[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1531237902.2708.76.camel@baylibre.com>
Date: Tue, 10 Jul 2018 17:51:42 +0200
From: Jerome Brunet <jbrunet@...libre.com>
To: Yixun Lan <yixun.lan@...ogic.com>,
Martin Blumenstingl <martin.blumenstingl@...glemail.com>,
jian.hu@...ogic.com
Cc: Neil Armstrong <narmstrong@...libre.com>,
qianggui.song@...ogic.com, sboyd@...nel.org, khilman@...libre.com,
mturquette@...libre.com, linux-kernel@...r.kernel.org,
bo.yang@...ogic.com, qiufang.dai@...ogic.com,
linux-arm-kernel@...ts.infradead.org, carlo@...one.org,
linux-amlogic@...ts.infradead.org, sunny.luo@...ogic.com,
linux-clk@...r.kernel.org, xingyu.chen@...ogic.com,
Rob Herring <robh@...nel.org>
Subject: Re: [PATCH 3/3] clk: meson-g12a: Add EE Clock controller driver
On Tue, 2018-07-10 at 20:59 +0800, Yixun Lan wrote:
> > > > > +static const struct of_device_id clkc_match_table[] = {
> > > > > + { .compatible = "amlogic,g12a-clkc" },
> > > >
> > > > I wonder if that should be "amlogic,meson-g12a-clkc" (I know that AXG
> > > > and GXBB/GXL/GXM omit the "meson-" prefix, but I am not sure if that's
> > > > correct - I'll CC Rob Herring on the dt-bindings patch)
> > > >
> > >
> > > we're fine with either way, since the g12a is recently new added SoC.
> > >
> > > but I guess those old compatibles which already exist won't be changed,
> > > as it will break the DT
> >
> > Yes, I doubt we are going to change the compatibles. With or without the meson
> > prefix, I suppose it going to stay that way.
> >
> > It is worth thinking about it for new drivers though.
> > We already have the "amlogic,", so do we really to prefix even more with "meson-
> > "
> >
> > Also, if we were really strict about that scheme, should meson8b compatible be
> >
> > "amlogic,meson-meson8b-foo-bar" ... the SoC is meson8b, not just 8b after all ?
> >
> > For the particular case of the clock controller, I don't care very much, as long
> > as the compatible is coherent with documentation.
> >
> > I might have a preference for the current compatible ... just for consistency
> > with the other clock controllers.
> >
>
> do you imply that we should just stick to 'amlogic,g12a-clkc'?
>
> it sound good to me, at least it's shorter than 'amlogic,meson-g12a-clkc'..
I did not imply anything.
Martin raised a valid point. I'm merely giving my opinion on the matter and I'm
interested to get others view on this.
Powered by blists - more mailing lists