Re: [PATCH] drm/panel-edp: Add BOE NV140FHD-N43 panel

[Doug Anderson]

Hi,

On Tue, Oct 13, 2026 at 7:49 AM Jun Yan <jerrysteve1101@xxxxxxxxx> wrote:
>
> > > @@ -1972,6 +1972,7 @@ static const struct edp_panel_entry edp_panels[] =
> > =3D {
> > >         EDP_PANEL_ENTRY('A', 'U', 'O', 0xd497, &delay_200_500_e50, "B120X=
> > AN01.0"),
> > >         EDP_PANEL_ENTRY('A', 'U', 'O', 0xf390, &delay_200_500_e50, "B140X=
> > TN07.7"),
> > >
> > > +       EDP_PANEL_ENTRY('B', 'O', 'E', 0x03f4, &delay_200_500_e50, "NV140=
> > FHD-N43"),
> > >         EDP_PANEL_ENTRY('B', 'O', 'E', 0x0607, &delay_200_500_e200, "Unkn=
> > own"),
> > >         EDP_PANEL_ENTRY('B', 'O', 'E', 0x0608, &delay_200_500_e50, "NT116=
> > WHM-N11"),
> > >         EDP_PANEL_ENTRY('B', 'O', 'E', 0x0609, &delay_200_500_e50_po2e200=
> > , "NT116WHM-N21 V4.1"),
> >
> > Things look pretty reasonable, but before landing I'd love a little
> > more background about this panel and where your timings come from. Do
> > you actually have a datasheet here, or are the timings a guess based
> > on nearby panels? This panel is also curious in a few other ways:
> >
>
> Thanks for your review.
>
> Regarding the NV140FHD-N43 panel: I could not find its dedicated datasheet.
> Instead, I referenced the similar NV140FHM-N43 panel [0], which shares the same
> resolution but is 6-bit.

Hard to say how similar a "FHD" vs "FHM" model is. Panel makers
sometimes use very similar names for very different panels. That being
said, we can look at the datasheet you pointed at.


> As I understand it, per section 8.0 POWER SEQUENCE:
>
> - hpd_absent maps to T3 (200ms)
> - unprepare maps to T10 (500ms)
> - enable maps to T17 (20ms).

The above doesn't appear to be correct. See the kernel doc for all
these. Unprepare is "T12-min on eDP timing diagrams.". Enable is
"(T6-min + max(T7-max, T8-min)) on eDP timing diagrams or the delay
after link training finishes until we can turn the backlight on and
see valid data."

You got the right value for unprepare. Enable should be 50, though.


> [0] https://raw.githubusercontent.com/yjun123/e-waste/master/sunniwell-z96a/datasheet/NV140FHM-N43_GZ.pdf
>
> Since delay_200_500_e20 is not defined, I'd avoid adding a new delay entry.
> Reusing delay_200_500_e50 is acceptable and works fine on hardware.

...so "delay_200_500_e50" seems right for the panel you pointed at.


> > 1. ID (0x03f4) is much lower than other panels.
> >
> > 2. The EDID contains the string "STARRY", which is weird for a BOE panel.
> >
>
> These two points are indeed strange and I have no good explanation for them.
> That said, I found a third‑party EDID for this same panel [1].
>
> It contains identical information to the EDID I extracted from the panel, so
> this is not an isolated case.
>
> [1] https://github.com/linuxhw/EDID/blob/9c0c1bffc9c0f1cb2044115149a5ecb1652803f8/Digital/BOE/BOE03F4/C26FE7F0F22E

Fair enough. I'm just going to append a note to your commit message
that the timings are a bit of a guess.

Reviewed-by: Douglas Anderson <dianders@xxxxxxxxxxxx>

Pushed to drm-misc-next:

[1/1] drm/panel-edp: Add BOE NV140FHD-N43 panel
      commit: d81e2537f8082cf416e3de7bcf758191e8ef771b