Quoting Kenneth Graunke (2017-08-02 21:43:33) > I'm not so sure. Table 8.25 of the GL 4.4 compatibility spec (page 257), > below glGetTexImage, shows this: > > +--------------------------------------------+ > | Base Internal Format | R | G | B | A | > +--------------------------------------------+ > | ALPHA | 0 | 0 | 0 | Ai | > | LUMINANCE (or 1) | Li | 0 | 0 | 1 | > | LUMINANCE_ALPHA (or 2) | Li | 0 | 0 | Ai | > | INTENSITY | Ii | 0 | 0 | 1 | > | RED | Ri | 0 | 0 | 1 | > | RG | Ri | Gi | 0 | 1 | > | RGB (or 3) | Ri | Gi | Bi | 1 | > | RGBA (or 4) | Ri | Gi | Bi | Ai | > +--------------------------------------------+ > > which indicates that the current bizarre <R, 0, 0, 1 or A> behavior > might actually be correct for glGetTexImage (?)
That would make a very nice comment above the swizzling in read_rgba_pixels(). formats.csv still has the old swizzling behaviour listed, but I don't know if that's relevant. -Chris _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
