2 Jun
2023
2 Jun
'23
9:52 a.m.
On Fri Jun 2 09:46:31 2023 +0000, Nikolay Sivov wrote:
Turns out 1D case works differently. It's using 2d resource for all profiles, using (x/w, 0.5) as a coordinate in SM4+, I think this justifies moving profile-specific logic out of codegen.c, back to the function.
Another option is to retain original 1D dim value and carry it all the way to lower_tex_proj(), but it will produce awkward combination of dim and coordinates types that is resolved too late and for each binary format, when writing declarations.