https://bugs.winehq.org/show_bug.cgi?id=46304
--- Comment #2 from Henri Verbeet hverbeet@gmail.com --- (In reply to Gert from comment #0)
I have not yet investigated why wine creates different code path for the two drivers, I guess that it depends on some extension(s) that may or may not be available.
I'm mostly on vacation, so haven't really looked at this, but part of the difference in the generated code is due to ARB_clip_control being available or not. (I.e., see the condition on the shader_glsl_fixup_position() call in shader_glsl_generate_vs_epilogue().)
Are you sure those are from the same D3D source shader though? The code is very similar, but reordering e.g. vs_c[20] vs. vs_c[16] is not something wined3d would do.
This also sounds suspiciously similar to bug 17423, any chance it's the same issue?