MLIR  20.0.0git
Public Member Functions | List of all members
VectorizationState Struct Reference

Contains the vectorization state and related methods used across the vectorization process of a given operation. More...

Public Member Functions

 VectorizationState (RewriterBase &rewriter)
 
LogicalResult initState (RewriterBase &rewriter, LinalgOp linalgOp, ArrayRef< int64_t > inputVectorSizes, ArrayRef< bool > inputScalableVecDims)
 Initializes the vectorization state, including the computation of the canonical vector shape for vectorization. More...
 
ArrayRef< int64_t > getCanonicalVecShape () const
 Returns the canonical vector shape used to vectorize the iteration space. More...
 
ArrayRef< bool > getScalableVecDims () const
 Returns the vector dimensions that are scalable in the canonical vector shape. More...
 
VectorType getCanonicalVecType (Type elementType, std::optional< AffineMap > dimPermutation=std::nullopt) const
 Returns a vector type of the provided elementType with the canonical vector shape and the corresponding fixed/scalable dimensions bit. More...
 
OperationmaskOperation (RewriterBase &rewriter, Operation *opToMask, LinalgOp linalgOp, std::optional< AffineMap > maybeMaskingMap=std::nullopt)
 Masks an operation with the canonical vector mask if the operation needs masking. More...
 

Detailed Description

Contains the vectorization state and related methods used across the vectorization process of a given operation.

Definition at line 186 of file Vectorization.cpp.

Constructor & Destructor Documentation

◆ VectorizationState()

VectorizationState::VectorizationState ( RewriterBase rewriter)
inline

Definition at line 187 of file Vectorization.cpp.

Member Function Documentation

◆ getCanonicalVecShape()

ArrayRef<int64_t> VectorizationState::getCanonicalVecShape ( ) const
inline

Returns the canonical vector shape used to vectorize the iteration space.

Definition at line 196 of file Vectorization.cpp.

◆ getCanonicalVecType()

VectorType VectorizationState::getCanonicalVecType ( Type  elementType,
std::optional< AffineMap dimPermutation = std::nullopt 
) const
inline

Returns a vector type of the provided elementType with the canonical vector shape and the corresponding fixed/scalable dimensions bit.

If dimPermutation is provided, the canonical vector dimensions are permuted accordingly.

Definition at line 206 of file Vectorization.cpp.

References mlir::get(), and vectorShape().

◆ getScalableVecDims()

ArrayRef<bool> VectorizationState::getScalableVecDims ( ) const
inline

Returns the vector dimensions that are scalable in the canonical vector shape.

Definition at line 200 of file Vectorization.cpp.

◆ initState()

LogicalResult VectorizationState::initState ( RewriterBase rewriter,
LinalgOp  linalgOp,
ArrayRef< int64_t >  inputVectorSizes,
ArrayRef< bool >  inputScalableVecDims 
)

Initializes the vectorization state, including the computation of the canonical vector shape for vectorization.

Definition at line 313 of file Vectorization.cpp.

References initState(), LDBG, and mlir::OpBuilder::setInsertionPoint().

Referenced by initState().

◆ maskOperation()

Operation * VectorizationState::maskOperation ( RewriterBase rewriter,
Operation opToMask,
LinalgOp  linalgOp,
std::optional< AffineMap maybeMaskingMap = std::nullopt 
)

Masks an operation with the canonical vector mask if the operation needs masking.

Returns the masked operation or the original operation if masking is not needed. If provided, the canonical mask for this operation is permuted using maybeMaskingMap.

Definition at line 430 of file Vectorization.cpp.

References mlir::detail::enumerate(), mlir::Operation::getResults(), LDBG, mlir::vector::maskOperation(), and mlir::RewriterBase::replaceAllUsesExcept().


The documentation for this struct was generated from the following file: