Classes
struct	NDVectorTypeInfo

struct	LLVMStructTypeStorage
	Type storage for LLVM structure types. More...

struct	LLVMTypeAndSizeStorage
	Common storage used for LLVM dialect types that need an element type and a number: arrays, fixed and scalable vectors. More...

class	DataLayoutImporter
	Helper class that translates an LLVM data layout to an MLIR data layout specification. More...

class	DebugImporter

class	DebugTranslation

class	LoopAnnotationImporter
	A helper class that converts llvm.loop metadata nodes into corresponding LoopAnnotationAttrs and llvm.access.group nodes into AccessGroupAttrs. More...

class	LoopAnnotationTranslation
	A helper class that converts LoopAnnotationAttrs and AccessGroupAttrs into corresponding llvm::MDNodes. More...

class	TypeFromLLVMIRTranslatorImpl
	Support for translating LLVM IR types to MLIR LLVM dialect types. More...

class	TypeToLLVMIRTranslatorImpl
	Support for translating MLIR LLVM dialect types to LLVM IR. More...

Functions
void	setNativeProperties (Operation *op, IntegerOverflowFlags overflowFlags)
	Handle generically setting flags as native properties on LLVM operations. More...

LogicalResult	oneToOneRewrite (Operation *op, StringRef targetOp, ValueRange operands, ArrayRef< NamedAttribute > targetAttrs, const LLVMTypeConverter &typeConverter, ConversionPatternRewriter &rewriter, IntegerOverflowFlags overflowFlags=IntegerOverflowFlags::none)
	Replaces the given operation "op" with a new operation of type "targetOp" and given operands. More...

LogicalResult	intrinsicRewrite (Operation *op, StringRef intrinsic, ValueRange operands, const LLVMTypeConverter &typeConverter, RewriterBase &rewriter)
	Replaces the given operation "op" with a call to an LLVM intrinsic with the specified name "intrinsic" and operands. More...

NDVectorTypeInfo	extractNDVectorTypeInfo (VectorType vectorType, const LLVMTypeConverter &converter)

SmallVector< int64_t, 4 >	getCoordinates (ArrayRef< int64_t > basis, unsigned linearIndex)

void	nDVectorIterate (const NDVectorTypeInfo &info, OpBuilder &builder, function_ref< void(ArrayRef< int64_t >)> fun)

LogicalResult	handleMultidimensionalVectors (Operation *op, ValueRange operands, const LLVMTypeConverter &typeConverter, std::function< Value(Type, ValueRange)> createOperand, ConversionPatternRewriter &rewriter)

LogicalResult	vectorOneToOneRewrite (Operation *op, StringRef targetOp, ValueRange operands, ArrayRef< NamedAttribute > targetAttrs, const LLVMTypeConverter &typeConverter, ConversionPatternRewriter &rewriter, IntegerOverflowFlags overflowFlags=IntegerOverflowFlags::none)

LogicalResult	verifyAccessGroupOpInterface (Operation *op)
	Verifies the access groups attribute of memory operations that implement the access group interface. More...

LogicalResult	verifyAliasAnalysisOpInterface (Operation *op)
	Verifies the alias analysis attributes of memory operations that implement the alias analysis interface. More...

LogicalResult	verifyDereferenceableOpInterface (Operation *op)
	Verifies that the operation implementing the dereferenceable interface has exactly one result of LLVM pointer type. More...

Type	parseType (DialectAsmParser &parser)
	Parses an LLVM dialect type. More...

void	printType (Type type, AsmPrinter &printer)
	Prints an LLVM Dialect type. More...

void	connectPHINodes (Region &region, const ModuleTranslation &state)
	For all blocks in the region that were converted to LLVM IR using the given ModuleTranslation, connect the PHI nodes of the corresponding LLVM IR blocks to the results of preceding blocks. More...

llvm::Constant *	getLLVMConstant (llvm::Type *llvmType, Attribute attr, Location loc, const ModuleTranslation &moduleTranslation)
	Create an LLVM IR constant of `llvmType` from the MLIR attribute `attr`. More...

llvm::CallInst *	createIntrinsicCall (llvm::IRBuilderBase &builder, llvm::Intrinsic::ID intrinsic, ArrayRef< llvm::Value * > args={}, ArrayRef< llvm::Type * > tys={})
	Creates a call to an LLVM IR intrinsic function with the given arguments. More...

