diff --git a/mlir/examples/toy/Ch6/CMakeLists.txt b/mlir/examples/toy/Ch6/CMakeLists.txt --- a/mlir/examples/toy/Ch6/CMakeLists.txt +++ b/mlir/examples/toy/Ch6/CMakeLists.txt @@ -42,8 +42,10 @@ MLIRCastInterfaces MLIRExecutionEngine MLIRIR + MLIRLLVMCommonConversion MLIRLLVMIR MLIRLLVMToLLVMIRTranslation + MLIRMemRef MLIRParser MLIRPass MLIRSideEffectInterfaces diff --git a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp --- a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp +++ b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp @@ -25,6 +25,9 @@ #include "toy/Passes.h" #include "mlir/Conversion/AffineToStandard/AffineToStandard.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/SCFToStandard/SCFToStandard.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" @@ -195,6 +198,7 @@ RewritePatternSet patterns(&getContext()); populateAffineToStdConversionPatterns(patterns); populateLoopToStdConversionPatterns(patterns); + populateMemRefToLLVMConversionPatterns(typeConverter, patterns); populateStdToLLVMConversionPatterns(typeConverter, patterns); // The only remaining operation to lower from the `toy` dialect, is the diff --git a/mlir/examples/toy/Ch7/CMakeLists.txt b/mlir/examples/toy/Ch7/CMakeLists.txt --- a/mlir/examples/toy/Ch7/CMakeLists.txt +++ b/mlir/examples/toy/Ch7/CMakeLists.txt @@ -42,7 +42,9 @@ MLIRCastInterfaces MLIRExecutionEngine MLIRIR + MLIRLLVMCommonConversion MLIRLLVMToLLVMIRTranslation + MLIRMemRef MLIRParser MLIRPass MLIRSideEffectInterfaces diff --git a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp --- a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp +++ b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp @@ -25,6 +25,9 @@ #include "toy/Passes.h" #include "mlir/Conversion/AffineToStandard/AffineToStandard.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/SCFToStandard/SCFToStandard.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" @@ -195,6 +198,7 @@ RewritePatternSet patterns(&getContext()); populateAffineToStdConversionPatterns(patterns); populateLoopToStdConversionPatterns(patterns); + populateMemRefToLLVMConversionPatterns(typeConverter, patterns); populateStdToLLVMConversionPatterns(typeConverter, patterns); // The only remaining operation to lower from the `toy` dialect, is the diff --git a/mlir/include/mlir/Conversion/ComplexToLLVM/ComplexToLLVM.h b/mlir/include/mlir/Conversion/ComplexToLLVM/ComplexToLLVM.h --- a/mlir/include/mlir/Conversion/ComplexToLLVM/ComplexToLLVM.h +++ b/mlir/include/mlir/Conversion/ComplexToLLVM/ComplexToLLVM.h @@ -9,13 +9,13 @@ #define MLIR_CONVERSION_COMPLEXTOLLVM_COMPLEXTOLLVM_H_ #include "mlir/Conversion/LLVMCommon/StructBuilder.h" -#include "mlir/Transforms/DialectConversion.h" namespace mlir { class LLVMTypeConverter; class ModuleOp; template class OperationPass; +class RewritePatternSet; class ComplexStructBuilder : public StructBuilder { public: diff --git a/mlir/include/mlir/Conversion/LLVMCommon/ConversionTarget.h b/mlir/include/mlir/Conversion/LLVMCommon/ConversionTarget.h new file mode 100644 --- /dev/null +++ b/mlir/include/mlir/Conversion/LLVMCommon/ConversionTarget.h @@ -0,0 +1,23 @@ +//===- ConversionTarget.h - LLVM dialect conversion target ------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef MLIR_CONVERSION_LLVMCOMMON_CONVERSIONTARGET_H +#define MLIR_CONVERSION_LLVMCOMMON_CONVERSIONTARGET_H + +#include "mlir/Transforms/DialectConversion.h" + +namespace mlir { +/// Derived class that automatically populates legalization information for +/// different LLVM ops. +class LLVMConversionTarget : public ConversionTarget { +public: + explicit LLVMConversionTarget(MLIRContext &ctx); +}; +} // namespace mlir + +#endif // MLIR_CONVERSION_LLVMCOMMON_CONVERSIONTARGET_H diff --git a/mlir/include/mlir/Conversion/LinalgToLLVM/LinalgToLLVM.h b/mlir/include/mlir/Conversion/LinalgToLLVM/LinalgToLLVM.h --- a/mlir/include/mlir/Conversion/LinalgToLLVM/LinalgToLLVM.h +++ b/mlir/include/mlir/Conversion/LinalgToLLVM/LinalgToLLVM.h @@ -8,14 +8,15 @@ #ifndef MLIR_CONVERSION_LINALGTOLLVM_LINALGTOLLVM_H_ #define MLIR_CONVERSION_LINALGTOLLVM_LINALGTOLLVM_H_ -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Transforms/DialectConversion.h" +#include namespace mlir { +class LLVMTypeConverter; class MLIRContext; class ModuleOp; template class OperationPass; +class RewritePatternSet; /// Populate the given list with patterns that convert from Linalg to LLVM. void populateLinalgToLLVMConversionPatterns(LLVMTypeConverter &converter, diff --git a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h b/mlir/include/mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h copy from mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h copy to mlir/include/mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h --- a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h +++ b/mlir/include/mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h @@ -1,52 +1,18 @@ -//===- ConvertStandardToLLVM.h - Convert to the LLVM dialect ----*- C++ -*-===// +//===- AllocLikeConversion.h - Convert allocation ops to LLVM ---*- C++ -*-===// // // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // See https://llvm.org/LICENSE.txt for license information. // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // //===----------------------------------------------------------------------===// -// -// Provides a dialect conversion targeting the LLVM IR dialect. By default, it -// converts Standard ops and types and provides hooks for dialect-specific -// extensions to the conversion. -// -//===----------------------------------------------------------------------===// -#ifndef MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H -#define MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H +#ifndef MLIR_CONVERSION_MEMREFTOLLVM_ALLOCLIKECONVERSION_H +#define MLIR_CONVERSION_MEMREFTOLLVM_ALLOCLIKECONVERSION_H #include "mlir/Conversion/LLVMCommon/Pattern.h" namespace mlir { -class LLVMTypeConverter; -class RewritePatternSet; - -/// Collect a set of patterns to convert memory-related operations from the -/// Standard dialect to the LLVM dialect, excluding non-memory-related -/// operations and FuncOp. -void populateStdToLLVMMemoryConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - -/// Collect a set of patterns to convert from the Standard dialect to the LLVM -/// dialect, excluding the memory-related operations. -void populateStdToLLVMNonMemoryConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - -/// Collect the default pattern to convert a FuncOp to the LLVM dialect. If -/// `emitCWrappers` is set, the pattern will also produce functions -/// that pass memref descriptors by pointer-to-structure in addition to the -/// default unpacked form. -void populateStdToLLVMFuncOpConversionPattern(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - -/// Collect the patterns to convert from the Standard dialect to LLVM. The -/// conversion patterns capture the LLVMTypeConverter and the LowerToLLVMOptions -/// by reference meaning the references have to remain alive during the entire -/// pattern lifetime. -void populateStdToLLVMConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - /// Lowering for AllocOp and AllocaOp. struct AllocLikeOpLLVMLowering : public ConvertToLLVMPattern { using ConvertToLLVMPattern::createIndexConstant; @@ -96,13 +62,6 @@ ConversionPatternRewriter &rewriter) const override; }; -/// Derived class that automatically populates legalization information for -/// different LLVM ops. -class LLVMConversionTarget : public ConversionTarget { -public: - explicit LLVMConversionTarget(MLIRContext &ctx); -}; - } // namespace mlir -#endif // MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H +#endif // MLIR_CONVERSION_MEMREFTOLLVM_ALLOCLIKECONVERSION_H diff --git a/mlir/include/mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h b/mlir/include/mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h new file mode 100644 --- /dev/null +++ b/mlir/include/mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h @@ -0,0 +1,27 @@ +//===- MemRefToLLVM.h - MemRef to LLVM dialect conversion -------*- C++ -*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#ifndef MLIR_CONVERSION_MEMREFTOLLVM_MEMREFTOLLVM_H +#define MLIR_CONVERSION_MEMREFTOLLVM_MEMREFTOLLVM_H + +#include + +namespace mlir { +class Pass; +class LLVMTypeConverter; +class RewritePatternSet; + +/// Collect a set of patterns to convert memory-related operations from the +/// MemRef dialect to the LLVM dialect. +void populateMemRefToLLVMConversionPatterns(LLVMTypeConverter &converter, + RewritePatternSet &patterns); + +std::unique_ptr createMemRefToLLVMPass(); +} // namespace mlir + +#endif // MLIR_CONVERSION_MEMREFTOLLVM_MEMREFTOLLVM_H diff --git a/mlir/include/mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h b/mlir/include/mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h --- a/mlir/include/mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h +++ b/mlir/include/mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h @@ -8,7 +8,7 @@ #ifndef MLIR_CONVERSION_OPENACCTOLLVM_CONVERTOPENACCTOLLVM_H #define MLIR_CONVERSION_OPENACCTOLLVM_CONVERTOPENACCTOLLVM_H -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/StructBuilder.h" #include namespace mlir { diff --git a/mlir/include/mlir/Conversion/Passes.h b/mlir/include/mlir/Conversion/Passes.h --- a/mlir/include/mlir/Conversion/Passes.h +++ b/mlir/include/mlir/Conversion/Passes.h @@ -23,6 +23,7 @@ #include "mlir/Conversion/LinalgToSPIRV/LinalgToSPIRVPass.h" #include "mlir/Conversion/LinalgToStandard/LinalgToStandard.h" #include "mlir/Conversion/MathToLibm/MathToLibm.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h" #include "mlir/Conversion/OpenACCToSCF/ConvertOpenACCToSCF.h" #include "mlir/Conversion/OpenMPToLLVM/ConvertOpenMPToLLVM.h" diff --git a/mlir/include/mlir/Conversion/Passes.td b/mlir/include/mlir/Conversion/Passes.td --- a/mlir/include/mlir/Conversion/Passes.td +++ b/mlir/include/mlir/Conversion/Passes.td @@ -255,6 +255,24 @@ let dependentDialects = ["StandardOpsDialect", "vector::VectorDialect"]; } +//===----------------------------------------------------------------------===// +// MemRefToLLVM +//===----------------------------------------------------------------------===// + +def ConvertMemRefToLLVM : Pass<"convert-memref-to-llvm", "ModuleOp"> { + let summary = "Convert operations from the MemRef dialect to the LLVM " + "dialect"; + let constructor = "mlir::createMemRefToLLVMPass()"; + let dependentDialects = ["LLVM::LLVMDialect"]; + let options = [ + Option<"useAlignedAlloc", "use-aligned-alloc", "bool", /*default=*/"false", + "Use aligned_alloc in place of malloc for heap allocations">, + Option<"indexBitwidth", "index-bitwidth", "unsigned", + /*default=kDeriveIndexBitwidthFromDataLayout*/"0", + "Bitwidth of the index type, 0 to use size of machine word">, + ]; +} + //===----------------------------------------------------------------------===// // OpenACCToSCF //===----------------------------------------------------------------------===// @@ -434,8 +452,6 @@ let constructor = "mlir::createLowerToLLVMPass()"; let dependentDialects = ["LLVM::LLVMDialect"]; let options = [ - Option<"useAlignedAlloc", "use-aligned-alloc", "bool", /*default=*/"false", - "Use aligned_alloc in place of malloc for heap allocations">, Option<"useBarePtrCallConv", "use-bare-ptr-memref-call-conv", "bool", /*default=*/"false", "Replace FuncOp's MemRef arguments with bare pointers to the MemRef " diff --git a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h b/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h --- a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h +++ b/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h @@ -15,24 +15,12 @@ #ifndef MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H #define MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H -#include "mlir/Conversion/LLVMCommon/Pattern.h" - namespace mlir { +class MLIRContext; class LLVMTypeConverter; class RewritePatternSet; -/// Collect a set of patterns to convert memory-related operations from the -/// Standard dialect to the LLVM dialect, excluding non-memory-related -/// operations and FuncOp. -void populateStdToLLVMMemoryConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - -/// Collect a set of patterns to convert from the Standard dialect to the LLVM -/// dialect, excluding the memory-related operations. -void populateStdToLLVMNonMemoryConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns); - /// Collect the default pattern to convert a FuncOp to the LLVM dialect. If /// `emitCWrappers` is set, the pattern will also produce functions /// that pass memref descriptors by pointer-to-structure in addition to the @@ -47,62 +35,6 @@ void populateStdToLLVMConversionPatterns(LLVMTypeConverter &converter, RewritePatternSet &patterns); -/// Lowering for AllocOp and AllocaOp. -struct AllocLikeOpLLVMLowering : public ConvertToLLVMPattern { - using ConvertToLLVMPattern::createIndexConstant; - using ConvertToLLVMPattern::getIndexType; - using ConvertToLLVMPattern::getVoidPtrType; - - explicit AllocLikeOpLLVMLowering(StringRef opName, - LLVMTypeConverter &converter) - : ConvertToLLVMPattern(opName, &converter.getContext(), converter) {} - -protected: - // Returns 'input' aligned up to 'alignment'. Computes - // bumped = input + alignement - 1 - // aligned = bumped - bumped % alignment - static Value createAligned(ConversionPatternRewriter &rewriter, Location loc, - Value input, Value alignment); - - /// Allocates the underlying buffer. Returns the allocated pointer and the - /// aligned pointer. - virtual std::tuple - allocateBuffer(ConversionPatternRewriter &rewriter, Location loc, - Value sizeBytes, Operation *op) const = 0; - -private: - static MemRefType getMemRefResultType(Operation *op) { - return op->getResult(0).getType().cast(); - } - - // An `alloc` is converted into a definition of a memref descriptor value and - // a call to `malloc` to allocate the underlying data buffer. The memref - // descriptor is of the LLVM structure type where: - // 1. the first element is a pointer to the allocated (typed) data buffer, - // 2. the second element is a pointer to the (typed) payload, aligned to the - // specified alignment, - // 3. the remaining elements serve to store all the sizes and strides of the - // memref using LLVM-converted `index` type. - // - // Alignment is performed by allocating `alignment` more bytes than - // requested and shifting the aligned pointer relative to the allocated - // memory. Note: `alignment - ` would actually be - // sufficient. If alignment is unspecified, the two pointers are equal. - - // An `alloca` is converted into a definition of a memref descriptor value and - // an llvm.alloca to allocate the underlying data buffer. - LogicalResult - matchAndRewrite(Operation *op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override; -}; - -/// Derived class that automatically populates legalization information for -/// different LLVM ops. -class LLVMConversionTarget : public ConversionTarget { -public: - explicit LLVMConversionTarget(MLIRContext &ctx); -}; - } // namespace mlir #endif // MLIR_CONVERSION_STANDARDTOLLVM_CONVERTSTANDARDTOLLVM_H diff --git a/mlir/lib/Conversion/AsyncToLLVM/AsyncToLLVM.cpp b/mlir/lib/Conversion/AsyncToLLVM/AsyncToLLVM.cpp --- a/mlir/lib/Conversion/AsyncToLLVM/AsyncToLLVM.cpp +++ b/mlir/lib/Conversion/AsyncToLLVM/AsyncToLLVM.cpp @@ -9,6 +9,8 @@ #include "mlir/Conversion/AsyncToLLVM/AsyncToLLVM.h" #include "../PassDetail.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Dialect/Async/IR/Async.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" diff --git a/mlir/lib/Conversion/AsyncToLLVM/CMakeLists.txt b/mlir/lib/Conversion/AsyncToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/AsyncToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/AsyncToLLVM/CMakeLists.txt @@ -12,6 +12,7 @@ LINK_LIBS PUBLIC MLIRAsync + MLIRLLVMCommonConversion MLIRLLVMIR MLIRStandardOpsTransforms MLIRStandardToLLVM diff --git a/mlir/lib/Conversion/CMakeLists.txt b/mlir/lib/Conversion/CMakeLists.txt --- a/mlir/lib/Conversion/CMakeLists.txt +++ b/mlir/lib/Conversion/CMakeLists.txt @@ -13,6 +13,7 @@ add_subdirectory(LinalgToStandard) add_subdirectory(LLVMCommon) add_subdirectory(MathToLibm) +add_subdirectory(MemRefToLLVM) add_subdirectory(OpenACCToLLVM) add_subdirectory(OpenACCToSCF) add_subdirectory(OpenMPToLLVM) diff --git a/mlir/lib/Conversion/ComplexToLLVM/CMakeLists.txt b/mlir/lib/Conversion/ComplexToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/ComplexToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/ComplexToLLVM/CMakeLists.txt @@ -15,6 +15,5 @@ MLIRLLVMCommonConversion MLIRLLVMIR MLIRStandardOpsTransforms - MLIRStandardToLLVM MLIRTransforms ) diff --git a/mlir/lib/Conversion/ComplexToLLVM/ComplexToLLVM.cpp b/mlir/lib/Conversion/ComplexToLLVM/ComplexToLLVM.cpp --- a/mlir/lib/Conversion/ComplexToLLVM/ComplexToLLVM.cpp +++ b/mlir/lib/Conversion/ComplexToLLVM/ComplexToLLVM.cpp @@ -9,7 +9,8 @@ #include "mlir/Conversion/ComplexToLLVM/ComplexToLLVM.h" #include "../PassDetail.