llvm-project
[HLSL][DirectX] Implement HLSL `mul` function and DXIL lowering of `llvm.matrix.multiply`
#184882
Merged
Merged
[HLSL][DirectX] Implement HLSL `mul` function and DXIL lowering of `llvm.matrix.multiply` #184882
Icohedron
merged
16 commits into
llvm:main
from
Icohedron:mul
Implement HLSL mul function
2491daa6
farzonl
commented
on 2026-03-05
Split mul into header and builtin implementations
4a105410
Remove redundant mul definitions. Make mul documentation comment cons…
41612f30
Icohedron
changed the title 
Merge remote-tracking branch 'upstream/main' into mul
9613b6d9
Relax CHECKs for double vector x double vector mul test
64f025eb
farzonl
approved these changes
on 2026-03-05
Icohedron
commented
on 2026-03-05
Make vector x vector case be implemented by the mul builtin
04e744f4
Apply clang-format
904d3079
Relax CHECKs for double vector x double vector mul codegen test
f0c2d715
farzonl
commented
on 2026-03-06
Revert "Relax CHECKs for double vector x double vector mul codegen test"
3e2700bf
Revert "Apply clang-format"
90320457
Revert "Make vector x vector case be implemented by the mul builtin"
f4e14d37
Use __builtin_hlsl_mad for double vector-vector mul
6a8e3bd3
Remove TODO for row-major indexing since it is no longer planned in t…
e86abe39
Adjust comments and logic of mul to be more clear
93b3cc32
farzonl
commented
on 2026-03-06
Use mad instead of builtin_hlsl_mad
0c988039
Use fmuladd when lowering double matrix multiply
fb067300
s-perron
approved these changes
on 2026-03-06
Login to write a write a comment.
Login via GitHub