[RFC] drm/panel/simple-edp: Add Samsung ATNA45DC02

[Rob Clark]

On Thu, Jul 18, 2024 at 9:45 AM Rob Clark <robdclark at gmail.com> wrote:
>
> On Thu, Jul 18, 2024 at 9:31 AM Doug Anderson <dianders at chromium.org> wrote:
> >
> > Hi,
> >
> > On Thu, Jul 18, 2024 at 9:25 AM Rob Clark <robdclark at gmail.com> wrote:
> > >
> > > On Thu, Jul 18, 2024 at 9:00 AM Doug Anderson <dianders at chromium.org> wrote:
> > > >
> > > > Hi,
> > > >
> > > > On Wed, Jul 17, 2024 at 6:09 PM Rob Clark <robdclark at gmail.com> wrote:
> > > > >
> > > > > On Wed, Jul 17, 2024 at 5:19 PM Doug Anderson <dianders at chromium.org> wrote:
> > > > > >
> > > > > > Hi,
> > > > > >
> > > > > > On Wed, Jul 17, 2024 at 2:58 PM Rob Clark <robdclark at gmail.com> wrote:
> > > > > > >
> > > > > > > From: Rob Clark <robdclark at chromium.org>
> > > > > > >
> > > > > > > Just a guess on the panel timings, but they appear to work.
> > > > > > >
> > > > > > > Signed-off-by: Rob Clark <robdclark at chromium.org>
> > > > > > > ---
> > > > > > > This adds the panel I have on my lenovo yoga slim 7x laptop.  I couldn't
> > > > > > > find any datasheet so timings is just a guess.  But AFAICT everything
> > > > > > > works fine.
> > > > > > >
> > > > > > >  drivers/gpu/drm/panel/panel-edp.c | 2 ++
> > > > > > >  1 file changed, 2 insertions(+)
> > > > > >
> > > > > > Given that this is a Samsung ATNA<mumble>, is there any chance it's an
> > > > > > OLED panel? Should it be supported with the Samsung OLED panel driver
> > > > > > like this:
> > > > > >
> > > > > > https://lore.kernel.org/r/20240715-x1e80100-crd-backlight-v2-0-31b7f2f658a3@linaro.org
> > > > >
> > > > > it is an OLED panel, and I did see that patchset and thought that
> > > > > samsung panel driver would work.  But changing the compat string on
> > > > > the yoga dts only gave me a black screen (and I didn't see any of the
> > > > > other mentioned problems with bl control or dpms with panel-edp).  So
> > > > > :shrug:?  It could be I overlooked something else, but it _seems_ like
> > > > > panel-edp is fine for this panel. Plus, it would avoid awkwardness if
> > > > > it turned out there were other non-samsung 2nd sources, but so far
> > > > > with a sample size of two 100% of these laptops have the same panel
> > > >
> > > > Hmm, OK. One question for you: are you using the "enable" GPIO in
> > > > panel-edp? IMO the code handling that GPIO in panel-edp is somewhat
> > > > dodgy, but it predates my deeper involvement with panels. I've never
> > > > seen an eDP panel that could use panel-edp where it was actually
> > > > required--every instance where someone thought it was required was
> > > > better modeled by using that GPIO as the backlight enable. On the
> > > > other hand, the "enable" GPIO in the Samsung OLED panel driver came
> > > > straight from the panel datasheet and it makes sense for it to be in
> > > > the panel driver since the backlight is handled straight by the
> > > > panel...
> > >
> > > hmm, at least current dts doesn't have an enable gpio.  Which could be
> > > why panel-samsung-atna33xc20 wasn't working.
> > >
> > > It is entirely possible we are relying on something left on by the bootloader.
> >
> > That would be my best guess. Is there any way for you to find out if
> > there's an enable GPIO?
>
> presumably it would be in the ACPI dump (which I've not much of a clue
> what I'm looking for there)
>

Ok, turns out it is wired up the same way as on the crd.. if I copy
those then panel-samsung-atna33xc20 is working

BR,
-R

> >
> > > > In any case, I guess if things are working it doesn't really hurt to
> > > > take it in panel-edp for now...
> > > >
> > >
> > > I wonder if using compatible="edp-panel" everywhere isn't a great idea
> > > if we want to switch drivers later.  But I guess that is already water
> > > under the bridge (so to speak)
> >
> > For panels that aren't OLED it's all very standard and we're kinda
> > forced to use something generic since manufacturers want lots of 2nd
> > (and 3rd and 4th and ...) sourcing. As far as I've been able to tell
> > you can't do 2nd sourcing between OLED panels and other panels since
> > the wires hooked up to the panels are a little different for the OLED
> > panels and the power sequencing is a bit different. It would also be
> > pretty obvious to an end user if some of their devices had an OLED
> > panel and some didn't. I'm not aware of OLED panels other than the
> > Samsung ones, but I haven't done any real research here...
> >
> > -Doug
> >