Re: [PATCH 0/4] drm/panel: simple: add Waveshare LCD panels

[Dmitry Baryshkov]

On Tue, Mar 31, 2026 at 05:11:43AM +0200, Marek Vasut wrote:
> On 3/31/26 3:22 AM, Dmitry Baryshkov wrote:
> > On Tue, 31 Mar 2026 at 00:07, Marek Vasut <marek.vasut@xxxxxxxxxxx> wrote:
> > > 
> > > On 3/30/26 3:25 PM, Dmitry Baryshkov wrote:
> > > > Waveshare have a serie of DSI panel kits with the DPI or LVDS panel
> > > > being attached to the DSI2DPI or DSI2LVDS bridge. Commit 80b0eb11f8e0
> > > > ("dt-bindings: display: panel: Add waveshare DPI panel support")
> > > > described two of them in the bindings and commit 46be11b678e0
> > > > ("drm/panel: simple: Add Waveshare 13.3" panel support") added
> > > > definitions for one of those panels. Add support for the rest of them.
> > > Can we by any chance use the icn6211 driver in tree for this ?
> > 
> > As far as I can see, no. Waveshare kits have an extra ASIC in front of
> > ICN6211 / ICN6202, which completely hides all programming. So far the
> > interface is really better expressed by the waveshare,dsi2dpi /
> > dsi2lvds: this way, even if they decide to change the actual
> > implementation (like they did for DPI -> LVDS), we won't have to worry
> > about it for as long as their programming interface remains stable.
> Hmmm, I've seen this before, but I don't think this is extra ASIC. The
> ICN6211 pulls its register settings from EEPROM, does it not ?

The kits that I have at hand don't have I2C EEPROMs. They have 25Q8
QSPI. Also note that neither ICN6211 nor ICN6202 datasheets describe I2C
master interface. It is defined as a purely I2C slave.

The kits have WSVTH01 chip (CPLD?) and 25Q8 QSPI NOR. I assume the
bridge settings are loaded from by the WSVTH01 and programmed into the
ICN bridge.

> But no matter how the ICN loads its configuration, this is not waveshare
> specific. I have another display from another manufacturer here which is
> also ICN6211 that is preprogrammed. Can we instead have some generic-dsi
> display , because I don't think these proprogrammed DSI displays are going
> to be only ICN6211 based and only made by waveshare ?

If it was just about the bridge config, I would completely agree with
you. However if you check the waveshare-dsi.c, you'd notice that it is
not directly related to either of ICN bridges. It controls backlight
and panel regulators.

For the reference, *technically* the signal sequence should be:
DSI host -> ICN6202 -> WSVTH01 -> panel. But describing it this way
would require additional changes to the bridge driver to stop it from
programming the bridge (over the DSI). As such I really assume that it's
easier to omit the ICN bridges from the picture.

-- 
With best wishes
Dmitry