2016-03-02 13:44 GMT+01:00 Henri Verbeet hverbeet@gmail.com:
On 2 March 2016 at 09:36, Józef Kucia joseph.kucia@gmail.com wrote:
Not that this is necessarily bad way to implement feature levels, but have you considered exposing wined3d_d3d_level from wined3d?
Oh, good idea. I guess for that to work properly the checks in d3d_level_from_caps() should be consistent with those in shader_glsl_get_caps() though (I'm thinking about the GLSL version "hacks" for SM4 and SM5 in d3d_level_from_caps()), not sure if that's fixable for the time being without regressing the fallbacks in the GPU recognition code. I can certainly give it a try though.
We'll probably want that at some point, yeah. The main issue I see with this patch though is that I think implementing d3d10_device_GetFeatureLevel() / d3d11_device_GetFeatureLevel() on top of this looks awkward.
Eh, good point. I'll fix that.