[virglrenderer-devel] regarding ARGB_copy_image and 3 component textures

[Gert Wollny]

Am Dienstag, den 10.07.2018, 08:57 +1000 schrieb Dave Airlie:
> 
> I'm going to have expend some quality time on this 3 component
> is always a pita, and I think there are some things about treating
> them different when texture buffer objects are in use vs when they
> aren't, as in you can store them in 4 components normally, except in
> some TBO cases,
I learned this the hard, way, (e.g. no swizzeling for TBO), but the
good news is that Eriks format cleanup patches actually solve this
problem (the with mesa part of the series). 

I have two piglit regressions with sRGB-linear conversion tests that
IMHO only passed before because sRGB was not treated properly, a bunch
of other tests pass now, and I have another patch in the pipeline that
might fix some more, but I think the bigger problem is that the
GL_DECODE_SRGB_EXT state is not passed properly.  

Some depth-format tests also flip, but it seems that this is a result
of some state not being set/cleared properly in the GL copy fallback,
because when I run the tests first thing after starting the VM they all
pass.

Best, 
Gert