Re: drm/bridge/imx8mp-hdmi-tx: Allow inexact pixel clock frequencies (Was: [PATCH V8 10/12] drm/bridge: imx: add bridge wrapper driver for i.MX8MP DWC HDMI)

[Dominique MARTINET <dominique.martinet@atmark-techno.com>]

Frieder Schrempf wrote on Tue, Aug 27, 2024 at 09:00:34AM +0200:
> > Using my PMS calculator and Rev 2 of the Tech Ref Manual, I think I
> > can generate you a table entry for 51.2MHz.  I don't have the ability
> > to test it.  I am still working on figuring out an algorithm to
> > calculate the fractional divider, but 51.2MHz x 5 does not' appear to
> > need the fractional clock divider, so I think I can just get away with
> > the standard PMS calculations.
> 
> Thanks Adam for all the efforts. The initial request for the 51.2 MHz
> clock was from Dominique, not me. I was just jumping onto the bandwagon
> with my own use-cases.

We still have that 51.2 MHz display around, so happy to test -- might
take a little bit of time as this already has a workaround but it's
definitely something I'd like to clean up so more than happy to help.

> > I still think it's a good idea to fall back to the PMS calculator if
> > the fractional stuff isn't found.  We could then determine which clock
> > frequency is closer between the nearest table entry or the PMS
> > calculator until someone can come up with an algorithm for it.
> 
> Yes, I think that's a good idea.

Seconded; we can keep the code as fallback of the table at first and if
it works well maybe even remove the table altogether later, but just
having fallback first would be great.

-- 
Dominique