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/Complex/IR/Complex.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" diff --git a/mlir/lib/Conversion/GPUCommon/CMakeLists.txt b/mlir/lib/Conversion/GPUCommon/CMakeLists.txt --- a/mlir/lib/Conversion/GPUCommon/CMakeLists.txt +++ b/mlir/lib/Conversion/GPUCommon/CMakeLists.txt @@ -34,6 +34,7 @@ MLIRIR MLIRLLVMCommonConversion MLIRLLVMIR + MLIRMemRefToLLVM MLIRPass MLIRSupport MLIRStandardToLLVM diff --git a/mlir/lib/Conversion/GPUCommon/GPUOpsLowering.h b/mlir/lib/Conversion/GPUCommon/GPUOpsLowering.h --- a/mlir/lib/Conversion/GPUCommon/GPUOpsLowering.h +++ b/mlir/lib/Conversion/GPUCommon/GPUOpsLowering.h @@ -8,7 +8,7 @@ #ifndef MLIR_CONVERSION_GPUCOMMON_GPUOPSLOWERING_H_ #define MLIR_CONVERSION_GPUCOMMON_GPUOPSLOWERING_H_ -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" diff --git a/mlir/lib/Conversion/GPUCommon/GPUToLLVMConversion.cpp b/mlir/lib/Conversion/GPUCommon/GPUToLLVMConversion.cpp --- a/mlir/lib/Conversion/GPUCommon/GPUToLLVMConversion.cpp +++ b/mlir/lib/Conversion/GPUCommon/GPUToLLVMConversion.cpp @@ -17,7 +17,9 @@ #include "../PassDetail.h" #include "mlir/Conversion/AsyncToLLVM/AsyncToLLVM.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" #include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h" @@ -316,6 +318,7 @@ target.addIllegalDialect(); populateVectorToLLVMConversionPatterns(converter, patterns); + populateMemRefToLLVMConversionPatterns(converter, patterns); populateStdToLLVMConversionPatterns(converter, patterns); populateAsyncStructuralTypeConversionsAndLegality(converter, patterns, target); diff --git a/mlir/lib/Conversion/GPUCommon/IndexIntrinsicsOpLowering.h b/mlir/lib/Conversion/GPUCommon/IndexIntrinsicsOpLowering.h --- a/mlir/lib/Conversion/GPUCommon/IndexIntrinsicsOpLowering.h +++ b/mlir/lib/Conversion/GPUCommon/IndexIntrinsicsOpLowering.h @@ -8,7 +8,7 @@ #ifndef MLIR_CONVERSION_GPUCOMMON_INDEXINTRINSICSOPLOWERING_H_ #define MLIR_CONVERSION_GPUCOMMON_INDEXINTRINSICSOPLOWERING_H_ -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "llvm/ADT/StringSwitch.h" diff --git a/mlir/lib/Conversion/GPUCommon/OpToFuncCallLowering.h b/mlir/lib/Conversion/GPUCommon/OpToFuncCallLowering.h --- a/mlir/lib/Conversion/GPUCommon/OpToFuncCallLowering.h +++ b/mlir/lib/Conversion/GPUCommon/OpToFuncCallLowering.h @@ -8,7 +8,7 @@ #ifndef MLIR_CONVERSION_GPUCOMMON_OPTOFUNCCALLLOWERING_H_ #define MLIR_CONVERSION_GPUCOMMON_OPTOFUNCCALLLOWERING_H_ -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/StandardOps/IR/Ops.h" diff --git a/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt b/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt --- a/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt +++ b/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt @@ -16,6 +16,7 @@ MLIRLLVMCommonConversion MLIRLLVMIR MLIRMemRef + MLIRMemRefToLLVM MLIRNVVMIR MLIRPass MLIRStandardToLLVM diff --git a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp --- a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp +++ b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp @@ -13,8 +13,10 @@ #include "mlir/Conversion/GPUToNVVM/GPUToNVVMPass.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" #include "mlir/Conversion/LLVMCommon/LoweringOptions.h" #include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/GPU/Passes.h" @@ -169,6 +171,7 @@ (void)applyPatternsAndFoldGreedily(m, std::move(patterns)); populateStdToLLVMConversionPatterns(converter, llvmPatterns); + populateMemRefToLLVMConversionPatterns(converter, llvmPatterns); populateGpuToNVVMConversionPatterns(converter, llvmPatterns); populateGpuWMMAToNVVMConversionPatterns(converter, llvmPatterns); LLVMConversionTarget target(getContext()); diff --git a/mlir/lib/Conversion/GPUToNVVM/WmmaOpsToNvvm.cpp b/mlir/lib/Conversion/GPUToNVVM/WmmaOpsToNvvm.cpp --- a/mlir/lib/Conversion/GPUToNVVM/WmmaOpsToNvvm.cpp +++ b/mlir/lib/Conversion/GPUToNVVM/WmmaOpsToNvvm.cpp @@ -11,7 +11,7 @@ // //===----------------------------------------------------------------------===// -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/LLVMIR/NVVMDialect.h" diff --git a/mlir/lib/Conversion/GPUToROCDL/CMakeLists.txt b/mlir/lib/Conversion/GPUToROCDL/CMakeLists.txt --- a/mlir/lib/Conversion/GPUToROCDL/CMakeLists.txt +++ b/mlir/lib/Conversion/GPUToROCDL/CMakeLists.txt @@ -14,6 +14,7 @@ MLIRGPUToGPURuntimeTransforms MLIRLLVMCommonConversion MLIRLLVMIR + MLIRMemRefToLLVM MLIRROCDLIR MLIRPass MLIRStandardToLLVM diff --git a/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp b/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp --- a/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp +++ b/mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp @@ -13,8 +13,10 @@ #include "mlir/Conversion/GPUToROCDL/GPUToROCDLPass.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" #include "mlir/Conversion/LLVMCommon/LoweringOptions.h" #include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h" #include "mlir/Conversion/VectorToROCDL/VectorToROCDL.h" @@ -73,6 +75,7 @@ populateVectorToLLVMConversionPatterns(converter, llvmPatterns); populateVectorToROCDLConversionPatterns(converter, llvmPatterns); populateStdToLLVMConversionPatterns(converter, llvmPatterns); + populateMemRefToLLVMConversionPatterns(converter, llvmPatterns); populateGpuToROCDLConversionPatterns(converter, llvmPatterns); LLVMConversionTarget target(getContext()); configureGpuToROCDLConversionLegality(target); diff --git a/mlir/lib/Conversion/LLVMCommon/CMakeLists.txt b/mlir/lib/Conversion/LLVMCommon/CMakeLists.txt --- a/mlir/lib/Conversion/LLVMCommon/CMakeLists.txt +++ b/mlir/lib/Conversion/LLVMCommon/CMakeLists.txt @@ -1,4 +1,5 @@ add_mlir_conversion_library(MLIRLLVMCommonConversion + ConversionTarget.cpp LoweringOptions.cpp MemRefBuilder.cpp Pattern.cpp diff --git a/mlir/lib/Conversion/LLVMCommon/ConversionTarget.cpp b/mlir/lib/Conversion/LLVMCommon/ConversionTarget.cpp new file mode 100644 --- /dev/null +++ b/mlir/lib/Conversion/LLVMCommon/ConversionTarget.cpp @@ -0,0 +1,18 @@ +//===- ConversionTarget.cpp - Target for converting to the LLVM dialect ---===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Dialect/LLVMIR/LLVMDialect.h" + +using namespace mlir; + +mlir::LLVMConversionTarget::LLVMConversionTarget(MLIRContext &ctx) + : ConversionTarget(ctx) { + this->addLegalDialect(); + this->addIllegalOp(); +} diff --git a/mlir/lib/Conversion/LinalgToLLVM/CMakeLists.txt b/mlir/lib/Conversion/LinalgToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/LinalgToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/LinalgToLLVM/CMakeLists.txt @@ -16,9 +16,9 @@ MLIRAffineToStandard MLIRIR MLIRLinalg + MLIRLLVMCommonConversion MLIRLLVMIR MLIRSCFToStandard - MLIRStandardToLLVM MLIRTransforms MLIRVectorToLLVM MLIRVectorToSCF diff --git a/mlir/lib/Conversion/LinalgToLLVM/LinalgToLLVM.cpp b/mlir/lib/Conversion/LinalgToLLVM/LinalgToLLVM.cpp --- a/mlir/lib/Conversion/LinalgToLLVM/LinalgToLLVM.cpp +++ b/mlir/lib/Conversion/LinalgToLLVM/LinalgToLLVM.cpp @@ -10,9 +10,10 @@ #include "../PassDetail.h" #include "mlir/Conversion/AffineToStandard/AffineToStandard.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" #include "mlir/Conversion/SCFToStandard/SCFToStandard.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h" #include "mlir/Conversion/VectorToSCF/VectorToSCF.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" diff --git a/mlir/lib/Conversion/MemRefToLLVM/AllocLikeConversion.cpp b/mlir/lib/Conversion/MemRefToLLVM/AllocLikeConversion.cpp new file mode 100644 --- /dev/null +++ b/mlir/lib/Conversion/MemRefToLLVM/AllocLikeConversion.cpp @@ -0,0 +1,47 @@ + +#include "mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h" +#include "mlir/Dialect/LLVMIR/LLVMDialect.h" + +using namespace mlir; + +Value AllocLikeOpLLVMLowering::createAligned( + ConversionPatternRewriter &rewriter, Location loc, Value input, + Value alignment) { + Value one = createIndexAttrConstant(rewriter, loc, alignment.getType(), 1); + Value bump = rewriter.create(loc, alignment, one); + Value bumped = rewriter.create(loc, input, bump); + Value mod = rewriter.create(loc, bumped, alignment); + return rewriter.create(loc, bumped, mod); +} + +LogicalResult AllocLikeOpLLVMLowering::matchAndRewrite( + Operation *op, ArrayRef operands, + ConversionPatternRewriter &rewriter) const { + MemRefType memRefType = getMemRefResultType(op); + if (!isConvertibleAndHasIdentityMaps(memRefType)) + return rewriter.notifyMatchFailure(op, "incompatible memref type"); + auto loc = op->getLoc(); + + // Get actual sizes of the memref as values: static sizes are constant + // values and dynamic sizes are passed to 'alloc' as operands. In case of + // zero-dimensional memref, assume a scalar (size 1). + SmallVector sizes; + SmallVector strides; + Value sizeBytes; + this->getMemRefDescriptorSizes(loc, memRefType, operands, rewriter, sizes, + strides, sizeBytes); + + // Allocate the underlying buffer. + Value allocatedPtr; + Value alignedPtr; + std::tie(allocatedPtr, alignedPtr) = + this->allocateBuffer(rewriter, loc, sizeBytes, op); + + // Create the MemRef descriptor. + auto memRefDescriptor = this->createMemRefDescriptor( + loc, memRefType, allocatedPtr, alignedPtr, sizes, strides, rewriter); + + // Return the final value of the descriptor. + rewriter.replaceOp(op, {memRefDescriptor}); + return success(); +} diff --git a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt b/mlir/lib/Conversion/MemRefToLLVM/CMakeLists.txt copy from mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt copy to mlir/lib/Conversion/MemRefToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/MemRefToLLVM/CMakeLists.txt @@ -1,12 +1,12 @@ -add_mlir_conversion_library(MLIRStandardToLLVM - StandardToLLVM.cpp +add_mlir_conversion_library(MLIRMemRefToLLVM + AllocLikeConversion.cpp + MemRefToLLVM.cpp ADDITIONAL_HEADER_DIRS - ${MLIR_MAIN_INCLUDE_DIR}/mlir/Conversion/StandardToLLVM + ${MLIR_MAIN_INCLUDE_DIR}/mlir/Conversion/MemRefToLLVM DEPENDS MLIRConversionPassIncGen - intrinsics_gen LINK_COMPONENTS Core @@ -15,8 +15,7 @@ MLIRAnalysis MLIRDataLayoutInterfaces MLIRLLVMCommonConversion - MLIRLLVMIR - MLIRMath MLIRMemRef + MLIRLLVMIR MLIRTransforms ) diff --git a/mlir/lib/Conversion/MemRefToLLVM/MemRefToLLVM.cpp b/mlir/lib/Conversion/MemRefToLLVM/MemRefToLLVM.cpp new file mode 100644 --- /dev/null +++ b/mlir/lib/Conversion/MemRefToLLVM/MemRefToLLVM.cpp @@ -0,0 +1,1400 @@ +//===- MemRefToLLVM.cpp - MemRef to LLVM dialect conversion ---------------===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" +#include "../PassDetail.h" +#include "mlir/Analysis/DataLayoutAnalysis.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/AllocLikeConversion.h" +#include "mlir/Dialect/LLVMIR/FunctionCallUtils.h" +#include "mlir/Dialect/LLVMIR/LLVMDialect.h" +#include "mlir/Dialect/MemRef/IR/MemRef.h" +#include "mlir/IR/AffineMap.h" +#include "mlir/IR/BlockAndValueMapping.h" + +using namespace mlir; + +namespace { + +struct AllocOpLowering : public AllocLikeOpLLVMLowering { + AllocOpLowering(LLVMTypeConverter &converter) + : AllocLikeOpLLVMLowering(memref::AllocOp::getOperationName(), + converter) {} + + std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, + Location loc, Value sizeBytes, + Operation *op) const override { + // Heap allocations. + memref::AllocOp allocOp = cast(op); + MemRefType memRefType = allocOp.getType(); + + Value alignment; + if (auto alignmentAttr = allocOp.alignment()) { + alignment = createIndexConstant(rewriter, loc, *alignmentAttr); + } else if (!memRefType.getElementType().isSignlessIntOrIndexOrFloat()) { + // In the case where no alignment is specified, we may want to override + // `malloc's` behavior. `malloc` typically aligns at the size of the + // biggest scalar on a target HW. For non-scalars, use the natural + // alignment of the LLVM type given by the LLVM DataLayout. + alignment = getSizeInBytes(loc, memRefType.getElementType(), rewriter); + } + + if (alignment) { + // Adjust the allocation size to consider alignment. + sizeBytes = rewriter.create(loc, sizeBytes, alignment); + } + + // Allocate the underlying buffer and store a pointer to it in the MemRef + // descriptor. + Type elementPtrType = this->getElementPtrType(memRefType); + auto allocFuncOp = LLVM::lookupOrCreateMallocFn( + allocOp->getParentOfType(), getIndexType()); + auto results = createLLVMCall(rewriter, loc, allocFuncOp, {sizeBytes}, + getVoidPtrType()); + Value allocatedPtr = + rewriter.create(loc, elementPtrType, results[0]); + + Value alignedPtr = allocatedPtr; + if (alignment) { + // Compute the aligned type pointer. + Value allocatedInt = + rewriter.create(loc, getIndexType(), allocatedPtr); + Value alignmentInt = + createAligned(rewriter, loc, allocatedInt, alignment); + alignedPtr = + rewriter.create(loc, elementPtrType, alignmentInt); + } + + return std::make_tuple(allocatedPtr, alignedPtr); + } +}; + +struct AlignedAllocOpLowering : public AllocLikeOpLLVMLowering { + AlignedAllocOpLowering(LLVMTypeConverter &converter) + : AllocLikeOpLLVMLowering(memref::AllocOp::getOperationName(), + converter) {} + + /// Returns the memref's element size in bytes using the data layout active at + /// `op`. + // TODO: there are other places where this is used. Expose publicly? + unsigned getMemRefEltSizeInBytes(MemRefType memRefType, Operation *op) const { + const DataLayout *layout = &defaultLayout; + if (const DataLayoutAnalysis *analysis = + getTypeConverter()->getDataLayoutAnalysis()) { + layout = &analysis->getAbove(op); + } + Type elementType = memRefType.getElementType(); + if (auto memRefElementType = elementType.dyn_cast()) + return getTypeConverter()->getMemRefDescriptorSize(memRefElementType, + *layout); + if (auto memRefElementType = elementType.dyn_cast()) + return getTypeConverter()->getUnrankedMemRefDescriptorSize( + memRefElementType, *layout); + return layout->getTypeSize(elementType); + } + + /// Returns true if the memref size in bytes is known to be a multiple of + /// factor assuming the data layout active at `op`. + bool isMemRefSizeMultipleOf(MemRefType type, uint64_t factor, + Operation *op) const { + uint64_t sizeDivisor = getMemRefEltSizeInBytes(type, op); + for (unsigned i = 0, e = type.getRank(); i < e; i++) { + if (type.isDynamic(type.getDimSize(i))) + continue; + sizeDivisor = sizeDivisor * type.getDimSize(i); + } + return sizeDivisor % factor == 0; + } + + /// Returns the alignment to be used for the allocation call itself. + /// aligned_alloc requires the allocation size to be a power of two, and the + /// allocation size to be a multiple of alignment, + int64_t getAllocationAlignment(memref::AllocOp allocOp) const { + if (Optional alignment = allocOp.alignment()) + return *alignment; + + // Whenever we don't have alignment set, we will use an alignment + // consistent with the element type; since the allocation size has to be a + // power of two, we will bump to the next power of two if it already isn't. + auto eltSizeBytes = getMemRefEltSizeInBytes(allocOp.getType(), allocOp); + return std::max(kMinAlignedAllocAlignment, + llvm::PowerOf2Ceil(eltSizeBytes)); + } + + std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, + Location loc, Value sizeBytes, + Operation *op) const override { + // Heap allocations. + memref::AllocOp allocOp = cast(op); + MemRefType memRefType = allocOp.getType(); + int64_t alignment = getAllocationAlignment(allocOp); + Value allocAlignment = createIndexConstant(rewriter, loc, alignment); + + // aligned_alloc requires size to be a multiple of alignment; we will pad + // the size to the next multiple if necessary. + if (!isMemRefSizeMultipleOf(memRefType, alignment, op)) + sizeBytes = createAligned(rewriter, loc, sizeBytes, allocAlignment); + + Type elementPtrType = this->getElementPtrType(memRefType); + auto allocFuncOp = LLVM::lookupOrCreateAlignedAllocFn( + allocOp->getParentOfType(), getIndexType()); + auto results = + createLLVMCall(rewriter, loc, allocFuncOp, {allocAlignment, sizeBytes}, + getVoidPtrType()); + Value allocatedPtr = + rewriter.create(loc, elementPtrType, results[0]); + + return std::make_tuple(allocatedPtr, allocatedPtr); + } + + /// The minimum alignment to use with aligned_alloc (has to be a power of 2). + static constexpr uint64_t kMinAlignedAllocAlignment = 16UL; + + /// Default layout to use in absence of the corresponding analysis. + DataLayout defaultLayout; +}; + +// Out of line definition, required till C++17. +constexpr uint64_t AlignedAllocOpLowering::kMinAlignedAllocAlignment; + +struct AllocaOpLowering : public AllocLikeOpLLVMLowering { + AllocaOpLowering(LLVMTypeConverter &converter) + : AllocLikeOpLLVMLowering(memref::AllocaOp::getOperationName(), + converter) {} + + /// Allocates the underlying buffer using the right call. `allocatedBytePtr` + /// is set to null for