maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/mlir/lib/Conversion/ArmNeon2dToIntr/CMakeLists.txt

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

21 lines
324 B
CMake
Raw Normal View History

2d Arm Neon sdot op, and lowering to the intrinsic. This adds Sdot2d op, which is similar to the usual Neon intrinsic except that it takes 2d vector operands, reflecting the structure of the arithmetic that it's performing: 4 separate 4-dimensional dot products, whence the vector<4x4xi8> shape. This also adds a new pass, arm-neon-2d-to-intr, lowering this new 2d op to the 1d intrinsic. Reviewed By: nicolasvasilache Differential Revision: https://reviews.llvm.org/D102504
2021-06-11 04:57:44 +08:00
add_mlir_conversion_library(MLIRArmNeon2dToIntr
ArmNeon2dToIntr.cpp
ADDITIONAL_HEADER_DIRS
${MLIR_MAIN_INCLUDE_DIR}/mlir/Conversion/ArmNeon2dToIntr
DEPENDS
MLIRConversionPassIncGen
LINK_COMPONENTS
Core
LINK_LIBS PUBLIC
[mlir] Add more missing dependencies after D118062 These used to be covered transitively, but now need to be explicit.
2022-01-25 14:11:53 +08:00
MLIRArithmetic
2d Arm Neon sdot op, and lowering to the intrinsic. This adds Sdot2d op, which is similar to the usual Neon intrinsic except that it takes 2d vector operands, reflecting the structure of the arithmetic that it's performing: 4 separate 4-dimensional dot products, whence the vector<4x4xi8> shape. This also adds a new pass, arm-neon-2d-to-intr, lowering this new 2d op to the 1d intrinsic. Reviewed By: nicolasvasilache Differential Revision: https://reviews.llvm.org/D102504
2021-06-11 04:57:44 +08:00
MLIRArmNeon
MLIRPass
MLIRTransforms
MLIRIR
[mlir] Add more missing dependencies after D118062 These used to be covered transitively, but now need to be explicit.
2022-01-25 14:11:53 +08:00
MLIRVector
2d Arm Neon sdot op, and lowering to the intrinsic. This adds Sdot2d op, which is similar to the usual Neon intrinsic except that it takes 2d vector operands, reflecting the structure of the arithmetic that it's performing: 4 separate 4-dimensional dot products, whence the vector<4x4xi8> shape. This also adds a new pass, arm-neon-2d-to-intr, lowering this new 2d op to the 1d intrinsic. Reviewed By: nicolasvasilache Differential Revision: https://reviews.llvm.org/D102504
2021-06-11 04:57:44 +08:00
)