Re: [Intel-gfx] [PATCH 1/4] drm: Fix color LUT rounding

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Nov 20, 2023 at 01:17:05PM +0000, Borah, Chaitanya Kumar wrote:
> 
> 
> > -----Original Message-----
> > From: Borah, Chaitanya Kumar
> > Sent: Monday, November 20, 2023 6:33 PM
> > To: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx>
> > Cc: intel-gfx@xxxxxxxxxxxxxxxxxxxxx; dri-devel@xxxxxxxxxxxxxxxxxxxxx; Jani
> > Nikula <jani.nikula@xxxxxxxxxxxxxxx>
> > Subject: RE: [Intel-gfx] [PATCH 1/4] drm: Fix color LUT rounding
> > 
> > Hello Ville,
> > 
> > > -----Original Message-----
> > > From: dri-devel <dri-devel-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of
> > > Ville Syrjälä
> > > Sent: Tuesday, October 31, 2023 9:37 PM
> > > To: Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx>
> > > Cc: intel-gfx@xxxxxxxxxxxxxxxxxxxxx; dri-devel@xxxxxxxxxxxxxxxxxxxxx
> > > Subject: Re: [Intel-gfx] [PATCH 1/4] drm: Fix color LUT rounding
> > >
> > > On Tue, Oct 31, 2023 at 11:15:35AM +0200, Jani Nikula wrote:
> > > > On Fri, 13 Oct 2023, Ville Syrjala <ville.syrjala@xxxxxxxxxxxxxxx> wrote:
> > > > > entrirely. But perhaps a better idea would be to follow the OpenGL
> > > > > int<->float conversion rules, in which case we get the following
> > > > > results:
> > > >
> > > > Do you have a pointer to the rules handy, I couldn't find it. :(
> > >
> > > Eg. '2.3.5 Fixed-Point Data Conversions' in GL 4.6 spec. The section
> > > number probably changes depending on which version of the spec you look
> > at.
> > >
> > 
> > This section particularly talks about conversion of normalized fixed point  to
> > floating point numbers and vice versa.
> > Pardon my limited knowledge on the topic but aren't we just doing a scaling
> > factor conversion(Q0.16 -> Q0.8) in these patches?
> > 
> > I could not draw a direct relation between the formulas in the section[1] and
> > what we are doing here.(but it could be just me!)
> 
> Scratch that! As I understand, in effect we are doing a Q0.16 Fixed Point -> Floating point -> Q0.8 Fixed Point conversion.

Yep, that's it.

> Correct me if I am wrong! Otherwise
> 
> LGTM.
> 
> Reviewed-by: Chaitanya Kumar Borah <chaitanya.kumar.borah@xxxxxxxxx>
> 
> 
> > 
> > Regards
> > 
> > Chaitanya
> > 
> > [1] https://registry.khronos.org/OpenGL/specs/gl/glspec46.core.pdf '2.3.5
> > Fixed-Point Data Conversions'
> > 
> > > >
> > > > Might also add the reference to the commit message and/or comment.
> > > >
> > > > BR,
> > > > Jani.
> > > >
> > > > --
> > > > Jani Nikula, Intel
> > >
> > > --
> > > Ville Syrjälä
> > > Intel

-- 
Ville Syrjälä
Intel



[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux