Chen Rabiner wrote:
> Hello guys,
> 
> My name is Chen, and I’m working on linux-opengl compatibility tests for
> embedded design. I am using MESA as reference code.
> 
> As far as i understand opengl require to sample at pixel centers (e.g.
> (0.5,0.5) (1.5, 0,5) and so on...) and check if that exact point lies inside
> a triangle.
> 
> Mesa swrast has had TRIANGLE_WALK_DOUBLE in config.h which allowed working
> with double precision numbers instead of fixed point and snapping
> coordinates . when i turn it on and off i get different results of the final
> picture.

I haven't tried the TRIANGLE_WALK_DOUBLE code in ages and don't know what shape 
it's in.


> my questions are 
> 1. Do both modes (e.g. double and fixed) test if a sampled pixel center is
> inside the triangle, from the calculation i'm not sure. 

Yes.

> 2. what is Mesa swrast tie breaker rule when a sampled point lies on a
> shared edge that is shared by two triangles?

I believe that if edge exactly hits the pixel center then for left/bottom edges 
the pixel is considered "in" and for right/top edges it's considered "out".


> 3. my guess is that fixed point snaps somewhere around the pixel center and
> not the exact center.

Fixed-point snapping could exactly hit a pixel center.


> I know that Mesa doesnt exactly samples, it does some complicated setup and
> than scanline using that setup but if someone can describe in words how this
> correspands to sampling i would be most grateful.

By "samples" do you mean point sampling or multisampling or something else?

The OpenGL spec explains point sampling, if that's what you're after.

-Brian

------------------------------------------------------------------------------
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
_______________________________________________
Mesa3d-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

Reply via email to