stack allocations. `accessAlignment` is set if + /// alignment is needed post allocation (for eg. in conjunction with malloc). + std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, + Location loc, Value sizeBytes, + Operation *op) const override { + + // With alloca, one gets a pointer to the element type right away. + // For stack allocations. + auto allocaOp = cast(op); + auto elementPtrType = this->getElementPtrType(allocaOp.getType()); + + auto allocatedElementPtr = rewriter.create( + loc, elementPtrType, sizeBytes, + allocaOp.alignment() ? *allocaOp.alignment() : 0); + + return std::make_tuple(allocatedElementPtr, allocatedElementPtr); + } +}; + +struct AllocaScopeOpLowering + : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::AllocaScopeOp allocaScopeOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + OpBuilder::InsertionGuard guard(rewriter); + Location loc = allocaScopeOp.getLoc(); + + // Split the current block before the AllocaScopeOp to create the inlining + // point. + auto *currentBlock = rewriter.getInsertionBlock(); + auto *remainingOpsBlock = + rewriter.splitBlock(currentBlock, rewriter.getInsertionPoint()); + Block *continueBlock; + if (allocaScopeOp.getNumResults() == 0) { + continueBlock = remainingOpsBlock; + } else { + continueBlock = rewriter.createBlock(remainingOpsBlock, + allocaScopeOp.getResultTypes()); + rewriter.create(loc, ValueRange(), remainingOpsBlock); + } + + // Inline body region. + Block *beforeBody = &allocaScopeOp.bodyRegion().front(); + Block *afterBody = &allocaScopeOp.bodyRegion().back(); + rewriter.inlineRegionBefore(allocaScopeOp.bodyRegion(), continueBlock); + + // Save stack and then branch into the body of the region. + rewriter.setInsertionPointToEnd(currentBlock); + auto stackSaveOp = + rewriter.create(loc, getVoidPtrType()); + rewriter.create(loc, ValueRange(), beforeBody); + + // Replace the alloca_scope return with a branch that jumps out of the body. + // Stack restore before leaving the body region. + rewriter.setInsertionPointToEnd(afterBody); + auto returnOp = + cast(afterBody->getTerminator()); + auto branchOp = rewriter.replaceOpWithNewOp( + returnOp, returnOp.results(), continueBlock); + + // Insert stack restore before jumping out the body of the region. + rewriter.setInsertionPoint(branchOp); + rewriter.create(loc, stackSaveOp); + + // Replace the op with values return from the body region. + rewriter.replaceOp(allocaScopeOp, continueBlock->getArguments()); + + return success(); + } +}; + +struct AssumeAlignmentOpLowering + : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern< + memref::AssumeAlignmentOp>::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::AssumeAlignmentOp op, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + memref::AssumeAlignmentOp::Adaptor transformed(operands); + Value memref = transformed.memref(); + unsigned alignment = op.alignment(); + auto loc = op.getLoc(); + + MemRefDescriptor memRefDescriptor(memref); + Value ptr = memRefDescriptor.alignedPtr(rewriter, memref.getLoc()); + + // Emit llvm.assume(memref.alignedPtr & (alignment - 1) == 0). Notice that + // the asserted memref.alignedPtr isn't used anywhere else, as the real + // users like load/store/views always re-extract memref.alignedPtr as they + // get lowered. + // + // This relies on LLVM's CSE optimization (potentially after SROA), since + // after CSE all memref.alignedPtr instances get de-duplicated into the same + // pointer SSA value. + auto intPtrType = + getIntPtrType(memRefDescriptor.getElementPtrType().getAddressSpace()); + Value zero = createIndexAttrConstant(rewriter, loc, intPtrType, 0); + Value mask = + createIndexAttrConstant(rewriter, loc, intPtrType, alignment - 1); + Value ptrValue = rewriter.create(loc, intPtrType, ptr); + rewriter.create( + loc, rewriter.create( + loc, LLVM::ICmpPredicate::eq, + rewriter.create(loc, ptrValue, mask), zero)); + + rewriter.eraseOp(op); + return success(); + } +}; + +// A `dealloc` is converted into a call to `free` on the underlying data buffer. +// The memref descriptor being an SSA value, there is no need to clean it up +// in any way. +struct DeallocOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + explicit DeallocOpLowering(LLVMTypeConverter &converter) + : ConvertOpToLLVMPattern(converter) {} + + LogicalResult + matchAndRewrite(memref::DeallocOp op, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + assert(operands.size() == 1 && "dealloc takes one operand"); + memref::DeallocOp::Adaptor transformed(operands); + + // Insert the `free` declaration if it is not already present. + auto freeFunc = LLVM::lookupOrCreateFreeFn(op->getParentOfType()); + MemRefDescriptor memref(transformed.memref()); + Value casted = rewriter.create( + op.getLoc(), getVoidPtrType(), + memref.allocatedPtr(rewriter, op.getLoc())); + rewriter.replaceOpWithNewOp( + op, TypeRange(), rewriter.getSymbolRefAttr(freeFunc), casted); + return success(); + } +}; + +// A `dim` is converted to a constant for static sizes and to an access to the +// size stored in the memref descriptor for dynamic sizes. +struct DimOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::DimOp dimOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + Type operandType = dimOp.source().getType(); + if (operandType.isa()) { + rewriter.replaceOp(dimOp, {extractSizeOfUnrankedMemRef( + operandType, dimOp, operands, rewriter)}); + + return success(); + } + if (operandType.isa()) { + rewriter.replaceOp(dimOp, {extractSizeOfRankedMemRef( + operandType, dimOp, operands, rewriter)}); + return success(); + } + llvm_unreachable("expected MemRefType or UnrankedMemRefType"); + } + +private: + Value extractSizeOfUnrankedMemRef(Type operandType, memref::DimOp dimOp, + ArrayRef operands, + ConversionPatternRewriter &rewriter) const { + Location loc = dimOp.getLoc(); + memref::DimOp::Adaptor transformed(operands); + + auto unrankedMemRefType = operandType.cast(); + auto scalarMemRefType = + MemRefType::get({}, unrankedMemRefType.getElementType()); + unsigned addressSpace = unrankedMemRefType.getMemorySpaceAsInt(); + + // Extract pointer to the underlying ranked descriptor and bitcast it to a + // memref descriptor pointer to minimize the number of GEP + // operations. + UnrankedMemRefDescriptor unrankedDesc(transformed.source()); + Value underlyingRankedDesc = unrankedDesc.memRefDescPtr(rewriter, loc); + Value scalarMemRefDescPtr = rewriter.create( + loc, + LLVM::LLVMPointerType::get(typeConverter->convertType(scalarMemRefType), + addressSpace), + underlyingRankedDesc); + + // Get pointer to offset field of memref descriptor. + Type indexPtrTy = LLVM::LLVMPointerType::get( + getTypeConverter()->getIndexType(), addressSpace); + Value two = rewriter.create( + loc, typeConverter->convertType(rewriter.getI32Type()), + rewriter.getI32IntegerAttr(2)); + Value offsetPtr = rewriter.create( + loc, indexPtrTy, scalarMemRefDescPtr, + ValueRange({createIndexConstant(rewriter, loc, 0), two})); + + // The size value that we have to extract can be obtained using GEPop with + // `dimOp.index() + 1` index argument. + Value idxPlusOne = rewriter.create( + loc, createIndexConstant(rewriter, loc, 1), transformed.index()); + Value sizePtr = rewriter.create(loc, indexPtrTy, offsetPtr, + ValueRange({idxPlusOne})); + return rewriter.create(loc, sizePtr); + } + + Optional getConstantDimIndex(memref::DimOp dimOp) const { + if (Optional idx = dimOp.getConstantIndex()) + return idx; + + if (auto constantOp = dimOp.index().getDefiningOp()) + return constantOp.value().cast().getValue().getSExtValue(); + + return llvm::None; + } + + Value extractSizeOfRankedMemRef(Type operandType, memref::DimOp dimOp, + ArrayRef operands, + ConversionPatternRewriter &rewriter) const { + Location loc = dimOp.getLoc(); + memref::DimOp::Adaptor transformed(operands); + // Take advantage if index is constant. + MemRefType memRefType = operandType.cast(); + if (Optional index = getConstantDimIndex(dimOp)) { + int64_t i = index.getValue(); + if (memRefType.isDynamicDim(i)) { + // extract dynamic size from the memref descriptor. + MemRefDescriptor descriptor(transformed.source()); + return descriptor.size(rewriter, loc, i); + } + // Use constant for static size. + int64_t dimSize = memRefType.getDimSize(i); + return createIndexConstant(rewriter, loc, dimSize); + } + Value index = transformed.index(); + int64_t rank = memRefType.getRank(); + MemRefDescriptor memrefDescriptor(transformed.source()); + return memrefDescriptor.size(rewriter, loc, index, rank); + } +}; + +/// Returns the LLVM type of the global variable given the memref type `type`. +static Type convertGlobalMemrefTypeToLLVM(MemRefType type, + LLVMTypeConverter &typeConverter) { + // LLVM type for a global memref will be a multi-dimension array. For + // declarations or uninitialized global memrefs, we can potentially flatten + // this to a 1D array. However, for memref.global's with an initial value, + // we do not intend to flatten the ElementsAttribute when going from std -> + // LLVM dialect, so the LLVM type needs to me a multi-dimension array. + Type elementType = typeConverter.convertType(type.getElementType()); + Type arrayTy = elementType; + // Shape has the outermost dim at index 0, so need to walk it backwards + for (int64_t dim : llvm::reverse(type.getShape())) + arrayTy = LLVM::LLVMArrayType::get(arrayTy, dim); + return arrayTy; +} + +/// GlobalMemrefOp is lowered to a LLVM Global Variable. +struct GlobalMemrefOpLowering + : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::GlobalOp global, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + MemRefType type = global.type().cast(); + if (!isConvertibleAndHasIdentityMaps(type)) + return failure(); + + Type arrayTy = convertGlobalMemrefTypeToLLVM(type, *getTypeConverter()); + + LLVM::Linkage linkage = + global.isPublic() ? LLVM::Linkage::External : LLVM::Linkage::Private; + + Attribute initialValue = nullptr; + if (!global.isExternal() && !global.isUninitialized()) { + auto elementsAttr = global.initial_value()->cast(); + initialValue = elementsAttr; + + // For scalar memrefs, the global variable created is of the element type, + // so unpack the elements attribute to extract the value. + if (type.getRank() == 0) + initialValue = elementsAttr.getValue({}); + } + + rewriter.replaceOpWithNewOp( + global, arrayTy, global.constant(), linkage, global.sym_name(), + initialValue, /*alignment=*/0, type.getMemorySpaceAsInt()); + return success(); + } +}; + +/// GetGlobalMemrefOp is lowered into a Memref descriptor with the pointer to +/// the first element stashed into the descriptor. This reuses +/// `AllocLikeOpLowering` to reuse the Memref descriptor construction. +struct GetGlobalMemrefOpLowering : public AllocLikeOpLLVMLowering { + GetGlobalMemrefOpLowering(LLVMTypeConverter &converter) + : AllocLikeOpLLVMLowering(memref::GetGlobalOp::getOperationName(), + converter) {} + + /// Buffer "allocation" for memref.get_global op is getting the address of + /// the global variable referenced. + std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, + Location loc, Value sizeBytes, + Operation *op) const override { + auto getGlobalOp = cast(op); + MemRefType type = getGlobalOp.result().getType().cast(); + unsigned memSpace = type.getMemorySpaceAsInt(); + + Type arrayTy = convertGlobalMemrefTypeToLLVM(type, *getTypeConverter()); + auto addressOf = rewriter.create( + loc, LLVM::LLVMPointerType::get(arrayTy, memSpace), getGlobalOp.name()); + + // Get the address of the first element in the array by creating a GEP with + // the address of the GV as the base, and (rank + 1) number of 0 indices. + Type elementType = typeConverter->convertType(type.getElementType()); + Type elementPtrType = LLVM::LLVMPointerType::get(elementType, memSpace); + + SmallVector operands = {addressOf}; + operands.insert(operands.end(), type.getRank() + 1, + createIndexConstant(rewriter, loc, 0)); + auto gep = rewriter.create(loc, elementPtrType, operands); + + // We do not expect the memref obtained using `memref.get_global` to be + // ever deallocated. Set the allocated pointer to be known bad value to + // help debug if that ever happens. + auto intPtrType = getIntPtrType(memSpace); + Value deadBeefConst = + createIndexAttrConstant(rewriter, op->getLoc(), intPtrType, 0xdeadbeef); + auto deadBeefPtr = + rewriter.create(loc, elementPtrType, deadBeefConst); + + // Both allocated and aligned pointers are same. We could potentially stash + // a nullptr for the allocated pointer since we do not expect any dealloc. + return std::make_tuple(deadBeefPtr, gep); + } +}; + +// Common base for load and store operations on MemRefs. Restricts the match +// to supported MemRef types. Provides functionality to emit code accessing a +// specific element of the underlying data buffer. +template +struct LoadStoreOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + using ConvertOpToLLVMPattern::isConvertibleAndHasIdentityMaps; + using Base = LoadStoreOpLowering; + + LogicalResult match(Derived op) const override { + MemRefType type = op.getMemRefType(); + return isConvertibleAndHasIdentityMaps(type) ? success() : failure(); + } +}; + +// Load operation is lowered to obtaining a pointer to the indexed element +// and loading it. +struct LoadOpLowering : public LoadStoreOpLowering { + using Base::Base; + + LogicalResult + matchAndRewrite(memref::LoadOp loadOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + memref::LoadOp::Adaptor transformed(operands); + auto type = loadOp.getMemRefType(); + + Value dataPtr = + getStridedElementPtr(loadOp.getLoc(), type, transformed.memref(), + transformed.indices(), rewriter); + rewriter.replaceOpWithNewOp(loadOp, dataPtr); + return success(); + } +}; + +// Store operation is lowered to obtaining a pointer to the indexed element, +// and storing the given value to it. +struct StoreOpLowering : public LoadStoreOpLowering { + using Base::Base; + + LogicalResult + matchAndRewrite(memref::StoreOp op, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto type = op.getMemRefType(); + memref::StoreOp::Adaptor transformed(operands); + + Value dataPtr = + getStridedElementPtr(op.getLoc(), type, transformed.memref(), + transformed.indices(), rewriter); + rewriter.replaceOpWithNewOp(op, transformed.value(), + dataPtr); + return success(); + } +}; + +// The prefetch operation is lowered in a way similar to the load operation +// except that the llvm.prefetch operation is used for replacement. +struct PrefetchOpLowering : public LoadStoreOpLowering { + using Base::Base; + + LogicalResult + matchAndRewrite(memref::PrefetchOp prefetchOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + memref::PrefetchOp::Adaptor transformed(operands); + auto type = prefetchOp.getMemRefType(); + auto loc = prefetchOp.getLoc(); + + Value dataPtr = getStridedElementPtr(loc, type, transformed.memref(), + transformed.indices(), rewriter); + + // Replace with llvm.prefetch. + auto llvmI32Type = typeConverter->convertType(rewriter.getIntegerType(32)); + auto isWrite = rewriter.create( + loc, llvmI32Type, rewriter.getI32IntegerAttr(prefetchOp.isWrite())); + auto localityHint = rewriter.create( + loc, llvmI32Type, + rewriter.getI32IntegerAttr(prefetchOp.localityHint())); + auto isData = rewriter.create( + loc, llvmI32Type, rewriter.getI32IntegerAttr(prefetchOp.isDataCache())); + + rewriter.replaceOpWithNewOp(prefetchOp, dataPtr, isWrite, + localityHint, isData); + return success(); + } +}; + +struct MemRefCastOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult match(memref::CastOp memRefCastOp) const override { + Type srcType = memRefCastOp.getOperand().getType(); + Type dstType = memRefCastOp.getType(); + + // memref::CastOp reduce to bitcast in the ranked MemRef case and can be + // used for type erasure. For now they must preserve underlying element type + // and require source and result type to have the same rank. Therefore, + // perform a sanity check that the underlying structs are the same. Once op + // semantics are relaxed we can revisit. + if (srcType.isa() && dstType.isa()) + return success(typeConverter->convertType(srcType) == + typeConverter->convertType(dstType)); + + // At least one of the operands is unranked type + assert(srcType.isa() || + dstType.isa()); + + // Unranked to unranked cast is disallowed + return !