Re: [PATCH RFC] drm/sun4i: rgb: Add 5% tolerance to dot clock frequency check

[Laurent Pinchart]

Hi Sean,

On Thursday 23 Feb 2017 10:54:37 Sean Paul wrote:
> On Wed, Dec 07, 2016 at 11:48:55AM +0200, Laurent Pinchart wrote:
> > On Wednesday 07 Dec 2016 10:26:25 Chen-Yu Tsai wrote:
> >> On Wed, Dec 7, 2016 at 1:29 AM, Maxime Ripard wrote:
> >>> On Thu, Nov 24, 2016 at 07:22:31PM +0800, Chen-Yu Tsai wrote:
> >>>> The panels shipped with Allwinner devices are very "generic", i.e.
> >>>> they do not have model numbers or reliable sources of information
> >>>> for the timings (that we know of) other than the fex files shipped
> >>>> on them. The dot clock frequency provided in the fex files have all
> >>>> been rounded to the nearest MHz, as that is the unit used in them.
> >>>> 
> >>>> We were using the simple panel "urt,umsh-8596md-t" as a substitute
> >>>> for the A13 Q8 tablets in the absence of a specific model for what
> >>>> may be many different but otherwise timing compatible panels. This
> >>>> was usable without any visual artifacts or side effects, until the
> >>>> dot clock rate check was added in commit bb43d40d7c83 ("drm/sun4i:
> >>>> rgb: Validate the clock rate").
> >>>> 
> >>>> The reason this check fails is because the dotclock frequency for
> >>>> this model is 33.26 MHz, which is not achievable with our dot clock
> >>>> hardware, and the rate returned by clk_round_rate deviates slightly,
> >>>> causing the driver to reject the display mode.
> >>>> 
> >>>> The LCD panels have some tolerance on the dot clock frequency, even
> >>>> if it's not specified in their datasheets.
> >>>> 
> >>>> This patch adds a 5% tolerence to the dot clock check.
> >>> 
> >>> As we discussed already, I really believe this is just as arbitrary as
> >>> the current behaviour.
> >> 
> >> Yes. I agree. This patch is mainly to give something that works for
> >> people who don't care about the details, and to get some feedback
> >> from people that do.
> >> 
> >>> Some panels require an exact frequency,
> > 
> > There's no such thing as an exact frequency, there will always be some
> > tolerance (and if your display controller can really generate an exact
> > frequency I'd be very interested in that hardware :-)).
> 
> There is such thing as exact frequency when you have to worry about panel
> warranty.

There's no such thing as exact frequency when you live in a world that has to 
comply with the laws of physics :-) This would require an infinitely precise 
clock, and there's no such thing.

> We are fortunate, since we can talk to the panel manufacturer and discuss
> which frequencies are tolerable inside the warranty. We usually hand pick a
> rate/mode within these constraints.
>
> Also, even though things *look* ok on the panel at a certain clock rate,
> running outside the specified clock could damage the hardware.

I'm curious, how so ? What happens to the panel if you feed it a clock outside 
of that range ?

> I don't think it's unreasonable to add tolerances to drm_panel, since they
> will differ in what is acceptable. The tricky part is teasing out what the
> tolerances are.

More than tricky, in the vast majority of cases we don't have that information 
at all :-/

> > This is something that has been bugging me for some time now. The problem
> > has been mostly ignored, or worked around in different ways by different
> > drivers. I'm afraid I have no generic solution available, but I think we
> > should try to agree on a common behaviour.
> > 
> > I don't believe it would be reasonable to request each panel to report a
> > tolerance, as the value is most of the time not available from the
> > documentation (when documentation is available). Worse, I'm pretty sure
> > that most panels documented as fixed timing can actually accept a wide
> > range of timings. The timings reported in the datasheet are just the
> > nominal values.
> > 
> > Panels that don't support multiple resolutions obviously require fixed
> > active h/v values. Even if they can tolerate some departure from the
> > nominal timings for the sync and porches lengths, it might not be very
> > useful to support that as I don't expect the display controllers and
> > encoders to be a limiting factor by not supporting the particular timings
> > that a panel considers as nominal. On the other hand, departing from the
> > nominal pixel clock frequency is needed as we can't achieve an exact
> > match, and even possibly to have some control over the frame rate
> > (although that might also require changing the sync and porches timings).
> > Without specific information about panel tolerance, do we have any option
> > other than picking an arbitrary value ?
> > 
> >>> some have a minimal frequency
> >>> but no maximum, some have a maximum frequency but no minimal, and I
> >>> guess most of them deviates by how much exactly they can take (and
> >>> possibly can take more easily a higher frequency, but are less
> >>> tolerant if you take a frequency lower than the nominal.
> >>> 
> >>> And we cannot remove that check entirely, since some bridges will
> >>> report out of range frequencies for higher modes that we know we
> >>> cannot reach.
> >> 
> >> I believe this should be handled by the bridge driver in the check
> >> callback? The callback I'm changing is attached to the connector,
> >> which I think doesn't get used if you have a bridge instead.
> >> And this only checks the pre-registered display modes, such as
> >> those specified in simple-panel or EDID.
> >> 
> >>> We could just try to see if the screen pixel clock frequency is out of
> >>> the pixel clock range we can generate, but then we will loop back on
> >>> how much out of range is it exactly, and is it within the screen
> >>> tolerancy.
> >>> 
> >>> We have an API to deal with the panel tolerancies in the DRM panel
> >>> framework, we can (and should) use it.
> >> 
> >> If you mean the get_timings callback, it's not very useful. Most of
> >> the panels in simple-panel do not use the display_timings structure,
> >> so they don't return anything. And I get that. The few datasheets
> >> I found don't list min/max tolerances for the dotclock.
> >> 
> >> The ones that do have the min/max the same as the recommended value.
> >> This may or may not be accurate. IIRC the one panel that had this
> >> that I did check didn't list min/max values in its datasheet.
> >> 
> >>> I'm not sure how others usually deal with this though. I think I
> >>> remember Eric telling me that for the RPi they just adjusted the
> >>> timings a bit, but they only really had a single panel to deal with.
> >>> 
> >>> Daniel, Eric, Laurent, Sean? Any ideas?
> >> 
> >> Yes! Feedback please! Between Maxime and me I think we only have a
> >> limited number of panels, with some overlap.

-- 
Regards,

Laurent Pinchart