Lowering for memory allocation ops. More...
#include "mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h"
Inheritance diagram for mlir::AllocationOpLLVMLowering:
Collaboration diagram for mlir::AllocationOpLLVMLowering:
Public Member Functions | |
| AllocationOpLLVMLowering (StringRef opName, LLVMTypeConverter &converter) | |
| Value | createIndexConstant (ConversionPatternRewriter &builder, Location loc, uint64_t value) const |
| Create an LLVM dialect operation defining the given index constant. More... | |
| Type | getIndexType () const |
| Gets the MLIR type wrapping the LLVM integer type whose bit width is defined by the used type converter. More... | |
| Type | getVoidPtrType () const |
| Get the MLIR type wrapping the LLVM i8* type. More... | |
| Public Member Functions inherited from mlir::ConvertToLLVMPattern | |
| ConvertToLLVMPattern (StringRef rootOpName, MLIRContext *context, LLVMTypeConverter &typeConverter, PatternBenefit benefit=1) | |
| Public Member Functions inherited from mlir::ConversionPattern | |
| virtual void | rewrite (Operation *op, ArrayRef< Value > operands, ConversionPatternRewriter &rewriter) const |
| Hook for derived classes to implement rewriting. More... | |
| virtual LogicalResult | matchAndRewrite (Operation *op, ArrayRef< Value > operands, ConversionPatternRewriter &rewriter) const |
| Hook for derived classes to implement combined matching and rewriting. More... | |
| LogicalResult | matchAndRewrite (Operation *op, PatternRewriter &rewriter) const final |
| Attempt to match and rewrite the IR root at the specified operation. More... | |
| TypeConverter * | getTypeConverter () 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 |
| Public Member Functions inherited from mlir::RewritePattern | |
| virtual | ~RewritePattern ()=default |
| virtual void | rewrite (Operation *op, PatternRewriter &rewriter) const |
| Rewrite the IR rooted at the specified operation with the result of this pattern, generating any new operations with the specified builder. More... | |
| virtual LogicalResult | match (Operation *op) const |
| 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::Pattern | |
| ArrayRef< OperationName > | getGeneratedOps () const |
| Return a list of operations that may be generated when rewriting an operation instance with this pattern. More... | |
| std::optional< OperationName > | getRootKind () const |
| Return the root node that this pattern matches. More... | |
| std::optional< TypeID > | getRootInterfaceID () const |
| Return the interface ID used to match the root operation of this pattern. More... | |
| std::optional< TypeID > | getRootTraitID () 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... | |
| MLIRContext * | getContext () 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) |
Protected Member Functions | |
| template<typename OpType > | |
| Value | getAlignment (ConversionPatternRewriter &rewriter, Location loc, OpType op) const |
| Computes the alignment for the given memory allocation op. More... | |
| template<typename OpType > | |
| int64_t | alignedAllocationGetAlignment (ConversionPatternRewriter &rewriter, Location loc, OpType op, const DataLayout *defaultLayout) const |
| Computes the alignment for aligned_alloc used to allocate the buffer for the memory allocation op. More... | |
| std::tuple< Value, Value > | allocateBufferManuallyAlign (ConversionPatternRewriter &rewriter, Location loc, Value sizeBytes, Operation *op, Value alignment) const |
| Allocates a memory buffer using an allocation method that doesn't guarantee alignment. More... | |
| Value | allocateBufferAutoAlign (ConversionPatternRewriter &rewriter, Location loc, Value sizeBytes, Operation *op, const DataLayout *defaultLayout, int64_t alignment) const |
| Allocates a memory buffer using an aligned allocation method. More... | |
| Protected Member Functions inherited from mlir::ConvertToLLVMPattern | |
| LLVM::LLVMDialect & | getDialect () const |
| Returns the LLVM dialect. More... | |
| LLVMTypeConverter * | getTypeConverter () 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... | |
| Protected Member Functions inherited from mlir::RewritePattern | |
| Pattern (StringRef rootName, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={}) | |
Inherit the base constructors from Pattern. More... | |
| Pattern (MatchAnyOpTypeTag tag, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={}) | |
Inherit the base constructors from Pattern. More... | |
| Pattern (MatchInterfaceOpTypeTag tag, TypeID interfaceID, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={}) | |
Inherit the base constructors from Pattern. More... | |
| Pattern (MatchTraitOpTypeTag tag, TypeID traitID, PatternBenefit benefit, MLIRContext *context, ArrayRef< StringRef > generatedNames={}) | |
Inherit the base constructors from Pattern. 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... | |
Static Protected Member Functions | |
| static Value | createAligned (ConversionPatternRewriter &rewriter, Location loc, Value input, Value alignment) |
| Computes the aligned value for 'input' as follows: bumped = input + alignement - 1 aligned = bumped - bumped % alignment. More... | |
| static MemRefType | getMemRefResultType (Operation *op) |
| 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... | |
Additional Inherited Members | |
| 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 Attributes inherited from mlir::ConversionPattern | |
| TypeConverter * | typeConverter = nullptr |
| An optional type converter for use by this pattern. More... | |
Detailed Description
Lowering for memory allocation ops.
