On Fri, 2006-09-22 at 14:06 +0100, Keith Whitwell wrote: > Michel Dänzer wrote: > > On Fri, 2006-09-22 at 11:14 +0100, Keith Whitwell wrote: > >> Maybe enable for the paths that are well understood and tested, but > >> where it is still behaving contrary to expectations, leave it disabled. > >> I'd say it is behaving as expected in the software mesa cases and > >> argb8888. > > > > Our confusion might have to do with the fact I now discovered that the > > hardware drivers only hit argb8888 and a8. :} Software Mesa also hits > > rgb and rgba8888, which the attached patch fixes, so I think all these > > make sense now. *phew* > > So, it looks like all that needs to be done is to extend the same logic > to the bgr888 and rgb888 formats, ie add a test on littleEndian and > adjust the dstMap in each case?
No, these seem to be byte streams. I meant to say 'rgb888' instead of just 'rgb' above, i.e. software Mesa passes glean with this patch for non-half-floats. > > On a possibly related note, software Mesa still seems to fail all > > combinations with type GL_HALF_FLOAT_ARB here, apparently there's an > > endianness issue somewhere in there as well. > > I'm happy to say that's not my fault... :) This is with the xlib driver BTW. I've been poking around in extract_float_rgba() but don't really have a clue yet what's going on. -- Earthling Michel Dänzer | http://tungstengraphics.com Libre software enthusiast | Debian, X and DRI developer ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Mesa3d-dev mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mesa3d-dev