(srcType.isa() && + dstType.isa()) + ? success() + : failure(); + } + + void rewrite(memref::CastOp memRefCastOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + memref::CastOp::Adaptor transformed(operands); + + auto srcType = memRefCastOp.getOperand().getType(); + auto dstType = memRefCastOp.getType(); + auto targetStructType = typeConverter->convertType(memRefCastOp.getType()); + auto loc = memRefCastOp.getLoc(); + + // For ranked/ranked case, just keep the original descriptor. + if (srcType.isa() && dstType.isa()) + return rewriter.replaceOp(memRefCastOp, {transformed.source()}); + + if (srcType.isa() && dstType.isa()) { + // Casting ranked to unranked memref type + // Set the rank in the destination from the memref type + // Allocate space on the stack and copy the src memref descriptor + // Set the ptr in the destination to the stack space + auto srcMemRefType = srcType.cast(); + int64_t rank = srcMemRefType.getRank(); + // ptr = AllocaOp sizeof(MemRefDescriptor) + auto ptr = getTypeConverter()->promoteOneMemRefDescriptor( + loc, transformed.source(), rewriter); + // voidptr = BitCastOp srcType* to void* + auto voidPtr = + rewriter.create(loc, getVoidPtrType(), ptr) + .getResult(); + // rank = ConstantOp srcRank + auto rankVal = rewriter.create( + loc, typeConverter->convertType(rewriter.getIntegerType(64)), + rewriter.getI64IntegerAttr(rank)); + // undef = UndefOp + UnrankedMemRefDescriptor memRefDesc = + UnrankedMemRefDescriptor::undef(rewriter, loc, targetStructType); + // d1 = InsertValueOp undef, rank, 0 + memRefDesc.setRank(rewriter, loc, rankVal); + // d2 = InsertValueOp d1, voidptr, 1 + memRefDesc.setMemRefDescPtr(rewriter, loc, voidPtr); + rewriter.replaceOp(memRefCastOp, (Value)memRefDesc); + + } else if (srcType.isa() && dstType.isa()) { + // Casting from unranked type to ranked. + // The operation is assumed to be doing a correct cast. If the destination + // type mismatches the unranked the type, it is undefined behavior. + UnrankedMemRefDescriptor memRefDesc(transformed.source()); + // ptr = ExtractValueOp src, 1 + auto ptr = memRefDesc.memRefDescPtr(rewriter, loc); + // castPtr = BitCastOp i8* to structTy* + auto castPtr = + rewriter + .create( + loc, LLVM::LLVMPointerType::get(targetStructType), ptr) + .getResult(); + // struct = LoadOp castPtr + auto loadOp = rewriter.create(loc, castPtr); + rewriter.replaceOp(memRefCastOp, loadOp.getResult()); + } else { + llvm_unreachable("Unsupported unranked memref to unranked memref cast"); + } + } +}; + +struct MemRefCopyOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::CopyOp op, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto loc = op.getLoc(); + memref::CopyOp::Adaptor adaptor(operands); + auto srcType = op.source().getType().cast(); + auto targetType = op.target().getType().cast(); + + // First make sure we have an unranked memref descriptor representation. + auto makeUnranked = [&, this](Value ranked, BaseMemRefType type) { + auto rank = rewriter.create( + loc, getIndexType(), rewriter.getIndexAttr(type.getRank())); + auto *typeConverter = getTypeConverter(); + auto ptr = + typeConverter->promoteOneMemRefDescriptor(loc, ranked, rewriter); + auto voidPtr = + rewriter.create(loc, getVoidPtrType(), ptr) + .getResult(); + auto unrankedType = + UnrankedMemRefType::get(type.getElementType(), type.getMemorySpace()); + return UnrankedMemRefDescriptor::pack(rewriter, loc, *typeConverter, + unrankedType, + ValueRange{rank, voidPtr}); + }; + + Value unrankedSource = srcType.hasRank() + ? makeUnranked(adaptor.source(), srcType) + : adaptor.source(); + Value unrankedTarget = targetType.hasRank() + ? makeUnranked(adaptor.target(), targetType) + : adaptor.target(); + + // Now promote the unranked descriptors to the stack. + auto one = rewriter.create(loc, getIndexType(), + rewriter.getIndexAttr(1)); + auto promote = [&](Value desc) { + auto ptrType = LLVM::LLVMPointerType::get(desc.getType()); + auto allocated = + rewriter.create(loc, ptrType, ValueRange{one}); + rewriter.create(loc, desc, allocated); + return allocated; + }; + + auto sourcePtr = promote(unrankedSource); + auto targetPtr = promote(unrankedTarget); + + auto elemSize = rewriter.create( + loc, getIndexType(), + rewriter.getIndexAttr(srcType.getElementTypeBitWidth() / 8)); + auto copyFn = LLVM::lookupOrCreateMemRefCopyFn( + op->getParentOfType(), getIndexType(), sourcePtr.getType()); + rewriter.create(loc, copyFn, + ValueRange{elemSize, sourcePtr, targetPtr}); + rewriter.eraseOp(op); + + return success(); + } +}; + +/// Extracts allocated, aligned pointers and offset from a ranked or unranked +/// memref type. In unranked case, the fields are extracted from the underlying +/// ranked descriptor. +static void extractPointersAndOffset(Location loc, + ConversionPatternRewriter &rewriter, + LLVMTypeConverter &typeConverter, + Value originalOperand, + Value convertedOperand, + Value *allocatedPtr, Value *alignedPtr, + Value *offset = nullptr) { + Type operandType = originalOperand.getType(); + if (operandType.isa()) { + MemRefDescriptor desc(convertedOperand); + *allocatedPtr = desc.allocatedPtr(rewriter, loc); + *alignedPtr = desc.alignedPtr(rewriter, loc); + if (offset != nullptr) + *offset = desc.offset(rewriter, loc); + return; + } + + unsigned memorySpace = + operandType.cast().getMemorySpaceAsInt(); + Type elementType = operandType.cast().getElementType(); + Type llvmElementType = typeConverter.convertType(elementType); + Type elementPtrPtrType = LLVM::LLVMPointerType::get( + LLVM::LLVMPointerType::get(llvmElementType, memorySpace)); + + // Extract pointer to the underlying ranked memref descriptor and cast it to + // ElemType**. + UnrankedMemRefDescriptor unrankedDesc(convertedOperand); + Value underlyingDescPtr = unrankedDesc.memRefDescPtr(rewriter, loc); + + *allocatedPtr = UnrankedMemRefDescriptor::allocatedPtr( + rewriter, loc, underlyingDescPtr, elementPtrPtrType); + *alignedPtr = UnrankedMemRefDescriptor::alignedPtr( + rewriter, loc, typeConverter, underlyingDescPtr, elementPtrPtrType); + if (offset != nullptr) { + *offset = UnrankedMemRefDescriptor::offset( + rewriter, loc, typeConverter, underlyingDescPtr, elementPtrPtrType); + } +} + +struct MemRefReinterpretCastOpLowering + : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern< + memref::ReinterpretCastOp>::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::ReinterpretCastOp castOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + memref::ReinterpretCastOp::Adaptor adaptor(operands, + castOp->getAttrDictionary()); + Type srcType = castOp.source().getType(); + + Value descriptor; + if (failed(convertSourceMemRefToDescriptor(rewriter, srcType, castOp, + adaptor, &descriptor))) + return failure(); + rewriter.replaceOp(castOp, {descriptor}); + return success(); + } + +private: + LogicalResult convertSourceMemRefToDescriptor( + ConversionPatternRewriter &rewriter, Type srcType, + memref::ReinterpretCastOp castOp, + memref::ReinterpretCastOp::Adaptor adaptor, Value *descriptor) const { + MemRefType targetMemRefType = + castOp.getResult().getType().cast(); + auto llvmTargetDescriptorTy = typeConverter->convertType(targetMemRefType) + .dyn_cast_or_null(); + if (!llvmTargetDescriptorTy) + return failure(); + + // Create descriptor. + Location loc = castOp.getLoc(); + auto desc = MemRefDescriptor::undef(rewriter, loc, llvmTargetDescriptorTy); + + // Set allocated and aligned pointers. + Value allocatedPtr, alignedPtr; + extractPointersAndOffset(loc, rewriter, *getTypeConverter(), + castOp.source(), adaptor.source(), &allocatedPtr, + &alignedPtr); + desc.setAllocatedPtr(rewriter, loc, allocatedPtr); + desc.setAlignedPtr(rewriter, loc, alignedPtr); + + // Set offset. + if (castOp.isDynamicOffset(0)) + desc.setOffset(rewriter, loc, adaptor.offsets()[0]); + else + desc.setConstantOffset(rewriter, loc, castOp.getStaticOffset(0)); + + // Set sizes and strides. + unsigned dynSizeId = 0; + unsigned dynStrideId = 0; + for (unsigned i = 0, e = targetMemRefType.getRank(); i < e; ++i) { + if (castOp.isDynamicSize(i)) + desc.setSize(rewriter, loc, i, adaptor.sizes()[dynSizeId++]); + else + desc.setConstantSize(rewriter, loc, i, castOp.getStaticSize(i)); + + if (castOp.isDynamicStride(i)) + desc.setStride(rewriter, loc, i, adaptor.strides()[dynStrideId++]); + else + desc.setConstantStride(rewriter, loc, i, castOp.getStaticStride(i)); + } + *descriptor = desc; + return success(); + } +}; + +struct MemRefReshapeOpLowering + : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::ReshapeOp reshapeOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto *op = reshapeOp.getOperation(); + memref::ReshapeOp::Adaptor adaptor(operands, op->getAttrDictionary()); + Type srcType = reshapeOp.source().getType(); + + Value descriptor; + if (failed(convertSourceMemRefToDescriptor(rewriter, srcType, reshapeOp, + adaptor, &descriptor))) + return failure(); + rewriter.replaceOp(op, {descriptor}); + return success(); + } + +private: + LogicalResult + convertSourceMemRefToDescriptor(ConversionPatternRewriter &rewriter, + Type srcType, memref::ReshapeOp reshapeOp, + memref::ReshapeOp::Adaptor adaptor, + Value *descriptor) const { + // Conversion for statically-known shape args is performed via + // `memref_reinterpret_cast`. + auto shapeMemRefType = reshapeOp.shape().getType().cast(); + if (shapeMemRefType.hasStaticShape()) + return failure(); + + // The shape is a rank-1 tensor with unknown length. + Location loc = reshapeOp.getLoc(); + MemRefDescriptor shapeDesc(adaptor.shape()); + Value resultRank = shapeDesc.size(rewriter, loc, 0); + + // Extract address space and element type. + auto targetType = + reshapeOp.getResult().getType().cast(); + unsigned addressSpace = targetType.getMemorySpaceAsInt(); + Type elementType = targetType.getElementType(); + + // Create the unranked memref descriptor that holds the ranked one. The + // inner descriptor is allocated on stack. + auto targetDesc = UnrankedMemRefDescriptor::undef( + rewriter, loc, typeConverter->convertType(targetType)); + targetDesc.setRank(rewriter, loc, resultRank); + SmallVector sizes; + UnrankedMemRefDescriptor::computeSizes(rewriter, loc, *getTypeConverter(), + targetDesc, sizes); + Value underlyingDescPtr = rewriter.create( + loc, getVoidPtrType(), sizes.front(), llvm::None); + targetDesc.setMemRefDescPtr(rewriter, loc, underlyingDescPtr); + + // Extract pointers and offset from the source memref. + Value allocatedPtr, alignedPtr, offset; + extractPointersAndOffset(loc, rewriter, *getTypeConverter(), + reshapeOp.source(), adaptor.source(), + &allocatedPtr, &alignedPtr, &offset); + + // Set pointers and offset. + Type llvmElementType = typeConverter->convertType(elementType); + auto elementPtrPtrType = LLVM::LLVMPointerType::get( + LLVM::LLVMPointerType::get(llvmElementType, addressSpace)); + UnrankedMemRefDescriptor::setAllocatedPtr(rewriter, loc, underlyingDescPtr, + elementPtrPtrType, allocatedPtr); + UnrankedMemRefDescriptor::setAlignedPtr(rewriter, loc, *getTypeConverter(), + underlyingDescPtr, + elementPtrPtrType, alignedPtr); + UnrankedMemRefDescriptor::setOffset(rewriter, loc, *getTypeConverter(), + underlyingDescPtr, elementPtrPtrType, + offset); + + // Use the offset pointer as base for further addressing. Copy over the new + // shape and compute strides. For this, we create a loop from rank-1 to 0. + Value targetSizesBase = UnrankedMemRefDescriptor::sizeBasePtr( + rewriter, loc, *getTypeConverter(), underlyingDescPtr, + elementPtrPtrType); + Value targetStridesBase = UnrankedMemRefDescriptor::strideBasePtr( + rewriter, loc, *getTypeConverter(), targetSizesBase, resultRank); + Value shapeOperandPtr = shapeDesc.alignedPtr(rewriter, loc); + Value oneIndex = createIndexConstant(rewriter, loc, 1); + Value resultRankMinusOne = + rewriter.create(loc, resultRank, oneIndex); + + Block *initBlock = rewriter.getInsertionBlock(); + Type indexType = getTypeConverter()->getIndexType(); + Block::iterator remainingOpsIt = std::next(rewriter.getInsertionPoint()); + + Block *condBlock = rewriter.createBlock(initBlock->getParent(), {}, + {indexType, indexType}); + + // Move the remaining initBlock ops to condBlock. + Block *remainingBlock = rewriter.splitBlock(initBlock, remainingOpsIt); + rewriter.mergeBlocks(remainingBlock, condBlock, ValueRange()); + + rewriter.setInsertionPointToEnd(initBlock); + rewriter.create(loc, ValueRange({resultRankMinusOne, oneIndex}), + condBlock); + rewriter.setInsertionPointToStart(condBlock); + Value indexArg = condBlock->getArgument(0); + Value strideArg = condBlock->getArgument(1); + + Value zeroIndex = createIndexConstant(rewriter, loc, 0); + Value pred = rewriter.create( + loc, IntegerType::get(rewriter.getContext(), 1), + LLVM::ICmpPredicate::sge, indexArg, zeroIndex); + + Block *bodyBlock = + rewriter.splitBlock(condBlock, rewriter.getInsertionPoint()); + rewriter.setInsertionPointToStart(bodyBlock); + + // Copy size from shape to descriptor. + Type llvmIndexPtrType = LLVM::LLVMPointerType::get(indexType); + Value sizeLoadGep = rewriter.create( + loc, llvmIndexPtrType, shapeOperandPtr, ValueRange{indexArg}); + Value size = rewriter.create(loc, sizeLoadGep); + UnrankedMemRefDescriptor::setSize(rewriter, loc, *getTypeConverter(), + targetSizesBase, indexArg, size); + + // Write stride value and compute next one. + UnrankedMemRefDescriptor::setStride(rewriter, loc, *getTypeConverter(), + targetStridesBase, indexArg, strideArg); + Value nextStride = rewriter.create(loc, strideArg, size); + + // Decrement loop counter and branch back. + Value decrement = rewriter.create(loc, indexArg, oneIndex); + rewriter.create(loc, ValueRange({decrement, nextStride}), + condBlock); + + Block *remainder = + rewriter.splitBlock(bodyBlock, rewriter.getInsertionPoint()); + + // Hook up the cond exit to the remainder. + rewriter.setInsertionPointToEnd(condBlock); + rewriter.create(loc, pred, bodyBlock, llvm::None, remainder, + llvm::None); + + // Reset position to beginning of new remainder block. + rewriter.setInsertionPointToStart(remainder); + + *descriptor = targetDesc; + return success(); + } +}; + +/// Conversion pattern that transforms a subview op into: +/// 1. An `llvm.mlir.undef` operation to create a memref descriptor +/// 2. Updates to the descriptor to introduce the data ptr, offset, size +/// and stride. +/// The subview op is replaced by the descriptor. +struct SubViewOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::SubViewOp subViewOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto loc = subViewOp.getLoc(); + + auto sourceMemRefType = subViewOp.source().getType().cast(); + auto sourceElementTy = + typeConverter->convertType(sourceMemRefType.getElementType()); + + auto viewMemRefType = subViewOp.getType(); + auto inferredType = memref::SubViewOp::inferResultType( + subViewOp.getSourceType(), + extractFromI64ArrayAttr(subViewOp.static_offsets()), + extractFromI64ArrayAttr(subViewOp.static_sizes()), + extractFromI64ArrayAttr(subViewOp.static_strides())) + .cast(); + auto targetElementTy = + typeConverter->convertType(viewMemRefType.getElementType()); + auto targetDescTy = typeConverter->convertType(viewMemRefType); + if (!sourceElementTy || !targetDescTy || !targetElementTy || + !LLVM::isCompatibleType(sourceElementTy) || + !LLVM::isCompatibleType(targetElementTy) || + !LLVM::isCompatibleType(targetDescTy)) + return failure(); + + // Extract the offset and strides from the type. + int64_t offset; + SmallVector strides; + auto successStrides = getStridesAndOffset(inferredType, strides, offset); + if (failed(successStrides)) + return failure(); + + // Create the descriptor. + if (!LLVM::isCompatibleType(operands.front().getType())) + return failure(); + MemRefDescriptor sourceMemRef(operands.front()); + auto targetMemRef = MemRefDescriptor::undef(rewriter, loc, targetDescTy); + + // Copy the buffer pointer from the old descriptor to the new one. + Value extracted = sourceMemRef.allocatedPtr(rewriter, loc); + Value bitcastPtr = rewriter.create( + loc, + LLVM::LLVMPointerType::get(targetElementTy, + viewMemRefType.getMemorySpaceAsInt()), + extracted); + targetMemRef.setAllocatedPtr(rewriter, loc, bitcastPtr); + + // Copy the aligned pointer from the old descriptor to the new one. + extracted = sourceMemRef.alignedPtr(rewriter, loc); + bitcastPtr = rewriter.create( + loc, + LLVM::LLVMPointerType::get(targetElementTy, + viewMemRefType.getMemorySpaceAsInt()), + extracted); + targetMemRef.setAlignedPtr(rewriter, loc, bitcastPtr); + + auto shape = viewMemRefType.getShape(); + auto inferredShape = inferredType.getShape(); + size_t inferredShapeRank = inferredShape.size(); + size_t resultShapeRank = shape.size(); + llvm::SmallDenseSet unusedDims = + computeRankReductionMask(inferredShape, shape).getValue(); + + // Extract strides needed to compute offset. + SmallVector strideValues; + strideValues.reserve(inferredShapeRank); + for (unsigned i = 0; i < inferredShapeRank; ++i) + strideValues.push_back(sourceMemRef.stride(rewriter, loc, i)); + + // Offset. + auto llvmIndexType = typeConverter->convertType(rewriter.getIndexType()); + if (!ShapedType::isDynamicStrideOrOffset(offset)) { + targetMemRef.setConstantOffset(rewriter, loc, offset); + } else { + Value baseOffset = sourceMemRef.offset(rewriter, loc); + // `inferredShapeRank` may be larger than the number of offset operands + // because of trailing semantics. In this case, the offset is guaranteed + // to be interpreted as 0 and we can just skip the extra dimensions. + for (unsigned i = 0, e = std::min(inferredShapeRank, + subViewOp.getMixedOffsets().size()); + i < e; ++i) { + Value offset = + // TODO: need OpFoldResult ODS adaptor to clean