Re: [PATCH] backlight: ktd253: Bring up in a known state

[Linus Walleij <linus.walleij@linaro.org>]

On Fri, Feb 5, 2021 at 6:27 PM Daniel Thompson
<daniel.thompson@linaro.org> wrote:
> On Tue, Jan 26, 2021 at 10:32:00PM +0100, Linus Walleij wrote:
> > The KTD253 backlight might already be on when the driver
> > is probed: then we don't really know what the current
> > ratio is and all light intensity settings will be off
> > relative to what it was at boot.
> >
> > To fix this, bring up the backlight OFF then move it to
> > the default backlight from there so we know the state.
> >
> > Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
>
> Hmnnn... horid, since the backlight will flicker during boot,
> but I recall that this bit of hardware is pretty horid anyway
> so it is not easily avoided.
>
> Therefore, slightly grudgingly,
> Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org>

Thanks, and I agree, this makes things better than before though.

The actual problem translates to the generic problem of bringing
misc boot-time conditions over to the kernel, and handing over
any type of hardware in a half-initialized state. We don't really have
a solution to that yet, many have tried.

For this specific case, one can imagine a "kinetic,boot-time-brightness"
(or similar) DT property, with the idea that the boot loader or
DT author knows what to put in there. I put that on my TODO
(the code in this patch will still needed if that property isn't provided).

Yours,
Linus Walleij
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel