MLIR  15.0.0git
Public Member Functions | List of all members
TransposeOpLowering Class Reference

Rewrite AVX2-specific vector.transpose, for the supported cases and depending on the TransposeLoweringOptions. More...

+ Inheritance diagram for TransposeOpLowering:
+ Collaboration diagram for TransposeOpLowering:

Public Member Functions

 TransposeOpLowering (LoweringOptions loweringOptions, MLIRContext *context, int benefit)
 
LogicalResult matchAndRewrite (vector::TransposeOp op, PatternRewriter &rewriter) const override
 
- Public Member Functions inherited from mlir::OpRewritePattern< vector::TransposeOp >
 OpRewritePattern (MLIRContext *context, PatternBenefit benefit=1, ArrayRef< StringRef > generatedNames={})
 Patterns must specify the root operation name they match against, and can also specify the benefit of the pattern matching and a list of generated ops. More...
 
- Public Member Functions inherited from mlir::detail::OpOrInterfaceRewritePatternBase< vector::TransposeOp >
void rewrite (Operation *op, PatternRewriter &rewriter) const final
 Wrappers around the RewritePattern methods that pass the derived op type. More...
 
virtual void rewrite (vector::TransposeOp op, PatternRewriter &rewriter) const
 Rewrite and Match methods that operate on the SourceOp type. More...
 
LogicalResult match (Operation *op) const final
 Attempt to match against code rooted at the specified operation, which is the same operation code as getRootKind(). More...
 
virtual LogicalResult match (vector::TransposeOp op) const
 
LogicalResult matchAndRewrite (Operation *op, PatternRewriter &rewriter) const final
 Attempt to match against code rooted at the specified operation, which is the same operation code as getRootKind(). More...
 
- Public Member Functions inherited from mlir::RewritePattern
virtual ~RewritePattern ()=default
 
- Public Member Functions inherited from mlir::Pattern
ArrayRef< OperationNamegetGeneratedOps () const
 Return a list of operations that may be generated when rewriting an operation instance with this pattern. More...
 
Optional< OperationNamegetRootKind () const
 Return the root node that this pattern matches. More...
 
Optional< TypeIDgetRootInterfaceID () const
 Return the interface ID used to match the root operation of this pattern. More...
 
Optional< TypeIDgetRootTraitID () const
 Return the trait ID used to match the root operation of this pattern. More...
 
PatternBenefit getBenefit () const
 Return the benefit (the inverse of "cost") of matching this pattern. More...
 
bool hasBoundedRewriteRecursion () const
 Returns true if this pattern is known to result in recursive application, i.e. More...
 
MLIRContextgetContext () const
 Return the MLIRContext used to create this pattern. More...
 
StringRef getDebugName () const
 Return a readable name for this pattern. More...
 
void setDebugName (StringRef name)
 Set the human readable debug name used for this pattern. More...
 
ArrayRef< StringRef > getDebugLabels () const
 Return the set of debug labels attached to this pattern. More...
 
void addDebugLabels (ArrayRef< StringRef > labels)
 Add the provided debug labels to this pattern. More...
 
void addDebugLabels (StringRef label)
 
- Public Member Functions inherited from mlir::ConvertOpToLLVMPattern< memref::TransposeOp >
 ConvertOpToLLVMPattern (LLVMTypeConverter &typeConverter, PatternBenefit benefit=1)
 
void rewrite (Operation *op, ArrayRef< Value > operands, ConversionPatternRewriter &rewriter) const final
 Wrappers around the RewritePattern methods that pass the derived op type. More...
 
virtual void rewrite (memref::TransposeOp op, OpAdaptor adaptor, ConversionPatternRewriter &rewriter) const
 
LogicalResult match (Operation *op) const final
 Attempt to match against code rooted at the specified operation, which is the same operation code as getRootKind(). More...
 
virtual LogicalResult match (memref::TransposeOp op) const
 Rewrite and Match methods that operate on the SourceOp type. More...
 
LogicalResult matchAndRewrite (Operation *op, ArrayRef< Value > operands, ConversionPatternRewriter &rewriter) const final
 Hook for derived classes to implement combined matching and rewriting. More...
 
virtual LogicalResult matchAndRewrite (memref::TransposeOp op, OpAdaptor adaptor, ConversionPatternRewriter &rewriter) const
 
- Public Member Functions inherited from mlir::ConvertToLLVMPattern
 ConvertToLLVMPattern (StringRef rootOpName, MLIRContext *context, LLVMTypeConverter &typeConverter, PatternBenefit benefit=1)
 
- Public Member Functions inherited from mlir::ConversionPattern
LogicalResult matchAndRewrite (Operation *op, PatternRewriter &rewriter) const final
 Attempt to match and rewrite the IR root at the specified operation. More...
 
TypeConvertergetTypeConverter () const
 Return the type converter held by this pattern, or nullptr if the pattern does not require type conversion. More...
 
template<typename ConverterTy >
std::enable_if_t< std::is_base_of< TypeConverter, ConverterTy >::value, ConverterTy * > getTypeConverter () const
 

Additional Inherited Members

- Public Types inherited from mlir::ConvertOpToLLVMPattern< memref::TransposeOp >
using OpAdaptor = typename memref::TransposeOp ::Adaptor
 
- Static Public Member Functions inherited from mlir::RewritePattern
template<typename T , typename... Args>
static std::unique_ptr< T > create (Args &&... args)
 This method provides a convenient interface for creating and initializing derived rewrite patterns of the given type T. More...
 
- Protected Member Functions inherited from mlir::Pattern
 Pattern (StringRef rootName, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={})
 Construct a pattern with a certain benefit that matches the operation with the given root name. More...
 
 Pattern (MatchAnyOpTypeTag tag, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={})
 Construct a pattern that may match any operation type. More...
 
 Pattern (MatchInterfaceOpTypeTag tag, TypeID interfaceID, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={})
 Construct a pattern that may match any operation that implements the interface defined by the provided interfaceID. More...
 
 Pattern (MatchTraitOpTypeTag tag, TypeID traitID, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={})
 Construct a pattern that may match any operation that implements the trait defined by the provided traitID. More...
 
void setHasBoundedRewriteRecursion (bool hasBoundedRecursionArg=true)
 Set the flag detailing if this pattern has bounded rewrite recursion or not. More...
 
- Protected Member Functions inherited from mlir::ConvertToLLVMPattern
LLVM::LLVMDialect & getDialect () const
 Returns the LLVM dialect. More...
 
LLVMTypeConvertergetTypeConverter () const
 
Type getIndexType () const
 Gets the MLIR type wrapping the LLVM integer type whose bit width is defined by the used type converter. More...
 
Type getIntPtrType (unsigned addressSpace=0) const
 Gets the MLIR type wrapping the LLVM integer type whose bit width corresponds to that of a LLVM pointer type. More...
 
Type getVoidType () const
 Gets the MLIR type wrapping the LLVM void type. More...
 
Type getVoidPtrType () const
 Get the MLIR type wrapping the LLVM i8* type. More...
 
Value createIndexConstant (ConversionPatternRewriter &builder, Location loc, uint64_t value) const
 Create an LLVM dialect operation defining the given index constant. More...
 
Value getStridedElementPtr (Location loc, MemRefType type, Value memRefDesc, ValueRange indices, ConversionPatternRewriter &rewriter) const
 
bool isConvertibleAndHasIdentityMaps (MemRefType type) const
 Returns if the given memref has identity maps and the element type is convertible to LLVM. More...
 
Type getElementPtrType (MemRefType type) const
 Returns the type of a pointer to an element of the memref. More...
 
void getMemRefDescriptorSizes (Location loc, MemRefType memRefType, ValueRange dynamicSizes, ConversionPatternRewriter &rewriter, SmallVectorImpl< Value > &sizes, SmallVectorImpl< Value > &strides, Value &sizeBytes) const
 Computes sizes, strides and buffer size in bytes of memRefType with identity layout. More...
 
Value getSizeInBytes (Location loc, Type type, ConversionPatternRewriter &rewriter) const
 Computes the size of type in bytes. More...
 
Value getNumElements (Location loc, ArrayRef< Value > shape, ConversionPatternRewriter &rewriter) const
 Computes total number of elements for the given shape. More...
 
MemRefDescriptor createMemRefDescriptor (Location loc, MemRefType memRefType, Value allocatedPtr, Value alignedPtr, ArrayRef< Value > sizes, ArrayRef< Value > strides, ConversionPatternRewriter &rewriter) const
 Creates and populates a canonical memref descriptor struct. More...
 
LogicalResult copyUnrankedDescriptors (OpBuilder &builder, Location loc, TypeRange origTypes, SmallVectorImpl< Value > &operands, bool toDynamic) const
 Copies the memory descriptor for any operands that were unranked descriptors originally to heap-allocated memory (if toDynamic is true) or to stack-allocated memory (otherwise). More...
 
- Protected Member Functions inherited from mlir::ConversionPattern
template<typename... Args>
 ConversionPattern (TypeConverter &typeConverter, Args &&...args)
 Construct a conversion pattern with the given converter, and forward the remaining arguments to RewritePattern. More...
 
- Static Protected Member Functions inherited from mlir::ConvertToLLVMPattern
static Value createIndexAttrConstant (OpBuilder &builder, Location loc, Type resultType, int64_t value)
 Create a constant Op producing a value of resultType from an index-typed integer attribute. More...
 
- Protected Attributes inherited from mlir::ConversionPattern
TypeConvertertypeConverter = nullptr
 An optional type converter for use by this pattern. More...
 

Detailed Description

Rewrite AVX2-specific vector.transpose, for the supported cases and depending on the TransposeLoweringOptions.

The lowering supports 2-D transpose cases and n-D cases that have been decomposed into 2-D transposition slices. For example, a 3-D transpose:

%0 = vector.transpose arg0, [2, 0, 1] : vector<1024x2048x4096xf32> to vector<4096x1024x2048xf32>

could be sliced into 2-D transposes by tiling two of its dimensions to one of the vector lengths supported by the AVX2 patterns (e.g., 4x8):

%0 = vector.transpose arg0, [2, 0, 1] : vector<1x4x8xf32> to vector<8x1x4xf32>

This lowering will analyze the n-D vector.transpose and determine if it's a supported 2-D transposition slice where any of the AVX2 patterns can be applied.

Definition at line 240 of file AVXTranspose.cpp.

Constructor & Destructor Documentation

◆ TransposeOpLowering()

TransposeOpLowering::TransposeOpLowering ( LoweringOptions  loweringOptions,
MLIRContext context,
int  benefit 
)
inline

Definition at line 244 of file AVXTranspose.cpp.

Member Function Documentation

◆ matchAndRewrite()

LogicalResult TransposeOpLowering::matchAndRewrite ( vector::TransposeOp  op,
PatternRewriter rewriter 
) const
inlineoverridevirtual

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