this up. + subViewOp.isDynamicOffset(i) + ? operands[subViewOp.getIndexOfDynamicOffset(i)] + : rewriter.create( + loc, llvmIndexType, + rewriter.getI64IntegerAttr(subViewOp.getStaticOffset(i))); + Value mul = rewriter.create(loc, offset, strideValues[i]); + baseOffset = rewriter.create(loc, baseOffset, mul); + } + targetMemRef.setOffset(rewriter, loc, baseOffset); + } + + // Update sizes and strides. + SmallVector mixedSizes = subViewOp.getMixedSizes(); + SmallVector mixedStrides = subViewOp.getMixedStrides(); + assert(mixedSizes.size() == mixedStrides.size() && + "expected sizes and strides of equal length"); + for (int i = inferredShapeRank - 1, j = resultShapeRank - 1; + i >= 0 && j >= 0; --i) { + if (unusedDims.contains(i)) + continue; + + // `i` may overflow subViewOp.getMixedSizes because of trailing semantics. + // In this case, the size is guaranteed to be interpreted as Dim and the + // stride as 1. + Value size, stride; + if (static_cast(i) >= mixedSizes.size()) { + // If the static size is available, use it directly. This is similar to + // the folding of dim(constant-op) but removes the need for dim to be + // aware of LLVM constants and for this pass to be aware of std + // constants. + int64_t staticSize = + subViewOp.source().getType().cast().getShape()[i]; + if (staticSize != ShapedType::kDynamicSize) { + size = rewriter.create( + loc, llvmIndexType, rewriter.getI64IntegerAttr(staticSize)); + } else { + Value pos = rewriter.create( + loc, llvmIndexType, rewriter.getI64IntegerAttr(i)); + size = rewriter.create( + loc, llvmIndexType, + rewriter.create(loc, subViewOp.source(), pos)); + } + stride = rewriter.create( + loc, llvmIndexType, rewriter.getI64IntegerAttr(1)); + } else { + // TODO: need OpFoldResult ODS adaptor to clean this up. + size = + subViewOp.isDynamicSize(i) + ? operands[subViewOp.getIndexOfDynamicSize(i)] + : rewriter.create( + loc, llvmIndexType, + rewriter.getI64IntegerAttr(subViewOp.getStaticSize(i))); + if (!ShapedType::isDynamicStrideOrOffset(strides[i])) { + stride = rewriter.create( + loc, llvmIndexType, rewriter.getI64IntegerAttr(strides[i])); + } else { + stride = subViewOp.isDynamicStride(i) + ? operands[subViewOp.getIndexOfDynamicStride(i)] + : rewriter.create( + loc, llvmIndexType, + rewriter.getI64IntegerAttr( + subViewOp.getStaticStride(i))); + stride = rewriter.create(loc, stride, strideValues[i]); + } + } + targetMemRef.setSize(rewriter, loc, j, size); + targetMemRef.setStride(rewriter, loc, j, stride); + j--; + } + + rewriter.replaceOp(subViewOp, {targetMemRef}); + return success(); + } +}; + +/// Conversion pattern that transforms a transpose op into: +/// 1. A function entry `alloca` operation to allocate a ViewDescriptor. +/// 2. A load of the ViewDescriptor from the pointer allocated in 1. +/// 3. Updates to the ViewDescriptor to introduce the data ptr, offset, size +/// and stride. Size and stride are permutations of the original values. +/// 4. A store of the resulting ViewDescriptor to the alloca'ed pointer. +/// The transpose op is replaced by the alloca'ed pointer. +class TransposeOpLowering : public ConvertOpToLLVMPattern { +public: + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + LogicalResult + matchAndRewrite(memref::TransposeOp transposeOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto loc = transposeOp.getLoc(); + memref::TransposeOpAdaptor adaptor(operands); + MemRefDescriptor viewMemRef(adaptor.in()); + + // No permutation, early exit. + if (transposeOp.permutation().isIdentity()) + return rewriter.replaceOp(transposeOp, {viewMemRef}), success(); + + auto targetMemRef = MemRefDescriptor::undef( + rewriter, loc, typeConverter->convertType(transposeOp.getShapedType())); + + // Copy the base and aligned pointers from the old descriptor to the new + // one. + targetMemRef.setAllocatedPtr(rewriter, loc, + viewMemRef.allocatedPtr(rewriter, loc)); + targetMemRef.setAlignedPtr(rewriter, loc, + viewMemRef.alignedPtr(rewriter, loc)); + + // Copy the offset pointer from the old descriptor to the new one. + targetMemRef.setOffset(rewriter, loc, viewMemRef.offset(rewriter, loc)); + + // Iterate over the dimensions and apply size/stride permutation. + for (auto en : llvm::enumerate(transposeOp.permutation().getResults())) { + int sourcePos = en.index(); + int targetPos = en.value().cast().getPosition(); + targetMemRef.setSize(rewriter, loc, targetPos, + viewMemRef.size(rewriter, loc, sourcePos)); + targetMemRef.setStride(rewriter, loc, targetPos, + viewMemRef.stride(rewriter, loc, sourcePos)); + } + + rewriter.replaceOp(transposeOp, {targetMemRef}); + return success(); + } +}; + +/// Conversion pattern that transforms an op into: +/// 1. An `llvm.mlir.undef` operation to create a memref descriptor +/// 2. Updates to the descriptor to introduce the data ptr, offset, size +/// and stride. +/// The view op is replaced by the descriptor. +struct ViewOpLowering : public ConvertOpToLLVMPattern { + using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; + + // Build and return the value for the idx^th shape dimension, either by + // returning the constant shape dimension or counting the proper dynamic size. + Value getSize(ConversionPatternRewriter &rewriter, Location loc, + ArrayRef shape, ValueRange dynamicSizes, + unsigned idx) const { + assert(idx < shape.size()); + if (!ShapedType::isDynamic(shape[idx])) + return createIndexConstant(rewriter, loc, shape[idx]); + // Count the number of dynamic dims in range [0, idx] + unsigned nDynamic = llvm::count_if(shape.take_front(idx), [](int64_t v) { + return ShapedType::isDynamic(v); + }); + return dynamicSizes[nDynamic]; + } + + // Build and return the idx^th stride, either by returning the constant stride + // or by computing the dynamic stride from the current `runningStride` and + // `nextSize`. The caller should keep a running stride and update it with the + // result returned by this function. + Value getStride(ConversionPatternRewriter &rewriter, Location loc, + ArrayRef strides, Value nextSize, + Value runningStride, unsigned idx) const { + assert(idx < strides.size()); + if (!MemRefType::isDynamicStrideOrOffset(strides[idx])) + return createIndexConstant(rewriter, loc, strides[idx]); + if (nextSize) + return runningStride + ? rewriter.create(loc, runningStride, nextSize) + : nextSize; + assert(!runningStride); + return createIndexConstant(rewriter, loc, 1); + } + + LogicalResult + matchAndRewrite(memref::ViewOp viewOp, ArrayRef operands, + ConversionPatternRewriter &rewriter) const override { + auto loc = viewOp.getLoc(); + memref::ViewOpAdaptor adaptor(operands); + + auto viewMemRefType = viewOp.getType(); + auto targetElementTy = + typeConverter->convertType(viewMemRefType.getElementType()); + auto targetDescTy = typeConverter->convertType(viewMemRefType); + if (!targetDescTy || !targetElementTy || + !LLVM::isCompatibleType(targetElementTy) || + !LLVM::isCompatibleType(targetDescTy)) + return viewOp.emitWarning("Target descriptor type not converted to LLVM"), + failure(); + + int64_t offset; + SmallVector strides; + auto successStrides = getStridesAndOffset(viewMemRefType, strides, offset); + if (failed(successStrides)) + return viewOp.emitWarning("cannot cast to non-strided shape"), failure(); + assert(offset == 0 && "expected offset to be 0"); + + // Create the descriptor. + MemRefDescriptor sourceMemRef(adaptor.source()); + auto targetMemRef = MemRefDescriptor::undef(rewriter, loc, targetDescTy); + + // Field 1: Copy the allocated pointer, used for malloc/free. + Value allocatedPtr = sourceMemRef.allocatedPtr(rewriter, loc); + auto srcMemRefType = viewOp.source().getType().cast(); + Value bitcastPtr = rewriter.create( + loc, + LLVM::LLVMPointerType::get(targetElementTy, + srcMemRefType.getMemorySpaceAsInt()), + allocatedPtr); + targetMemRef.setAllocatedPtr(rewriter, loc, bitcastPtr); + + // Field 2: Copy the actual aligned pointer to payload. + Value alignedPtr = sourceMemRef.alignedPtr(rewriter, loc); + alignedPtr = rewriter.create(loc, alignedPtr.getType(), + alignedPtr, adaptor.byte_shift()); + bitcastPtr = rewriter.create( + loc, + LLVM::LLVMPointerType::get(targetElementTy, + srcMemRefType.getMemorySpaceAsInt()), + alignedPtr); + targetMemRef.setAlignedPtr(rewriter, loc, bitcastPtr); + + // Field 3: The offset in the resulting type must be 0. This is because of + // the type change: an offset on srcType* may not be expressible as an + // offset on dstType*. + targetMemRef.setOffset(rewriter, loc, + createIndexConstant(rewriter, loc, offset)); + + // Early exit for 0-D corner case. + if (viewMemRefType.getRank() == 0) + return rewriter.replaceOp(viewOp, {targetMemRef}), success(); + + // Fields 4 and 5: Update sizes and strides. + if (strides.back() != 1) + return viewOp.emitWarning("cannot cast to non-contiguous shape"), + failure(); + Value stride = nullptr, nextSize = nullptr; + for (int i = viewMemRefType.getRank() - 1; i >= 0; --i) { + // Update size. + Value size = + getSize(rewriter, loc, viewMemRefType.getShape(), adaptor.sizes(), i); + targetMemRef.setSize(rewriter, loc, i, size); + // Update stride. + stride = getStride(rewriter, loc, strides, nextSize, stride, i); + targetMemRef.setStride(rewriter, loc, i, stride); + nextSize = size; + } + + rewriter.replaceOp(viewOp, {targetMemRef}); + return success(); + } +}; + +} // namespace + +void mlir::populateMemRefToLLVMConversionPatterns(LLVMTypeConverter &converter, + RewritePatternSet &patterns) { + // clang-format off + patterns.add< + AllocaOpLowering, + AllocOpLowering, + AllocaScopeOpLowering, + AssumeAlignmentOpLowering, + DimOpLowering, + DeallocOpLowering, + GlobalMemrefOpLowering, + GetGlobalMemrefOpLowering, + LoadOpLowering, + MemRefCastOpLowering, + MemRefCopyOpLowering, + MemRefReinterpretCastOpLowering, + MemRefReshapeOpLowering, + PrefetchOpLowering, + StoreOpLowering, + SubViewOpLowering, + TransposeOpLowering, + ViewOpLowering>(converter); + // clang-format on + auto allocLowering = converter.getOptions().allocLowering; + if (allocLowering == LowerToLLVMOptions::AllocLowering::AlignedAlloc) + patterns.add(converter); + else if (allocLowering == LowerToLLVMOptions::AllocLowering::Malloc) + patterns.add(converter); +} + +namespace { +struct MemRefToLLVMPass : public ConvertMemRefToLLVMBase { + MemRefToLLVMPass() = default; + + void runOnOperation() override { + Operation *op = getOperation(); + const auto &dataLayoutAnalysis = getAnalysis(); + LowerToLLVMOptions options(&getContext(), + dataLayoutAnalysis.getAtOrAbove(op)); + options.allocLowering = + (useAlignedAlloc ? LowerToLLVMOptions::AllocLowering::AlignedAlloc + : LowerToLLVMOptions::AllocLowering::Malloc); + if (indexBitwidth != kDeriveIndexBitwidthFromDataLayout) + options.overrideIndexBitwidth(indexBitwidth); + + LLVMTypeConverter typeConverter(&getContext(), options, + &dataLayoutAnalysis); + RewritePatternSet patterns(&getContext()); + populateMemRefToLLVMConversionPatterns(typeConverter, patterns); + LLVMConversionTarget target(getContext()); + target.addLegalOp(); + if (failed(applyPartialConversion(op, target, std::move(patterns)))) + signalPassFailure(); + } +}; +} // namespace + +std::unique_ptr mlir::createMemRefToLLVMPass() { + return std::make_unique(); +} diff --git a/mlir/lib/Conversion/OpenACCToLLVM/CMakeLists.txt b/mlir/lib/Conversion/OpenACCToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/OpenACCToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/OpenACCToLLVM/CMakeLists.txt @@ -10,7 +10,7 @@ LINK_LIBS PUBLIC MLIRIR MLIRLLVMIR + MLIRMemRefToLLVM MLIROpenACC - MLIRStandardToLLVM MLIRTransforms ) diff --git a/mlir/lib/Conversion/OpenACCToLLVM/OpenACCToLLVM.cpp b/mlir/lib/Conversion/OpenACCToLLVM/OpenACCToLLVM.cpp --- a/mlir/lib/Conversion/OpenACCToLLVM/OpenACCToLLVM.cpp +++ b/mlir/lib/Conversion/OpenACCToLLVM/OpenACCToLLVM.cpp @@ -7,9 +7,11 @@ //===----------------------------------------------------------------------===// #include "../PassDetail.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Conversion/OpenACCToLLVM/ConvertOpenACCToLLVM.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/OpenACC/OpenACC.h" +#include "mlir/IR/Builders.h" using namespace mlir; diff --git a/mlir/lib/Conversion/OpenMPToLLVM/CMakeLists.txt b/mlir/lib/Conversion/OpenMPToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/OpenMPToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/OpenMPToLLVM/CMakeLists.txt @@ -15,6 +15,7 @@ MLIRIR MLIRLLVMCommonConversion MLIRLLVMIR + MLIRMemRefToLLVM MLIROpenMP MLIRStandardToLLVM MLIRTransforms diff --git a/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp b/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp --- a/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp +++ b/mlir/lib/Conversion/OpenMPToLLVM/OpenMPToLLVM.cpp @@ -9,7 +9,9 @@ #include "mlir/Conversion/OpenMPToLLVM/ConvertOpenMPToLLVM.h" #include "../PassDetail.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" #include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" @@ -62,6 +64,7 @@ // Convert to OpenMP operations with LLVM IR dialect RewritePatternSet patterns(&getContext()); LLVMTypeConverter converter(&getContext()); + populateMemRefToLLVMConversionPatterns(converter, patterns); populateStdToLLVMConversionPatterns(converter, patterns); populateOpenMPToLLVMConversionPatterns(converter, patterns); diff --git a/mlir/lib/Conversion/SPIRVToLLVM/CMakeLists.txt b/mlir/lib/Conversion/SPIRVToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/SPIRVToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/SPIRVToLLVM/CMakeLists.txt @@ -16,6 +16,7 @@ MLIRSPIRVUtils MLIRLLVMCommonConversion MLIRLLVMIR + MLIRMemRefToLLVM MLIRStandardToLLVM MLIRIR MLIRTransforms diff --git a/mlir/lib/Conversion/SPIRVToLLVM/ConvertLaunchFuncToLLVMCalls.cpp b/mlir/lib/Conversion/SPIRVToLLVM/ConvertLaunchFuncToLLVMCalls.cpp --- a/mlir/lib/Conversion/SPIRVToLLVM/ConvertLaunchFuncToLLVMCalls.cpp +++ b/mlir/lib/Conversion/SPIRVToLLVM/ConvertLaunchFuncToLLVMCalls.cpp @@ -13,7 +13,9 @@ #include "../PassDetail.h" #include "mlir/Conversion/LLVMCommon/LoweringOptions.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Conversion/LLVMCommon/TypeConverter.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/SPIRVToLLVM/SPIRVToLLVM.h" #include "mlir/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" @@ -282,6 +284,7 @@ auto *context = module.getContext(); RewritePatternSet patterns(context); LLVMTypeConverter typeConverter(context, options); + populateMemRefToLLVMConversionPatterns(typeConverter, patterns); populateStdToLLVMConversionPatterns(typeConverter, patterns); patterns.add(typeConverter); diff --git a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp --- a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp +++ b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVM.cpp @@ -11,8 +11,8 @@ //===----------------------------------------------------------------------===// #include "mlir/Conversion/SPIRVToLLVM/SPIRVToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h" #include "mlir/Dialect/SPIRV/IR/SPIRVOps.h" diff --git a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.cpp b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.cpp --- a/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.cpp +++ b/mlir/lib/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.cpp @@ -12,9 +12,8 @@ #include "mlir/Conversion/SPIRVToLLVM/SPIRVToLLVMPass.h" #include "../PassDetail.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" #include "mlir/Conversion/SPIRVToLLVM/SPIRVToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h" diff --git a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt b/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt @@ -17,6 +17,5 @@ MLIRLLVMCommonConversion MLIRLLVMIR MLIRMath - MLIRMemRef MLIRTransforms ) diff --git a/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp b/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp --- a/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp +++ b/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp @@ -13,6 +13,7 @@ #include "../PassDetail.h" #include "mlir/Analysis/DataLayoutAnalysis.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" #include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Conversion/LLVMCommon/VectorPattern.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" @@ -20,7 +21,6 @@ #include "mlir/Dialect/LLVMIR/FunctionCallUtils.