Hi Chris!

Am Sonntag 15 Juni 2008 14:01:09 schrieb Chris Rankin:
> Since R300 now supports GL_ARB_shadow I thought I'd try running the
> shadowtex program. When I start the program, all I see is a black window.
> However, I can then cycle through the various comparison modes to get the
> following results:
>
> GL_LESS: blank
> GL_GEQUAL: coloured shapes
> GL_GREATER: coloured shapes
> GL_EQUAL: blank
> GL_NOTEQUAL: blank
> GL_ALWAYS: coloured shapes
> GL_NEVER: blank
> GL_LEQUAL: blank
>
> Are these correct, please? I would have thought that the program would at
> least *start* in a mode that generated some output.

This is a known problem, and it's on my todo list.

So far I have only been able to do a preliminary analysis, which points to an 
incorrect "distance from light source" calculation (as opposed to a problem 
in the ARB_shadow implementation); my evidence for this is:
1. Check out the 'd' mode of shadowtex; the values appear to be smaller than 
they should be (by comparing to software rendered Mesa).
2. Change the Z bias ('b'/'B') by an insane amount, and things will start to 
become visible.
So it appears that some other part of the pipeline is at fault here (perhaps 
it's texgen, but glean/texgen passes), or perhaps some problem with how the 
different parts interact.

Once I understand the problem I will isolate it into a Piglit test and then 
produce a fix, as soon as my next "go into R300 hacking mode for 24+ hours" 
cycle starts.

cu,
Nicolai

> Cycling though "fragment program modes" gives nothing except a blank window
> and the following message:
>
> Using fragment program without "OPTION ARB_fragment_program_shadow"
>
> Cheers,
> Chris

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
Mesa3d-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev

Reply via email to