When given (e.g.) 3x 16-bit components to store on a device that isn't using native 16-bit loads and stores, we should be lowering that into one 32-bit store and one masked store. Instead, the logic here ends up returning that the best we can do is one 8-byte store, which is clearly wrong. Stores should round down, loads should round up. Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26293> |
||
|---|---|---|
| .. | ||
| dxcapi.h | ||
| dxil_buffer.c | ||
| dxil_buffer.h | ||
| dxil_buffer_test.c | ||
| dxil_container.c | ||
| dxil_container.h | ||
| dxil_dump.c | ||
| dxil_dump.h | ||
| dxil_dump_decls.h | ||
| dxil_enums.c | ||
| dxil_enums.h | ||
| dxil_function.c | ||
| dxil_function.h | ||
| dxil_internal.h | ||
| dxil_module.c | ||
| dxil_module.h | ||
| dxil_nir.c | ||
| dxil_nir.h | ||
| dxil_nir_algebraic.py | ||
| dxil_nir_lower_int_cubemaps.c | ||
| dxil_nir_lower_int_cubemaps.h | ||
| dxil_nir_lower_int_samplers.c | ||
| dxil_nir_lower_int_samplers.h | ||
| dxil_nir_lower_vs_vertex_conversion.c | ||
| dxil_nir_tess.c | ||
| dxil_signature.c | ||
| dxil_signature.h | ||
| dxil_validator.cpp | ||
| dxil_validator.h | ||
| dxil_versions.h | ||
| meson.build | ||
| nir_to_dxil.c | ||
| nir_to_dxil.h | ||