llvm::CallInst *	createIntrinsicCall (llvm::IRBuilderBase &builder, ModuleTranslation &moduleTranslation, Operation *intrOp, llvm::Intrinsic::ID intrinsic, unsigned numResults, ArrayRef< unsigned > overloadedResults, ArrayRef< unsigned > overloadedOperands, ArrayRef< unsigned > immArgPositions, ArrayRef< StringLiteral > immArgAttrNames)
	Creates a call to a LLVM IR intrinsic defined by LLVM_IntrOpBase. More...

static llvm::ArrayRef< std::pair< llvm::Attribute::AttrKind, llvm::StringRef > >	getAttrKindToNameMapping ()
	Returns a list of pairs that each hold a mapping from LLVM attribute kinds to their corresponding string name in LLVM IR dialect. More...

static llvm::DenseMap< llvm::StringRef, llvm::Attribute::AttrKind >	getAttrNameToKindMapping ()
	Returns a dense map from LLVM attribute name to their kind in LLVM IR dialect. More...

FloatType	getFloatType (MLIRContext *context, unsigned width)
	Returns a supported MLIR floating point type of the given bit width or null if the bit width is not supported. More...

Function Documentation

◆ connectPHINodes()

void mlir::LLVM::detail::connectPHINodes	(	Region &	region,
		const ModuleTranslation &	state
	)

For all blocks in the region that were converted to LLVM IR using the given ModuleTranslation, connect the PHI nodes of the corresponding LLVM IR blocks to the results of preceding blocks.

Connect the PHI nodes to the results of preceding blocks.

Definition at line 875 of file ModuleTranslation.cpp.

Referenced by convertOmpOpRegions().

◆ createIntrinsicCall() [1/2]

llvm::CallInst * mlir::LLVM::detail::createIntrinsicCall	(	llvm::IRBuilderBase &	builder,
		llvm::Intrinsic::ID	intrinsic,
		ArrayRef< llvm::Value * >	args = `{}`,
		ArrayRef< llvm::Type * >	tys = `{}`
	)

Creates a call to an LLVM IR intrinsic function with the given arguments.

Definition at line 902 of file ModuleTranslation.cpp.

◆ createIntrinsicCall() [2/2]

llvm::CallInst * mlir::LLVM::detail::createIntrinsicCall	(	llvm::IRBuilderBase &	builder,
		ModuleTranslation &	moduleTranslation,
		Operation *	intrOp,
		llvm::Intrinsic::ID	intrinsic,
		unsigned	numResults,
		ArrayRef< unsigned >	overloadedResults,
		ArrayRef< unsigned >	overloadedOperands,
		ArrayRef< unsigned >	immArgPositions,
		ArrayRef< StringLiteral >	immArgAttrNames
	)

Creates a call to a LLVM IR intrinsic defined by LLVM_IntrOpBase.

This resolves the overloads, and maps mixed MLIR value and attribute arguments to LLVM values.

Definition at line 911 of file ModuleTranslation.cpp.

References mlir::LLVM::ModuleTranslation::convertType(), mlir::Operation::getAttr(), getLLVMConstant(), mlir::Operation::getLoc(), mlir::Operation::getOperands(), mlir::Operation::getResult(), mlir::Value::getType(), mlir::getType(), and mlir::LLVM::ModuleTranslation::lookupValues().

◆ extractNDVectorTypeInfo()

LLVM::detail::NDVectorTypeInfo mlir::LLVM::detail::extractNDVectorTypeInfo	(	VectorType	vectorType,
		const LLVMTypeConverter &	converter
	)

Definition at line 19 of file VectorPattern.cpp.

References mlir::LLVM::detail::NDVectorTypeInfo::arraySizes, mlir::LLVMTypeConverter::convertType(), getNumElements(), mlir::LLVM::isCompatibleType(), mlir::LLVM::isCompatibleVectorType(), mlir::LLVM::detail::NDVectorTypeInfo::llvm1DVectorTy, and mlir::LLVM::detail::NDVectorTypeInfo::llvmNDVectorTy.

Referenced by handleMultidimensionalVectors().

◆ getAttrKindToNameMapping()

static llvm::ArrayRef<std::pair<llvm::Attribute::AttrKind, llvm::StringRef> > mlir::LLVM::detail::getAttrKindToNameMapping ( )

static

Returns a list of pairs that each hold a mapping from LLVM attribute kinds to their corresponding string name in LLVM IR dialect.

Definition at line 22 of file AttrKindDetail.h.

Referenced by getAttrNameToKindMapping().

