[PATCH 1/5] sunxi: video: No double clock on DE2

[Vasily Khoruzhick]

On Wed, Sep 16, 2020 at 9:26 AM Martin Cerveny <martin at c-home.cz> wrote:
>
>
>
> On Wed, 16 Sep 2020, Maxime Ripard wrote:
>
> > On Wed, Sep 16, 2020 at 04:10:48PM +0200, Martin Cerveny wrote:
> >> Weird code or comment. This is variant is tested on V3s.
> >>
> >> Signed-off-by: Martin Cerveny <m.cerveny at computer.org>
> >
> > Generally speaking your commit logs are fairly concise, but it really
> > becomes an issue when you're allegedly fixing a bug.
> >
> > There's a bunch of questions here that are completely up in the air:
> >
> >  - What issue are you actually trying to fix, how can one reproduce it
> >  - You claim that there's no double clock on the DE2, according to
> >    what?
> >  - DE2 is used on way more SoCs than just the V3s, did you check/test
> >    on those SoCs as well?
> >
> > In general, a good commit log should not explain what you're doing but
> > *why* you're doing it. The what can be quite easily figured out from the
> > patch content, the why can't today, and it will be even harder in a
> > year's time.
> >
> > Maxime
> >
>
> I cannot issue more information on the problem:
> - check original code - the "1x pll" was commented by #ifndef
>    (only "2x pll" is computed)
> - but original author added comment "/* No double clock on DE2 */"
> - V3s does not support "2x pll" in TCON_CLK_REG
>
> So I wrote where I tested this patch.
> It should be discussed with original author.

Sorry, I don't remember the details on why the driver uses
pll-video0-2x, that's the code from 2017

> Regards.
>
> CC: Vasily Khoruzhick <anarsoul at gmail.com>