From: <bug...@fr...> - 2007-03-01 02:50:31
|
http://bugs.freedesktop.org/show_bug.cgi?id=10024 ------- Comment #55 from z3r...@gm... 2007-02-28 18:50 PST ------- I just tried your patch, and if I disable the scaling that I do manually in the fragment program (MUL tmp.xy, tmp, program.local[0];) it will work just as if I still had the scaling. So basically the driver seems to do the scaling then, but I think this is wrong, with the blob I must manually scale the fragment.position with a MUL. Also, it seems a non-power-of-two adjust is needed for both drivers, so I'm applying that via a MUL now, too. I don't think the driver should do the scaling automatically; yes, it works for my purposes, but I'm not sure it's compliant with the specification. If I only apply the first patch to move the fragment.position calculation into the fragment program, then I can do the scaling/non-power-of-two adjust myself and get almost prefect results. As I showed before, I can get the mirror to work correctly when it's on the floor, however the vertical mirror doesn't seem to work correctly. On the vertical mirror it's almost as if it's looking from behind the mirror, which shouldn't happen. When I apply correct scaling/non-power-of-two adjust, it looks like the front of the mirror is the same as the back of the mirror. Both sides appear to have the back-of-mirror texture applied to them. The back of the mirror does have this texture applied to it, but the front of the mirror only gets this texture as a result of the mirroring. If this makes sense. It's really weird. I can't explain how the texture or the fragment.position would cause it to appear to be looking from behind the mirror. So perhaps this is another non-fragment.position related bug? Also, just letting you know that I will be away from Friday to Sunday, and possibly Monday (local timezone), but I'll try to reply as soon as possible. -- Configure bugmail: http://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee. |