◆ getAttrNameToKindMapping()

static llvm::DenseMap<llvm::StringRef, llvm::Attribute::AttrKind> mlir::LLVM::detail::getAttrNameToKindMapping ( )

static

Returns a dense map from LLVM attribute name to their kind in LLVM IR dialect.

Definition at line 69 of file AttrKindDetail.h.

References getAttrKindToNameMapping().

Referenced by mlir::LLVM::ModuleTranslation::convertParameterAttrs().

◆ getCoordinates()

SmallVector< int64_t, 4 > mlir::LLVM::detail::getCoordinates	(	ArrayRef< int64_t >	basis,
		unsigned	linearIndex
	)

Definition at line 48 of file VectorPattern.cpp.

Referenced by nDVectorIterate().

◆ getFloatType()

FloatType mlir::LLVM::detail::getFloatType	(	MLIRContext *	context,
		unsigned	width
	)

Returns a supported MLIR floating point type of the given bit width or null if the bit width is not supported.

Definition at line 28 of file DataLayoutImporter.cpp.

References mlir::get().

Referenced by getScalarConstantAsAttr().

◆ getLLVMConstant()

llvm::Constant * mlir::LLVM::detail::getLLVMConstant	(	llvm::Type *	llvmType,
		Attribute	attr,
		Location	loc,
		const ModuleTranslation &	moduleTranslation
	)

Create an LLVM IR constant of llvmType from the MLIR attribute attr.

This currently supports integer, floating point, splat and dense element attributes and combinations thereof. Also, an array attribute with two elements is supported to represent a complex constant. In case of error, report it to loc and return nullptr.

Definition at line 572 of file ModuleTranslation.cpp.

References buildSequentialConstant(), convertDenseElementsAttr(), convertDenseResourceElementsAttr(), mlir::emitError(), mlir::get(), getInnermostElementType(), mlir::LLVM::ModuleTranslation::getLLVMContext(), and mlir::LLVM::ModuleTranslation::lookupFunction().

Referenced by convertDenseElementsAttr(), createIntrinsicCall(), and createVL().

◆ handleMultidimensionalVectors()

LogicalResult mlir::LLVM::detail::handleMultidimensionalVectors	(	Operation *	op,
		ValueRange	operands,
		const LLVMTypeConverter &	typeConverter,
		std::function< Value(Type, ValueRange)>	createOperand,
		ConversionPatternRewriter &	rewriter
	)

Definition at line 80 of file VectorPattern.cpp.

References mlir::OpBuilder::create(), mlir::detail::enumerate(), extractNDVectorTypeInfo(), mlir::Operation::getLoc(), mlir::Operation::getResult(), mlir::Value::getType(), nDVectorIterate(), and mlir::ConversionPatternRewriter::replaceOp().

Referenced by mlir::impl::scalarizeVectorOp().

◆ intrinsicRewrite()

LogicalResult mlir::LLVM::detail::intrinsicRewrite	(	Operation *	op,
		StringRef	intrinsic,
		ValueRange	operands,
		const LLVMTypeConverter &	typeConverter,
		RewriterBase &	rewriter
	)

Replaces the given operation "op" with a call to an LLVM intrinsic with the specified name "intrinsic" and operands.

The rewrite performs a simple one-to-one matching between the op and LLVM intrinsic. For example:

%res = intr.op %val : vector<16xf32>

can be converted to

%res = llvm.call_intrinsic "intrinsic"(%val)

The provided operands must be LLVM-compatible.

Upholds a convention that multi-result operations get converted into an operation returning the LLVM IR structure type, in which case individual values are first extracted before replacing the original results.

Definition at line 356 of file Pattern.cpp.

References mlir::OpBuilder::create(), mlir::Operation::getAttrDictionary(), mlir::Operation::getLoc(), mlir::Operation::getNumResults(), mlir::Operation::getResultTypes(), mlir::Builder::getStringAttr(), mlir::Value::getType(), mlir::LLVM::isCompatibleType(), mlir::LLVMTypeConverter::packOperationResults(), mlir::RewriterBase::replaceOp(), and mlir::Operation::setAttrs().

◆ nDVectorIterate()

void mlir::LLVM::detail::nDVectorIterate	(	const NDVectorTypeInfo &	info,
		OpBuilder &	builder,
		function_ref< void(ArrayRef< int64_t >)>	fun
	)

Definition at line 64 of file VectorPattern.cpp.

References mlir::LLVM::detail::NDVectorTypeInfo::arraySizes, and getCoordinates().

Referenced by handleMultidimensionalVectors().