Definition at line 17 of file AllocLikeConversion.h.
Constructor & Destructor Documentation
◆ AllocationOpLLVMLowering()
|
inlineexplicit |
Definition at line 22 of file AllocLikeConversion.h.
Member Function Documentation
◆ alignedAllocationGetAlignment()
template<typename OpType >
|
inlineprotected |
Computes the alignment for aligned_alloc used to allocate the buffer for the memory allocation op.
Aligned_alloc requires the allocation size to be a power of two, and the allocation size to be a multiple of the alignment.
Definition at line 61 of file AllocLikeConversion.h.
References max().
◆ allocateBufferAutoAlign()
|
protected |
Allocates a memory buffer using an aligned allocation method.
Definition at line 113 of file AllocLikeConversion.cpp.
References mlir::OpBuilder::create(), createAligned(), createIndexConstant(), mlir::ConvertToLLVMPattern::getElementPtrType(), getIndexType(), getMemRefResultType(), mlir::Operation::getParentOfType(), and mlir::ConvertToLLVMPattern::getTypeConverter().
◆ allocateBufferManuallyAlign()
|
protected |
Allocates a memory buffer using an allocation method that doesn't guarantee alignment.
Returns the pointer and its aligned value.
Definition at line 53 of file AllocLikeConversion.cpp.
References mlir::OpBuilder::create(), createAligned(), mlir::ConvertToLLVMPattern::getElementPtrType(), getIndexType(), getMemRefResultType(), mlir::Operation::getParentOfType(), and mlir::ConvertToLLVMPattern::getTypeConverter().
◆ createAligned()
|
staticprotected |
Computes the aligned value for 'input' as follows: bumped = input + alignement - 1 aligned = bumped - bumped % alignment.
Definition at line 43 of file AllocLikeConversion.cpp.
References mlir::OpBuilder::create(), mlir::ConvertToLLVMPattern::createIndexAttrConstant(), mlir::Value::getType(), and mlir::mod().
Referenced by allocateBufferAutoAlign(), and allocateBufferManuallyAlign().
◆ createIndexConstant()
| Value ConvertToLLVMPattern::createIndexConstant |
Create an LLVM dialect operation defining the given index constant.
Definition at line 69 of file Pattern.cpp.
References mlir::MemRefDescriptor::alignedPtr(), mlir::OpBuilder::create(), mlir::ConvertToLLVMPattern::createIndexConstant(), mlir::MemRefDescriptor::getElementPtrType(), mlir::getStridesAndOffset(), mlir::MemRefDescriptor::offset(), and mlir::MemRefDescriptor::stride().
Referenced by allocateBufferAutoAlign(), and getAlignment().
◆ getAlignment()
template<typename OpType >
|
inlineprotected |
Computes the alignment for the given memory allocation op.
Definition at line 39 of file AllocLikeConversion.h.
References createIndexConstant(), mlir::ConvertToLLVMPattern::getSizeInBytes(), and mlir::Value::getType().
◆ getIndexType()
| Type ConvertToLLVMPattern::getIndexType |
Gets the MLIR type wrapping the LLVM integer type whose bit width is defined by the used type converter.
Definition at line 51 of file Pattern.cpp.
Referenced by allocateBufferAutoAlign(), and allocateBufferManuallyAlign().
◆ getMemRefResultType()
|
inlinestaticprotected |
Definition at line 33 of file AllocLikeConversion.h.
References mlir::Type::cast(), mlir::Operation::getResult(), and mlir::Value::getType().
Referenced by allocateBufferAutoAlign(), and allocateBufferManuallyAlign().
◆ getVoidPtrType()
| Type ConvertToLLVMPattern::getVoidPtrType |
Get the MLIR type wrapping the LLVM i8* type.
Definition at line 61 of file Pattern.cpp.
The documentation for this struct was generated from the following files:
- include/mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h
- lib/Conversion/MemRefToLLVM/AllocLikeConversion.cpp