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/Math/IR/Math.h" -#include "mlir/Dialect/MemRef/IR/MemRef.h" #include "mlir/Dialect/StandardOps/IR/Ops.h" #include "mlir/Dialect/Utils/StaticValueUtils.h" #include "mlir/IR/Attributes.h" @@ -30,7 +30,6 @@ #include "mlir/IR/MLIRContext.h" #include "mlir/IR/PatternMatch.h" #include "mlir/IR/TypeUtilities.h" -#include "mlir/Interfaces/DataLayoutInterfaces.h" #include "mlir/Support/LogicalResult.h" #include "mlir/Support/MathExtras.h" #include "mlir/Transforms/DialectConversion.h" @@ -499,226 +498,6 @@ } }; -struct AllocOpLowering : public AllocLikeOpLLVMLowering { - AllocOpLowering(LLVMTypeConverter &converter) - : AllocLikeOpLLVMLowering(memref::AllocOp::getOperationName(), - converter) {} - - std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, - Location loc, Value sizeBytes, - Operation *op) const override { - // Heap allocations. - memref::AllocOp allocOp = cast(op); - MemRefType memRefType = allocOp.getType(); - - Value alignment; - if (auto alignmentAttr = allocOp.alignment()) { - alignment = createIndexConstant(rewriter, loc, *alignmentAttr); - } else if (!memRefType.getElementType().isSignlessIntOrIndexOrFloat()) { - // In the case where no alignment is specified, we may want to override - // `malloc's` behavior. `malloc` typically aligns at the size of the - // biggest scalar on a target HW. For non-scalars, use the natural - // alignment of the LLVM type given by the LLVM DataLayout. - alignment = getSizeInBytes(loc, memRefType.getElementType(), rewriter); - } - - if (alignment) { - // Adjust the allocation size to consider alignment. - sizeBytes = rewriter.create(loc, sizeBytes, alignment); - } - - // Allocate the underlying buffer and store a pointer to it in the MemRef - // descriptor. - Type elementPtrType = this->getElementPtrType(memRefType); - auto allocFuncOp = LLVM::lookupOrCreateMallocFn( - allocOp->getParentOfType(), getIndexType()); - auto results = createLLVMCall(rewriter, loc, allocFuncOp, {sizeBytes}, - getVoidPtrType()); - Value allocatedPtr = - rewriter.create(loc, elementPtrType, results[0]); - - Value alignedPtr = allocatedPtr; - if (alignment) { - // Compute the aligned type pointer. - Value allocatedInt = - rewriter.create(loc, getIndexType(), allocatedPtr); - Value alignmentInt = - createAligned(rewriter, loc, allocatedInt, alignment); - alignedPtr = - rewriter.create(loc, elementPtrType, alignmentInt); - } - - return std::make_tuple(allocatedPtr, alignedPtr); - } -}; - -struct AlignedAllocOpLowering : public AllocLikeOpLLVMLowering { - AlignedAllocOpLowering(LLVMTypeConverter &converter) - : AllocLikeOpLLVMLowering(memref::AllocOp::getOperationName(), - converter) {} - - /// Returns the memref's element size in bytes using the data layout active at - /// `op`. - // TODO: there are other places where this is used. Expose publicly? - unsigned getMemRefEltSizeInBytes(MemRefType memRefType, Operation *op) const { - const DataLayout *layout = &defaultLayout; - if (const DataLayoutAnalysis *analysis = - getTypeConverter()->getDataLayoutAnalysis()) { - layout = &analysis->getAbove(op); - } - Type elementType = memRefType.getElementType(); - if (auto memRefElementType = elementType.dyn_cast()) - return getTypeConverter()->getMemRefDescriptorSize(memRefElementType, - *layout); - if (auto memRefElementType = elementType.dyn_cast()) - return getTypeConverter()->getUnrankedMemRefDescriptorSize( - memRefElementType, *layout); - return layout->getTypeSize(elementType); - } - - /// Returns true if the memref size in bytes is known to be a multiple of - /// factor assuming the data layout active at `op`. - bool isMemRefSizeMultipleOf(MemRefType type, uint64_t factor, - Operation *op) const { - uint64_t sizeDivisor = getMemRefEltSizeInBytes(type, op); - for (unsigned i = 0, e = type.getRank(); i < e; i++) { - if (type.isDynamic(type.getDimSize(i))) - continue; - sizeDivisor = sizeDivisor * type.getDimSize(i); - } - return sizeDivisor % factor == 0; - } - - /// Returns the alignment to be used for the allocation call itself. - /// aligned_alloc requires the allocation size to be a power of two, and the - /// allocation size to be a multiple of alignment, - int64_t getAllocationAlignment(memref::AllocOp allocOp) const { - if (Optional alignment = allocOp.alignment()) - return *alignment; - - // Whenever we don't have alignment set, we will use an alignment - // consistent with the element type; since the allocation size has to be a - // power of two, we will bump to the next power of two if it already isn't. - auto eltSizeBytes = getMemRefEltSizeInBytes(allocOp.getType(), allocOp); - return std::max(kMinAlignedAllocAlignment, - llvm::PowerOf2Ceil(eltSizeBytes)); - } - - std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, - Location loc, Value sizeBytes, - Operation *op) const override { - // Heap allocations. - memref::AllocOp allocOp = cast(op); - MemRefType memRefType = allocOp.getType(); - int64_t alignment = getAllocationAlignment(allocOp); - Value allocAlignment = createIndexConstant(rewriter, loc, alignment); - - // aligned_alloc requires size to be a multiple of alignment; we will pad - // the size to the next multiple if necessary. - if (!isMemRefSizeMultipleOf(memRefType, alignment, op)) - sizeBytes = createAligned(rewriter, loc, sizeBytes, allocAlignment); - - Type elementPtrType = this->getElementPtrType(memRefType); - auto allocFuncOp = LLVM::lookupOrCreateAlignedAllocFn( - allocOp->getParentOfType(), getIndexType()); - auto results = - createLLVMCall(rewriter, loc, allocFuncOp, {allocAlignment, sizeBytes}, - getVoidPtrType()); - Value allocatedPtr = - rewriter.create(loc, elementPtrType, results[0]); - - return std::make_tuple(allocatedPtr, allocatedPtr); - } - - /// The minimum alignment to use with aligned_alloc (has to be a power of 2). - static constexpr uint64_t kMinAlignedAllocAlignment = 16UL; - - /// Default layout to use in absence of the corresponding analysis. - DataLayout defaultLayout; -}; - -// Out of line definition, required till C++17. -constexpr uint64_t AlignedAllocOpLowering::kMinAlignedAllocAlignment; - -struct AllocaOpLowering : public AllocLikeOpLLVMLowering { - AllocaOpLowering(LLVMTypeConverter &converter) - : AllocLikeOpLLVMLowering(memref::AllocaOp::getOperationName(), - converter) {} - - /// Allocates the underlying buffer using the right call. `allocatedBytePtr` - /// is set to null for stack allocations. `accessAlignment` is set if - /// alignment is needed post allocation (for eg. in conjunction with malloc). - std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, - Location loc, Value sizeBytes, - Operation *op) const override { - - // With alloca, one gets a pointer to the element type right away. - // For stack allocations. - auto allocaOp = cast(op); - auto elementPtrType = this->getElementPtrType(allocaOp.getType()); - - auto allocatedElementPtr = rewriter.create( - loc, elementPtrType, sizeBytes, - allocaOp.alignment() ? *allocaOp.alignment() : 0); - - return std::make_tuple(allocatedElementPtr, allocatedElementPtr); - } -}; - -struct AllocaScopeOpLowering - : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::AllocaScopeOp allocaScopeOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - OpBuilder::InsertionGuard guard(rewriter); - Location loc = allocaScopeOp.getLoc(); - - // Split the current block before the AllocaScopeOp to create the inlining - // point. - auto *currentBlock = rewriter.getInsertionBlock(); - auto *remainingOpsBlock = - rewriter.splitBlock(currentBlock, rewriter.getInsertionPoint()); - Block *continueBlock; - if (allocaScopeOp.getNumResults() == 0) { - continueBlock = remainingOpsBlock; - } else { - continueBlock = rewriter.createBlock(remainingOpsBlock, - allocaScopeOp.getResultTypes()); - rewriter.create(loc, remainingOpsBlock); - } - - // Inline body region. - Block *beforeBody = &allocaScopeOp.bodyRegion().front(); - Block *afterBody = &allocaScopeOp.bodyRegion().back(); - rewriter.inlineRegionBefore(allocaScopeOp.bodyRegion(), continueBlock); - - // Save stack and then branch into the body of the region. - rewriter.setInsertionPointToEnd(currentBlock); - auto stackSaveOp = - rewriter.create(loc, getVoidPtrType()); - rewriter.create(loc, beforeBody); - - // Replace the alloca_scope return with a branch that jumps out of the body. - // Stack restore before leaving the body region. - rewriter.setInsertionPointToEnd(afterBody); - auto returnOp = - cast(afterBody->getTerminator()); - auto branchOp = rewriter.replaceOpWithNewOp( - returnOp, continueBlock, returnOp.results()); - - // Insert stack restore before jumping out the body of the region. - rewriter.setInsertionPoint(branchOp); - rewriter.create(loc, stackSaveOp); - - // Replace the op with values return from the body region. - rewriter.replaceOp(allocaScopeOp, continueBlock->getArguments()); - - return success(); - } -}; - /// Copies the shaped descriptor part to (if `toDynamic` is set) or from /// (otherwise) the dynamically allocated memory for any operands that were /// unranked descriptors originally. @@ -877,130 +656,6 @@ using Super::Super; }; -// A `dealloc` is converted into a call to `free` on the underlying data buffer. -// The memref descriptor being an SSA value, there is no need to clean it up -// in any way. -struct DeallocOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - explicit DeallocOpLowering(LLVMTypeConverter &converter) - : ConvertOpToLLVMPattern(converter) {} - - LogicalResult - matchAndRewrite(memref::DeallocOp op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - assert(operands.size() == 1 && "dealloc takes one operand"); - memref::DeallocOp::Adaptor transformed(operands); - - // Insert the `free` declaration if it is not already present. - auto freeFunc = LLVM::lookupOrCreateFreeFn(op->getParentOfType()); - MemRefDescriptor memref(transformed.memref()); - Value casted = rewriter.create( - op.getLoc(), getVoidPtrType(), - memref.allocatedPtr(rewriter, op.getLoc())); - rewriter.replaceOpWithNewOp( - op, TypeRange(), rewriter.getSymbolRefAttr(freeFunc), casted); - return success(); - } -}; - -/// Returns the LLVM type of the global variable given the memref type `type`. -static Type convertGlobalMemrefTypeToLLVM(MemRefType type, - LLVMTypeConverter &typeConverter) { - // LLVM type for a global memref will be a multi-dimension array. For - // declarations or uninitialized global memrefs, we can potentially flatten - // this to a 1D array. However, for memref.global's with an initial value, - // we do not intend to flatten the ElementsAttribute when going from std -> - // LLVM dialect, so the LLVM type needs to me a multi-dimension array. - Type elementType = typeConverter.convertType(type.getElementType()); - Type arrayTy = elementType; - // Shape has the outermost dim at index 0, so need to walk it backwards - for (int64_t dim : llvm::reverse(type.getShape())) - arrayTy = LLVM::LLVMArrayType::get(arrayTy, dim); - return arrayTy; -} - -/// GlobalMemrefOp is lowered to a LLVM Global Variable. -struct GlobalMemrefOpLowering - : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::GlobalOp global, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - MemRefType type = global.type().cast(); - if (!isConvertibleAndHasIdentityMaps(type)) - return failure(); - - Type arrayTy = convertGlobalMemrefTypeToLLVM(type, *getTypeConverter()); - - LLVM::Linkage linkage = - global.isPublic() ? LLVM::Linkage::External : LLVM::Linkage::Private; - - Attribute initialValue = nullptr; - if (!global.isExternal() && !global.isUninitialized()) { - auto elementsAttr = global.initial_value()->cast(); - initialValue = elementsAttr; - - // For scalar memrefs, the global variable created is of the element type, - // so unpack the elements attribute to extract the value. - if (type.getRank() == 0) - initialValue = elementsAttr.getValue({}); - } - - rewriter.replaceOpWithNewOp( - global, arrayTy, global.constant(), linkage, global.sym_name(), - initialValue, /*alignment=*/0, type.getMemorySpaceAsInt()); - return success(); - } -}; - -/// GetGlobalMemrefOp is lowered into a Memref descriptor with the pointer to -/// the first element stashed into the descriptor. This reuses -/// `AllocLikeOpLowering` to reuse the Memref descriptor construction. -struct GetGlobalMemrefOpLowering : public AllocLikeOpLLVMLowering { - GetGlobalMemrefOpLowering(LLVMTypeConverter &converter) - : AllocLikeOpLLVMLowering(memref::GetGlobalOp::getOperationName(), - converter) {} - - /// Buffer "allocation" for memref.get_global op is getting the address of - /// the global variable referenced. - std::tuple allocateBuffer(ConversionPatternRewriter &rewriter, - Location loc, Value sizeBytes, - Operation *op) const override { - auto getGlobalOp = cast(op); - MemRefType type = getGlobalOp.result().getType().cast(); - unsigned memSpace = type.getMemorySpaceAsInt(); - - Type arrayTy = convertGlobalMemrefTypeToLLVM(type, *getTypeConverter()); - auto addressOf = rewriter.create( - loc, LLVM::LLVMPointerType::get(arrayTy, memSpace), getGlobalOp.name()); - - // Get the address of the first element in the array by creating a GEP with - // the address of the GV as the base, and (rank + 1) number of 0 indices. - Type elementType = typeConverter->convertType(type.getElementType()); - Type elementPtrType = LLVM::LLVMPointerType::get(elementType, memSpace); - - SmallVector operands = {addressOf}; - operands.insert(operands.end(), type.getRank() + 1, - createIndexConstant(rewriter, loc, 0)); - auto gep = rewriter.create(loc, elementPtrType, operands); - - // We do not expect the memref obtained using `memref.get_global` to be - // ever deallocated. Set the allocated pointer to be known bad value to - // help debug if that ever happens. - auto intPtrType = getIntPtrType(memSpace); - Value deadBeefConst = - createIndexAttrConstant(rewriter, op->getLoc(), intPtrType, 0xdeadbeef); - auto deadBeefPtr = - rewriter.create(loc, elementPtrType, deadBeefConst); - - // Both allocated and aligned pointers are same. We could potentially stash - // a nullptr for the allocated pointer since we do not expect any dealloc. - return std::make_tuple(deadBeefPtr, gep); - } -}; - // A `expm1` is converted into `exp - 1`. struct ExpM1OpLowering : public ConvertOpToLLVMPattern { using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; @@ -1164,417 +819,6 @@ } }; -struct MemRefCastOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult match(memref::CastOp memRefCastOp) const override { - Type srcType = memRefCastOp.getOperand().getType(); - Type dstType = memRefCastOp.getType(); - - // memref::CastOp reduce to bitcast in the ranked MemRef case and can be - // used for type erasure. For now they must preserve underlying element type - // and require source and result type to have the same rank. Therefore, - // perform a sanity check that the underlying structs are the same. Once op - // semantics are relaxed we can revisit. - if (srcType.isa() && dstType.isa()) - return success(typeConverter->convertType(srcType) == - typeConverter->convertType(dstType)); - - // At least one of the operands is unranked type - assert(srcType.isa() || - dstType.isa()); - - // Unranked to unranked cast is disallowed - return !