◆ oneToOneRewrite()

LogicalResult mlir::LLVM::detail::oneToOneRewrite	(	Operation *	op,
		StringRef	targetOp,
		ValueRange	operands,
		ArrayRef< NamedAttribute >	targetAttrs,
		const LLVMTypeConverter &	typeConverter,
		ConversionPatternRewriter &	rewriter,
		IntegerOverflowFlags	overflowFlags = `IntegerOverflowFlags::none`
	)

Replaces the given operation "op" with a new operation of type "targetOp" and given operands.

Definition at line 316 of file Pattern.cpp.

References mlir::OpBuilder::create(), mlir::ConversionPatternRewriter::eraseOp(), mlir::Operation::getLoc(), mlir::Operation::getNumResults(), mlir::Operation::getResult(), mlir::Operation::getResultTypes(), mlir::Builder::getStringAttr(), mlir::LLVMTypeConverter::packOperationResults(), mlir::ConversionPatternRewriter::replaceOp(), and setNativeProperties().

Referenced by mlir::OneToOneConvertToLLVMPattern< SourceOp, TargetOp >::matchAndRewrite(), and vectorOneToOneRewrite().

◆ parseType()

Type mlir::LLVM::detail::parseType ( DialectAsmParser & parser )

Parses an LLVM dialect type.

Parses one of the LLVM dialect types.

Definition at line 293 of file LLVMTypeSyntax.cpp.

◆ printType()

void mlir::LLVM::detail::printType	(	Type	type,
		AsmPrinter &	printer
	)

Prints an LLVM Dialect type.

Prints the given LLVM dialect type recursively.

This leverages closedness of the LLVM dialect type system to avoid printing the dialect prefix repeatedly. For recursive structures, only prints the name of the structure when printing a self-reference. Note that this does not apply to sibling references. For example, struct<"a", (ptr<struct<"a">>)> struct<"c", (ptr<struct<"b", (ptr<struct<"c">>)>>, ptr<struct<"b", (ptr<struct<"c">>)>>)> note that "b" is printed twice.

Definition at line 99 of file LLVMTypeSyntax.cpp.

Referenced by dispatchPrint().

◆ setNativeProperties()

void mlir::LLVM::detail::setNativeProperties	(	Operation *	op,
		IntegerOverflowFlags	overflowFlags
	)

Handle generically setting flags as native properties on LLVM operations.

Definition at line 308 of file Pattern.cpp.

Referenced by oneToOneRewrite(), and vectorOneToOneRewrite().

◆ vectorOneToOneRewrite()

LogicalResult mlir::LLVM::detail::vectorOneToOneRewrite	(	Operation *	op,
		StringRef	targetOp,
		ValueRange	operands,
		ArrayRef< NamedAttribute >	targetAttrs,
		const LLVMTypeConverter &	typeConverter,
		ConversionPatternRewriter &	rewriter,
		IntegerOverflowFlags	overflowFlags = `IntegerOverflowFlags::none`
	)

Definition at line 106 of file VectorPattern.cpp.

References mlir::OpBuilder::create(), mlir::Operation::getLoc(), mlir::Operation::getResult(), mlir::Builder::getStringAttr(), mlir::ValueRange::getType(), mlir::ValueRange::getTypes(), handleMultidimensionalVectors(), mlir::LLVM::isCompatibleType(), oneToOneRewrite(), and setNativeProperties().

Referenced by mlir::VectorConvertToLLVMPattern< SourceOp, TargetOp, AttrConvert >::matchAndRewrite().

◆ verifyAccessGroupOpInterface()

LogicalResult mlir::LLVM::detail::verifyAccessGroupOpInterface ( Operation * op )

Verifies the access groups attribute of memory operations that implement the access group interface.

Definition at line 35 of file LLVMInterfaces.cpp.

◆ verifyAliasAnalysisOpInterface()

LogicalResult mlir::LLVM::detail::verifyAliasAnalysisOpInterface ( Operation * op )

Verifies the alias analysis attributes of memory operations that implement the alias analysis interface.

Definition at line 49 of file LLVMInterfaces.cpp.

◆ verifyDereferenceableOpInterface()

LogicalResult mlir::LLVM::detail::verifyDereferenceableOpInterface ( Operation * op )

Verifies that the operation implementing the dereferenceable interface has exactly one result of LLVM pointer type.

Definition at line 72 of file LLVMInterfaces.cpp.

References mlir::Operation::emitOpError(), mlir::Operation::getNumResults(), mlir::Operation::getResult(), and mlir::Value::getType().

Classes

Functions