when it's a pattern we don't want.

Rob