On Wed, 2003-07-09 at 16:10, Michel Dänzer wrote: > On Wed, 2003-07-09 at 14:11, Martin Spott wrote: > > Rupert Levene <[EMAIL PROTECTED]> wrote: > > > On Thu, Jul 03, 2003 at 01:07:28AM +0200, Michel Dnzer wrote: > > > > >> PS: http://levene.dyndns.org/invisible-text/normal.png does show a > > >> driver problem - the lighting is wrong with HW TCL - but you didn't > > >> report that... ;) > > > > > Well I'd assumed that _this_ was a software problem since it's a new > > > in the latest version =) Is this a known driver problem? > > Yes. HW TCL seems to be severely broken in general in the Radeon > drivers, which can be worked around by setting the R200_NO_TCL (or > RADEON_TCL_FORCE_DISABLE) environment variable. > > > If the scene looks to be too dark, then you can assume it's a driver > > problem. > > Why don't you just take a look at the screenshot? The red tiles aren't > supposed to be red. This particular problem seems to be caused (or at > least amplified) by many light sources.
I see this kind of thing with my own 3D software (see http://qdn.qubesoft.com) running on the r200 dri driver. I have a feeling that its something to do with modifying light parameters. In my 3D scenes, we render large numbers of objects, potentially each with different lighting configurations, depending on which light objects are in range of a given object. We call glLightX to modify the parameters of the GL lights many times in a single frame. This appears to be the main problem with the r200 rendering errors. FWIW, my software works very nicely with nvidia hardware/drivers. ------------------------------------------------------- This SF.Net email sponsored by: Parasoft Error proof Web apps, automate testing & more. Download & eval WebKing and get a free book. www.parasoft.com/bulletproofapps1 _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