(srcType.isa() && - dstType.isa()) - ? success() - : failure(); - } - - void rewrite(memref::CastOp memRefCastOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - memref::CastOp::Adaptor transformed(operands); - - auto srcType = memRefCastOp.getOperand().getType(); - auto dstType = memRefCastOp.getType(); - auto targetStructType = typeConverter->convertType(memRefCastOp.getType()); - auto loc = memRefCastOp.getLoc(); - - // For ranked/ranked case, just keep the original descriptor. - if (srcType.isa() && dstType.isa()) - return rewriter.replaceOp(memRefCastOp, {transformed.source()}); - - if (srcType.isa() && dstType.isa()) { - // Casting ranked to unranked memref type - // Set the rank in the destination from the memref type - // Allocate space on the stack and copy the src memref descriptor - // Set the ptr in the destination to the stack space - auto srcMemRefType = srcType.cast(); - int64_t rank = srcMemRefType.getRank(); - // ptr = AllocaOp sizeof(MemRefDescriptor) - auto ptr = getTypeConverter()->promoteOneMemRefDescriptor( - loc, transformed.source(), rewriter); - // voidptr = BitCastOp srcType* to void* - auto voidPtr = - rewriter.create(loc, getVoidPtrType(), ptr) - .getResult(); - // rank = ConstantOp srcRank - auto rankVal = rewriter.create( - loc, typeConverter->convertType(rewriter.getIntegerType(64)), - rewriter.getI64IntegerAttr(rank)); - // undef = UndefOp - UnrankedMemRefDescriptor memRefDesc = - UnrankedMemRefDescriptor::undef(rewriter, loc, targetStructType); - // d1 = InsertValueOp undef, rank, 0 - memRefDesc.setRank(rewriter, loc, rankVal); - // d2 = InsertValueOp d1, voidptr, 1 - memRefDesc.setMemRefDescPtr(rewriter, loc, voidPtr); - rewriter.replaceOp(memRefCastOp, (Value)memRefDesc); - - } else if (srcType.isa() && dstType.isa()) { - // Casting from unranked type to ranked. - // The operation is assumed to be doing a correct cast. If the destination - // type mismatches the unranked the type, it is undefined behavior. - UnrankedMemRefDescriptor memRefDesc(transformed.source()); - // ptr = ExtractValueOp src, 1 - auto ptr = memRefDesc.memRefDescPtr(rewriter, loc); - // castPtr = BitCastOp i8* to structTy* - auto castPtr = - rewriter - .create( - loc, LLVM::LLVMPointerType::get(targetStructType), ptr) - .getResult(); - // struct = LoadOp castPtr - auto loadOp = rewriter.create(loc, castPtr); - rewriter.replaceOp(memRefCastOp, loadOp.getResult()); - } else { - llvm_unreachable("Unsupported unranked memref to unranked memref cast"); - } - } -}; - -struct MemRefCopyOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::CopyOp op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto loc = op.getLoc(); - memref::CopyOp::Adaptor adaptor(operands); - auto srcType = op.source().getType().cast(); - auto targetType = op.target().getType().cast(); - - // First make sure we have an unranked memref descriptor representation. - auto makeUnranked = [&, this](Value ranked, BaseMemRefType type) { - auto rank = rewriter.create( - loc, getIndexType(), rewriter.getIndexAttr(type.getRank())); - auto *typeConverter = getTypeConverter(); - auto ptr = - typeConverter->promoteOneMemRefDescriptor(loc, ranked, rewriter); - auto voidPtr = - rewriter.create(loc, getVoidPtrType(), ptr) - .getResult(); - auto unrankedType = - UnrankedMemRefType::get(type.getElementType(), type.getMemorySpace()); - return UnrankedMemRefDescriptor::pack(rewriter, loc, *typeConverter, - unrankedType, - ValueRange{rank, voidPtr}); - }; - - Value unrankedSource = srcType.hasRank() - ? makeUnranked(adaptor.source(), srcType) - : adaptor.source(); - Value unrankedTarget = targetType.hasRank() - ? makeUnranked(adaptor.target(), targetType) - : adaptor.target(); - - // Now promote the unranked descriptors to the stack. - auto one = rewriter.create(loc, getIndexType(), - rewriter.getIndexAttr(1)); - auto promote = [&](Value desc) { - auto ptrType = LLVM::LLVMPointerType::get(desc.getType()); - auto allocated = - rewriter.create(loc, ptrType, ValueRange{one}); - rewriter.create(loc, desc, allocated); - return allocated; - }; - - auto sourcePtr = promote(unrankedSource); - auto targetPtr = promote(unrankedTarget); - - auto elemSize = rewriter.create( - loc, getIndexType(), - rewriter.getIndexAttr(srcType.getElementTypeBitWidth() / 8)); - auto copyFn = LLVM::lookupOrCreateMemRefCopyFn( - op->getParentOfType(), getIndexType(), sourcePtr.getType()); - rewriter.create(loc, copyFn, - ValueRange{elemSize, sourcePtr, targetPtr}); - rewriter.eraseOp(op); - - return success(); - } -}; - -/// Extracts allocated, aligned pointers and offset from a ranked or unranked -/// memref type. In unranked case, the fields are extracted from the underlying -/// ranked descriptor. -static void extractPointersAndOffset(Location loc, - ConversionPatternRewriter &rewriter, - LLVMTypeConverter &typeConverter, - Value originalOperand, - Value convertedOperand, - Value *allocatedPtr, Value *alignedPtr, - Value *offset = nullptr) { - Type operandType = originalOperand.getType(); - if (operandType.isa()) { - MemRefDescriptor desc(convertedOperand); - *allocatedPtr = desc.allocatedPtr(rewriter, loc); - *alignedPtr = desc.alignedPtr(rewriter, loc); - if (offset != nullptr) - *offset = desc.offset(rewriter, loc); - return; - } - - unsigned memorySpace = - operandType.cast().getMemorySpaceAsInt(); - Type elementType = operandType.cast().getElementType(); - Type llvmElementType = typeConverter.convertType(elementType); - Type elementPtrPtrType = LLVM::LLVMPointerType::get( - LLVM::LLVMPointerType::get(llvmElementType, memorySpace)); - - // Extract pointer to the underlying ranked memref descriptor and cast it to - // ElemType**. - UnrankedMemRefDescriptor unrankedDesc(convertedOperand); - Value underlyingDescPtr = unrankedDesc.memRefDescPtr(rewriter, loc); - - *allocatedPtr = UnrankedMemRefDescriptor::allocatedPtr( - rewriter, loc, underlyingDescPtr, elementPtrPtrType); - *alignedPtr = UnrankedMemRefDescriptor::alignedPtr( - rewriter, loc, typeConverter, underlyingDescPtr, elementPtrPtrType); - if (offset != nullptr) { - *offset = UnrankedMemRefDescriptor::offset( - rewriter, loc, typeConverter, underlyingDescPtr, elementPtrPtrType); - } -} - -struct MemRefReinterpretCastOpLowering - : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern< - memref::ReinterpretCastOp>::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::ReinterpretCastOp castOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - memref::ReinterpretCastOp::Adaptor adaptor(operands, - castOp->getAttrDictionary()); - Type srcType = castOp.source().getType(); - - Value descriptor; - if (failed(convertSourceMemRefToDescriptor(rewriter, srcType, castOp, - adaptor, &descriptor))) - return failure(); - rewriter.replaceOp(castOp, {descriptor}); - return success(); - } - -private: - LogicalResult convertSourceMemRefToDescriptor( - ConversionPatternRewriter &rewriter, Type srcType, - memref::ReinterpretCastOp castOp, - memref::ReinterpretCastOp::Adaptor adaptor, Value *descriptor) const { - MemRefType targetMemRefType = - castOp.getResult().getType().cast(); - auto llvmTargetDescriptorTy = typeConverter->convertType(targetMemRefType) - .dyn_cast_or_null(); - if (!llvmTargetDescriptorTy) - return failure(); - - // Create descriptor. - Location loc = castOp.getLoc(); - auto desc = MemRefDescriptor::undef(rewriter, loc, llvmTargetDescriptorTy); - - // Set allocated and aligned pointers. - Value allocatedPtr, alignedPtr; - extractPointersAndOffset(loc, rewriter, *getTypeConverter(), - castOp.source(), adaptor.source(), &allocatedPtr, - &alignedPtr); - desc.setAllocatedPtr(rewriter, loc, allocatedPtr); - desc.setAlignedPtr(rewriter, loc, alignedPtr); - - // Set offset. - if (castOp.isDynamicOffset(0)) - desc.setOffset(rewriter, loc, adaptor.offsets()[0]); - else - desc.setConstantOffset(rewriter, loc, castOp.getStaticOffset(0)); - - // Set sizes and strides. - unsigned dynSizeId = 0; - unsigned dynStrideId = 0; - for (unsigned i = 0, e = targetMemRefType.getRank(); i < e; ++i) { - if (castOp.isDynamicSize(i)) - desc.setSize(rewriter, loc, i, adaptor.sizes()[dynSizeId++]); - else - desc.setConstantSize(rewriter, loc, i, castOp.getStaticSize(i)); - - if (castOp.isDynamicStride(i)) - desc.setStride(rewriter, loc, i, adaptor.strides()[dynStrideId++]); - else - desc.setConstantStride(rewriter, loc, i, castOp.getStaticStride(i)); - } - *descriptor = desc; - return success(); - } -}; - -struct MemRefReshapeOpLowering - : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::ReshapeOp reshapeOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto *op = reshapeOp.getOperation(); - memref::ReshapeOp::Adaptor adaptor(operands, op->getAttrDictionary()); - Type srcType = reshapeOp.source().getType(); - - Value descriptor; - if (failed(convertSourceMemRefToDescriptor(rewriter, srcType, reshapeOp, - adaptor, &descriptor))) - return failure(); - rewriter.replaceOp(op, {descriptor}); - return success(); - } - -private: - LogicalResult - convertSourceMemRefToDescriptor(ConversionPatternRewriter &rewriter, - Type srcType, memref::ReshapeOp reshapeOp, - memref::ReshapeOp::Adaptor adaptor, - Value *descriptor) const { - // Conversion for statically-known shape args is performed via - // `memref_reinterpret_cast`. - auto shapeMemRefType = reshapeOp.shape().getType().cast(); - if (shapeMemRefType.hasStaticShape()) - return failure(); - - // The shape is a rank-1 tensor with unknown length. - Location loc = reshapeOp.getLoc(); - MemRefDescriptor shapeDesc(adaptor.shape()); - Value resultRank = shapeDesc.size(rewriter, loc, 0); - - // Extract address space and element type. - auto targetType = - reshapeOp.getResult().getType().cast(); - unsigned addressSpace = targetType.getMemorySpaceAsInt(); - Type elementType = targetType.getElementType(); - - // Create the unranked memref descriptor that holds the ranked one. The - // inner descriptor is allocated on stack. - auto targetDesc = UnrankedMemRefDescriptor::undef( - rewriter, loc, typeConverter->convertType(targetType)); - targetDesc.setRank(rewriter, loc, resultRank); - SmallVector sizes; - UnrankedMemRefDescriptor::computeSizes(rewriter, loc, *getTypeConverter(), - targetDesc, sizes); - Value underlyingDescPtr = rewriter.create( - loc, getVoidPtrType(), sizes.front(), llvm::None); - targetDesc.setMemRefDescPtr(rewriter, loc, underlyingDescPtr); - - // Extract pointers and offset from the source memref. - Value allocatedPtr, alignedPtr, offset; - extractPointersAndOffset(loc, rewriter, *getTypeConverter(), - reshapeOp.source(), adaptor.source(), - &allocatedPtr, &alignedPtr, &offset); - - // Set pointers and offset. - Type llvmElementType = typeConverter->convertType(elementType); - auto elementPtrPtrType = LLVM::LLVMPointerType::get( - LLVM::LLVMPointerType::get(llvmElementType, addressSpace)); - UnrankedMemRefDescriptor::setAllocatedPtr(rewriter, loc, underlyingDescPtr, - elementPtrPtrType, allocatedPtr); - UnrankedMemRefDescriptor::setAlignedPtr(rewriter, loc, *getTypeConverter(), - underlyingDescPtr, - elementPtrPtrType, alignedPtr); - UnrankedMemRefDescriptor::setOffset(rewriter, loc, *getTypeConverter(), - underlyingDescPtr, elementPtrPtrType, - offset); - - // Use the offset pointer as base for further addressing. Copy over the new - // shape and compute strides. For this, we create a loop from rank-1 to 0. - Value targetSizesBase = UnrankedMemRefDescriptor::sizeBasePtr( - rewriter, loc, *getTypeConverter(), underlyingDescPtr, - elementPtrPtrType); - Value targetStridesBase = UnrankedMemRefDescriptor::strideBasePtr( - rewriter, loc, *getTypeConverter(), targetSizesBase, resultRank); - Value shapeOperandPtr = shapeDesc.alignedPtr(rewriter, loc); - Value oneIndex = createIndexConstant(rewriter, loc, 1); - Value resultRankMinusOne = - rewriter.create(loc, resultRank, oneIndex); - - Block *initBlock = rewriter.getInsertionBlock(); - Type indexType = getTypeConverter()->getIndexType(); - Block::iterator remainingOpsIt = std::next(rewriter.getInsertionPoint()); - - Block *condBlock = rewriter.createBlock(initBlock->getParent(), {}, - {indexType, indexType}); - - // Iterate over the remaining ops in initBlock and move them to condBlock. - BlockAndValueMapping map; - for (auto it = remainingOpsIt, e = initBlock->end(); it != e; ++it) { - rewriter.clone(*it, map); - rewriter.eraseOp(&*it); - } - - rewriter.setInsertionPointToEnd(initBlock); - rewriter.create(loc, ValueRange({resultRankMinusOne, oneIndex}), - condBlock); - rewriter.setInsertionPointToStart(condBlock); - Value indexArg = condBlock->getArgument(0); - Value strideArg = condBlock->getArgument(1); - - Value zeroIndex = createIndexConstant(rewriter, loc, 0); - Value pred = rewriter.create( - loc, IntegerType::get(rewriter.getContext(), 1), - LLVM::ICmpPredicate::sge, indexArg, zeroIndex); - - Block *bodyBlock = - rewriter.splitBlock(condBlock, rewriter.getInsertionPoint()); - rewriter.setInsertionPointToStart(bodyBlock); - - // Copy size from shape to descriptor. - Type llvmIndexPtrType = LLVM::LLVMPointerType::get(indexType); - Value sizeLoadGep = rewriter.create( - loc, llvmIndexPtrType, shapeOperandPtr, ValueRange{indexArg}); - Value size = rewriter.create(loc, sizeLoadGep); - UnrankedMemRefDescriptor::setSize(rewriter, loc, *getTypeConverter(), - targetSizesBase, indexArg, size); - - // Write stride value and compute next one. - UnrankedMemRefDescriptor::setStride(rewriter, loc, *getTypeConverter(), - targetStridesBase, indexArg, strideArg); - Value nextStride = rewriter.create(loc, strideArg, size); - - // Decrement loop counter and branch back. - Value decrement = rewriter.create(loc, indexArg, oneIndex); - rewriter.create(loc, ValueRange({decrement, nextStride}), - condBlock); - - Block *remainder = - rewriter.splitBlock(bodyBlock, rewriter.getInsertionPoint()); - - // Hook up the cond exit to the remainder. - rewriter.setInsertionPointToEnd(condBlock); - rewriter.create(loc, pred, bodyBlock, llvm::None, remainder, - llvm::None); - - // Reset position to beginning of new remainder block. - rewriter.setInsertionPointToStart(remainder); - - *descriptor = targetDesc; - return success(); - } -}; - struct DialectCastOpLowering : public ConvertOpToLLVMPattern { using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; @@ -1592,96 +836,6 @@ } }; -// A `dim` is converted to a constant for static sizes and to an access to the -// size stored in the memref descriptor for dynamic sizes. -struct DimOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::DimOp dimOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - Type operandType = dimOp.source().getType(); - if (operandType.isa()) { - rewriter.replaceOp(dimOp, {extractSizeOfUnrankedMemRef( - operandType, dimOp, operands, rewriter)}); - - return success(); - } - if (operandType.isa()) { - rewriter.replaceOp(dimOp, {extractSizeOfRankedMemRef( - operandType, dimOp, operands, rewriter)}); - return success(); - } - llvm_unreachable("expected MemRefType or UnrankedMemRefType"); - } - -private: - Value extractSizeOfUnrankedMemRef(Type operandType, memref::DimOp dimOp, - ArrayRef operands, - ConversionPatternRewriter &rewriter) const { - Location loc = dimOp.getLoc(); - memref::DimOp::Adaptor transformed(operands); - - auto unrankedMemRefType = operandType.cast(); - auto scalarMemRefType = - MemRefType::get({}, unrankedMemRefType.getElementType()); - unsigned addressSpace = unrankedMemRefType.getMemorySpaceAsInt(); - - // Extract pointer to the underlying ranked descriptor and bitcast it to a - // memref descriptor pointer to minimize the number of GEP - // operations. - UnrankedMemRefDescriptor unrankedDesc(transformed.source()); - Value underlyingRankedDesc = unrankedDesc.memRefDescPtr(rewriter, loc); - Value scalarMemRefDescPtr = rewriter.create( - loc, - LLVM::LLVMPointerType::get(typeConverter->convertType(scalarMemRefType), - addressSpace), - underlyingRankedDesc); - - // Get pointer to offset field of memref descriptor. - Type indexPtrTy = LLVM::LLVMPointerType::get( - getTypeConverter()->getIndexType(), addressSpace); - Value two = rewriter.create( - loc, typeConverter->convertType(rewriter.getI32Type()), - rewriter.getI32IntegerAttr(2)); - Value offsetPtr = rewriter.create( - loc, indexPtrTy, scalarMemRefDescPtr, - ValueRange({createIndexConstant(rewriter, loc, 0), two})); - - // The size value that we have to extract can be obtained using GEPop with - // `dimOp.index() + 1` index argument. - Value idxPlusOne = rewriter.create( - loc, createIndexConstant(rewriter, loc, 1), transformed.index()); - Value sizePtr = rewriter.create(loc, indexPtrTy, offsetPtr, - ValueRange({idxPlusOne})); - return rewriter.create(loc, sizePtr); - } - - Value extractSizeOfRankedMemRef(Type operandType, memref::DimOp dimOp, - ArrayRef operands, - ConversionPatternRewriter &rewriter) const { - Location loc = dimOp.getLoc(); - memref::DimOp::Adaptor transformed(operands); - // Take advantage if index is constant. - MemRefType memRefType = operandType.cast(); - if (Optional index = dimOp.getConstantIndex()) { - int64_t i = index.getValue(); - if (memRefType.isDynamicDim(i)) { - // extract dynamic size from the memref descriptor. - MemRefDescriptor descriptor(transformed.source()); - return descriptor.size(rewriter, loc, i); - } - // Use constant for static size. - int64_t dimSize = memRefType.getDimSize(i); - return createIndexConstant(rewriter, loc, dimSize); - } - Value index = dimOp.index(); - int64_t rank = memRefType.getRank(); - MemRefDescriptor memrefDescriptor(transformed.source()); - return memrefDescriptor.size(rewriter, loc, index, rank); - } -}; - struct RankOpLowering : public ConvertOpToLLVMPattern { using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; @@ -1719,76 +873,6 @@ } }; -// Load operation is lowered to obtaining a pointer to the indexed element -// and loading it. -struct LoadOpLowering : public LoadStoreOpLowering { - using Base::Base; - - LogicalResult - matchAndRewrite(memref::LoadOp loadOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - memref::LoadOp::Adaptor transformed(operands); - auto type = loadOp.getMemRefType(); - - Value dataPtr = - getStridedElementPtr(loadOp.getLoc(), type, transformed.memref(), - transformed.indices(), rewriter); - rewriter.replaceOpWithNewOp(loadOp, dataPtr); - return success(); - } -}; - -// Store operation is lowered to obtaining a pointer to the indexed element, -// and storing the given value to it. -struct StoreOpLowering : public LoadStoreOpLowering { - using Base::Base; - - LogicalResult - matchAndRewrite(memref::StoreOp op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto type = op.getMemRefType(); - memref::StoreOp::Adaptor transformed(operands); - - Value dataPtr = - getStridedElementPtr(op.getLoc(), type, transformed.memref(), - transformed.indices(), rewriter); - rewriter.replaceOpWithNewOp(op, transformed.value(), - dataPtr); - return success(); - } -}; - -// The prefetch operation is lowered in a way similar to the load operation -// except that the llvm.prefetch operation is used for replacement. -struct PrefetchOpLowering : public LoadStoreOpLowering { - using Base::Base; - - LogicalResult - matchAndRewrite(memref::PrefetchOp prefetchOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - memref::PrefetchOp::Adaptor transformed(operands); - auto type = prefetchOp.getMemRefType(); - auto loc = prefetchOp.getLoc(); - - Value dataPtr = getStridedElementPtr(loc, type, transformed.memref(), - transformed.indices(), rewriter); - - // Replace with llvm.prefetch. - auto llvmI32Type = typeConverter->convertType(rewriter.getIntegerType(32)); - auto isWrite = rewriter.create( - loc, llvmI32Type, rewriter.getI32IntegerAttr(prefetchOp.isWrite())); - auto localityHint = rewriter.create( - loc, llvmI32Type, - rewriter.getI32IntegerAttr(prefetchOp.localityHint())); - auto isData = rewriter.create( - loc, llvmI32Type, rewriter.getI32IntegerAttr(prefetchOp.isDataCache())); - - rewriter.replaceOpWithNewOp(prefetchOp, dataPtr, isWrite, - localityHint, isData); - return success(); - } -}; - // The lowering of index_cast becomes an integer conversion since index becomes // an integer. If the bit width of the source and target integer types is the // same, just erase the cast. If the target type is wider, sign-extend the @@ -2090,369 +1174,6 @@ } }; -/// Conversion pattern that transforms a subview op into: -/// 1. An `llvm.mlir.undef` operation to create a memref descriptor -/// 2. Updates to the descriptor to introduce the data ptr, offset, size -/// and stride. -/// The subview op is replaced by the descriptor. -struct SubViewOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::SubViewOp subViewOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto loc = subViewOp.getLoc(); - - auto sourceMemRefType = subViewOp.source().getType().cast(); - auto sourceElementTy = - typeConverter->convertType(sourceMemRefType.getElementType()); - - auto viewMemRefType = subViewOp.getType(); - auto inferredType = memref::SubViewOp::inferResultType( - subViewOp.getSourceType(), - extractFromI64ArrayAttr(subViewOp.static_offsets()), - extractFromI64ArrayAttr(subViewOp.static_sizes()), - extractFromI64ArrayAttr(subViewOp.static_strides())) - .cast(); - auto targetElementTy = - typeConverter->convertType(viewMemRefType.getElementType()); - auto targetDescTy = typeConverter->convertType(viewMemRefType); - if (!sourceElementTy || !targetDescTy || !targetElementTy || - !LLVM::isCompatibleType(sourceElementTy) || - !LLVM::isCompatibleType(targetElementTy) || - !LLVM::isCompatibleType(targetDescTy)) - return failure(); - - // Extract the offset and strides from the type. - int64_t offset; - SmallVector strides; - auto successStrides = getStridesAndOffset(inferredType, strides, offset); - if (failed(successStrides)) - return failure(); - - // Create the descriptor. - if (!LLVM::isCompatibleType(operands.front().getType())) - return failure(); - MemRefDescriptor sourceMemRef(operands.front()); - auto targetMemRef = MemRefDescriptor::undef(rewriter, loc, targetDescTy); - - // Copy the buffer pointer from the old descriptor to the new one. - Value extracted = sourceMemRef.allocatedPtr(rewriter, loc); - Value bitcastPtr = rewriter.create( - loc, - LLVM::LLVMPointerType::get(targetElementTy, - viewMemRefType.getMemorySpaceAsInt()), - extracted); - targetMemRef.setAllocatedPtr(rewriter, loc, bitcastPtr); - - // Copy the aligned pointer from the old descriptor to the new one. - extracted = sourceMemRef.alignedPtr(rewriter, loc); - bitcastPtr = rewriter.create( - loc, - LLVM::LLVMPointerType::get(targetElementTy, - viewMemRefType.getMemorySpaceAsInt()), - extracted); - targetMemRef.setAlignedPtr(rewriter, loc, bitcastPtr); - - auto shape = viewMemRefType.getShape(); - auto inferredShape = inferredType.getShape(); - size_t inferredShapeRank = inferredShape.size(); - size_t resultShapeRank = shape.size(); - llvm::SmallDenseSet unusedDims = - computeRankReductionMask(inferredShape, shape).getValue(); - - // Extract strides needed to compute offset. - SmallVector strideValues; - strideValues.reserve(inferredShapeRank); - for (unsigned i = 0; i < inferredShapeRank; ++i) - strideValues.push_back(sourceMemRef.stride(rewriter, loc, i)); - - // Offset. - auto llvmIndexType = typeConverter->convertType(rewriter.getIndexType()); - if (!ShapedType::isDynamicStrideOrOffset(offset)) { - targetMemRef.setConstantOffset(rewriter, loc, offset); - } else { - Value baseOffset = sourceMemRef.offset(rewriter, loc); - // `inferredShapeRank` may be larger than the number of offset operands - // because of trailing semantics. In this case, the offset is guaranteed - // to be interpreted as 0 and we can just skip the extra dimensions. - for (unsigned i = 0, e = std::min(inferredShapeRank, - subViewOp.getMixedOffsets().size()); - i < e; ++i) { - Value offset = - // TODO: need OpFoldResult ODS adaptor to clean this up. - subViewOp.isDynamicOffset(i) - ? operands[subViewOp.getIndexOfDynamicOffset(i)] - : rewriter.create( - loc, llvmIndexType, - rewriter.getI64IntegerAttr(subViewOp.getStaticOffset(i))); - Value mul = rewriter.create(loc, offset, strideValues[i]); - baseOffset = rewriter.create(loc, baseOffset, mul); - } - targetMemRef.setOffset(rewriter, loc, baseOffset); - } - - // Update sizes and strides. - SmallVector mixedSizes = subViewOp.getMixedSizes(); - SmallVector mixedStrides = subViewOp.getMixedStrides(); - assert(mixedSizes.size() == mixedStrides.size() && - "expected sizes and strides of equal length"); - for (int i = inferredShapeRank - 1, j = resultShapeRank - 1; - i >= 0 && j >= 0; --i) { - if (unusedDims.contains(i)) - continue; - - // `i` may overflow subViewOp.getMixedSizes because of trailing semantics. - // In this case, the size is guaranteed to be interpreted as Dim and the - // stride as 1. - Value size, stride; - if (static_cast(i) >= mixedSizes.size()) { - size = rewriter.create( - loc, llvmIndexType, - rewriter.create(loc, subViewOp.source(), i)); - stride = rewriter.create( - loc, llvmIndexType, rewriter.getI64IntegerAttr(1)); - } else { - // TODO: need OpFoldResult ODS adaptor to clean this up. - size = - subViewOp.isDynamicSize(i) - ? operands[subViewOp.getIndexOfDynamicSize(i)] - : rewriter.create( - loc, llvmIndexType, - rewriter.getI64IntegerAttr(subViewOp.getStaticSize(i))); - if (!ShapedType::isDynamicStrideOrOffset(strides[i])) { - stride = rewriter.create( - loc, llvmIndexType, rewriter.getI64IntegerAttr(strides[i])); - } else { - stride = subViewOp.isDynamicStride(i) - ? operands[subViewOp.getIndexOfDynamicStride(i)] - : rewriter.create( - loc, llvmIndexType, - rewriter.getI64IntegerAttr( - subViewOp.getStaticStride(i))); - stride = rewriter.create(loc, stride, strideValues[i]); - } - } - targetMemRef.setSize(rewriter, loc, j, size); - targetMemRef.setStride(rewriter, loc, j, stride); - j--; - } - - rewriter.replaceOp(subViewOp, {targetMemRef}); - return success(); - } -}; - -/// Conversion pattern that transforms a transpose op into: -/// 1. A function entry `alloca` operation to allocate a ViewDescriptor. -/// 2. A load of the ViewDescriptor from the pointer allocated in 1. -/// 3. Updates to the ViewDescriptor to introduce the data ptr, offset, size -/// and stride. Size and stride are permutations of the original values. -/// 4. A store of the resulting ViewDescriptor to the alloca'ed pointer. -/// The transpose op is replaced by the alloca'ed pointer. -class TransposeOpLowering : public ConvertOpToLLVMPattern { -public: - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::TransposeOp transposeOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto loc = transposeOp.getLoc(); - memref::TransposeOpAdaptor adaptor(operands); - MemRefDescriptor viewMemRef(adaptor.in()); - - // No permutation, early exit. - if (transposeOp.permutation().isIdentity()) - return rewriter.replaceOp(transposeOp, {viewMemRef}), success(); - - auto targetMemRef = MemRefDescriptor::undef( - rewriter, loc, typeConverter->convertType(transposeOp.getShapedType())); - - // Copy the base and aligned pointers from the old descriptor to the new - // one. - targetMemRef.setAllocatedPtr(rewriter, loc, - viewMemRef.allocatedPtr(rewriter, loc)); - targetMemRef.setAlignedPtr(rewriter, loc, - viewMemRef.alignedPtr(rewriter, loc)); - - // Copy the offset pointer from the old descriptor to the new one. - targetMemRef.setOffset(rewriter, loc, viewMemRef.offset(rewriter, loc)); - - // Iterate over the dimensions and apply size/stride permutation. - for (auto en : llvm::enumerate(transposeOp.permutation().getResults())) { - int sourcePos = en.index(); - int targetPos = en.value().cast().getPosition(); - targetMemRef.setSize(rewriter, loc, targetPos, - viewMemRef.size(rewriter, loc, sourcePos)); - targetMemRef.setStride(rewriter, loc, targetPos, - viewMemRef.stride(rewriter, loc, sourcePos)); - } - - rewriter.replaceOp(transposeOp, {targetMemRef}); - return success(); - } -}; - -/// Conversion pattern that transforms an op into: -/// 1. An `llvm.mlir.undef` operation to create a memref descriptor -/// 2. Updates to the descriptor to introduce the data ptr, offset, size -/// and stride. -/// The view op is replaced by the descriptor. -struct ViewOpLowering : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern::ConvertOpToLLVMPattern; - - // Build and return the value for the idx^th shape dimension, either by - // returning the constant shape dimension or counting the proper dynamic size. - Value getSize(ConversionPatternRewriter &rewriter, Location loc, - ArrayRef shape, ValueRange dynamicSizes, - unsigned idx) const { - assert(idx < shape.size()); - if (!ShapedType::isDynamic(shape[idx])) - return createIndexConstant(rewriter, loc, shape[idx]); - // Count the number of dynamic dims in range [0, idx] - unsigned nDynamic = llvm::count_if(shape.take_front(idx), [](int64_t v) { - return ShapedType::isDynamic(v); - }); - return dynamicSizes[nDynamic]; - } - - // Build and return the idx^th stride, either by returning the constant stride - // or by computing the dynamic stride from the current `runningStride` and - // `nextSize`. The caller should keep a running stride and update it with the - // result returned by this function. - Value getStride(ConversionPatternRewriter &rewriter, Location loc, - ArrayRef strides, Value nextSize, - Value runningStride, unsigned idx) const { - assert(idx < strides.size()); - if (!MemRefType::isDynamicStrideOrOffset(strides[idx])) - return createIndexConstant(rewriter, loc, strides[idx]); - if (nextSize) - return runningStride - ? rewriter.create(loc, runningStride, nextSize) - : nextSize; - assert(!runningStride); - return createIndexConstant(rewriter, loc, 1); - } - - LogicalResult - matchAndRewrite(memref::ViewOp viewOp, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - auto loc = viewOp.getLoc(); - memref::ViewOpAdaptor adaptor(operands); - - auto viewMemRefType = viewOp.getType(); - auto targetElementTy = - typeConverter->convertType(viewMemRefType.getElementType()); - auto targetDescTy = typeConverter->convertType(viewMemRefType); - if (!targetDescTy || !targetElementTy || - !LLVM::isCompatibleType(targetElementTy) || - !LLVM::isCompatibleType(targetDescTy)) - return viewOp.emitWarning("Target descriptor type not converted to LLVM"), - failure(); - - int64_t offset; - SmallVector strides; - auto successStrides = getStridesAndOffset(viewMemRefType, strides, offset); - if (failed(successStrides)) - return viewOp.emitWarning("cannot cast to non-strided shape"), failure(); - assert(offset == 0 && "expected offset to be 0"); - - // Create the descriptor. - MemRefDescriptor sourceMemRef(adaptor.source()); - auto targetMemRef = MemRefDescriptor::undef(rewriter, loc, targetDescTy); - - // Field 1: Copy the allocated pointer, used for malloc/free. - Value allocatedPtr = sourceMemRef.allocatedPtr(rewriter, loc); - auto srcMemRefType = viewOp.source().getType().cast(); - Value bitcastPtr = rewriter.create( - loc, - LLVM::LLVMPointerType::get(targetElementTy, - srcMemRefType.getMemorySpaceAsInt()), - allocatedPtr); - targetMemRef.setAllocatedPtr(rewriter, loc, bitcastPtr); - - // Field 2: Copy the actual aligned pointer to payload. - Value alignedPtr = sourceMemRef.alignedPtr(rewriter, loc); - alignedPtr = rewriter.create(loc, alignedPtr.getType(), - alignedPtr, adaptor.byte_shift()); - bitcastPtr = rewriter.create( - loc, - LLVM::LLVMPointerType::get(targetElementTy, - srcMemRefType.getMemorySpaceAsInt()), - alignedPtr); - targetMemRef.setAlignedPtr(rewriter, loc, bitcastPtr); - - // Field 3: The offset in the resulting type must be 0. This is because of - // the type change: an offset on srcType* may not be expressible as an - // offset on dstType*. - targetMemRef.setOffset(rewriter, loc, - createIndexConstant(rewriter, loc, offset)); - - // Early exit for 0-D corner case. - if (viewMemRefType.getRank() == 0) - return rewriter.replaceOp(viewOp, {targetMemRef}), success(); - - // Fields 4 and 5: Update sizes and strides. - if (strides.back() != 1) - return viewOp.emitWarning("cannot cast to non-contiguous shape"), - failure(); - Value stride = nullptr, nextSize = nullptr; - for (int i = viewMemRefType.getRank() - 1; i >= 0; --i) { - // Update size. - Value size = - getSize(rewriter, loc, viewMemRefType.getShape(), adaptor.sizes(), i); - targetMemRef.setSize(rewriter, loc, i, size); - // Update stride. - stride = getStride(rewriter, loc, strides, nextSize, stride, i); - targetMemRef.setStride(rewriter, loc, i, stride); - nextSize = size; - } - - rewriter.replaceOp(viewOp, {targetMemRef}); - return success(); - } -}; - -struct AssumeAlignmentOpLowering - : public ConvertOpToLLVMPattern { - using ConvertOpToLLVMPattern< - memref::AssumeAlignmentOp>::ConvertOpToLLVMPattern; - - LogicalResult - matchAndRewrite(memref::AssumeAlignmentOp op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const override { - memref::AssumeAlignmentOp::Adaptor transformed(operands); - Value memref = transformed.memref(); - unsigned alignment = op.alignment(); - auto loc = op.getLoc(); - - MemRefDescriptor memRefDescriptor(memref); - Value ptr = memRefDescriptor.alignedPtr(rewriter, memref.getLoc()); - - // Emit llvm.assume(memref.alignedPtr & (alignment - 1) == 0). Notice that - // the asserted memref.alignedPtr isn't used anywhere else, as the real - // users like load/store/views always re-extract memref.alignedPtr as they - // get lowered. - // - // This relies on LLVM's CSE optimization (potentially after SROA), since - // after CSE all memref.alignedPtr instances get de-duplicated into the same - // pointer SSA value. - auto intPtrType = - getIntPtrType(memRefDescriptor.getElementPtrType().getAddressSpace()); - Value zero = createIndexAttrConstant(rewriter, loc, intPtrType, 0); - Value mask = - createIndexAttrConstant(rewriter, loc, intPtrType, alignment - 1); - Value ptrValue = rewriter.create(loc, intPtrType, ptr); - rewriter.create( - loc, rewriter.create( - loc, LLVM::ICmpPredicate::eq, - rewriter.create(loc, ptrValue, mask), zero)); - - rewriter.eraseOp(op); - return success(); - } -}; - } // namespace /// Try to match the kind of a std.atomic_rmw to determine whether to use a @@ -2627,17 +1348,22 @@ } // namespace -/// Collect a set of patterns to convert from the Standard dialect to LLVM. -void mlir::populateStdToLLVMNonMemoryConversionPatterns( +void mlir::populateStdToLLVMFuncOpConversionPattern( LLVMTypeConverter &converter, RewritePatternSet &patterns) { - // FIXME: this should be tablegen'ed + if (converter.getOptions().useBarePtrCallConv) + patterns.add(converter); + else + patterns.add(converter); +} + +void mlir::populateStdToLLVMConversionPatterns(LLVMTypeConverter &converter, + RewritePatternSet &patterns) { + populateStdToLLVMFuncOpConversionPattern(converter, patterns); // clang-format off patterns.add< AbsFOpLowering, AddFOpLowering, AddIOpLowering, - AllocaOpLowering, - AllocaScopeOpLowering, AndOpLowering, AssertOpLowering, AtomicRMWOpLowering, @@ -2673,8 +1399,8 @@ NegFOpLowering, OrOpLowering, PowFOpLowering, - PrefetchOpLowering, RemFOpLowering, + RankOpLowering, ReturnOpLowering, RsqrtOpLowering, SIToFPOpLowering, @@ -2701,47 +1427,6 @@ // clang-format on } -void mlir::populateStdToLLVMMemoryConversionPatterns( - LLVMTypeConverter &converter, RewritePatternSet &patterns) { - // clang-format off - patterns.add< - AssumeAlignmentOpLowering, - DimOpLowering, - GlobalMemrefOpLowering, - GetGlobalMemrefOpLowering, - LoadOpLowering, - MemRefCastOpLowering, - MemRefCopyOpLowering, - MemRefReinterpretCastOpLowering, - MemRefReshapeOpLowering, - RankOpLowering, - StoreOpLowering, - SubViewOpLowering, - TransposeOpLowering, - ViewOpLowering>(converter); - // clang-format on - auto allocLowering = converter.getOptions().allocLowering; - if (allocLowering == LowerToLLVMOptions::AllocLowering::AlignedAlloc) - patterns.add(converter); - else if (allocLowering == LowerToLLVMOptions::AllocLowering::Malloc) - patterns.add(converter); -} - -void mlir::populateStdToLLVMFuncOpConversionPattern( - LLVMTypeConverter &converter, RewritePatternSet &patterns) { - if (converter.getOptions().useBarePtrCallConv) - patterns.add(converter); - else - patterns.add(converter); -} - -void mlir::populateStdToLLVMConversionPatterns(LLVMTypeConverter &converter, - RewritePatternSet &patterns) { - populateStdToLLVMFuncOpConversionPattern(converter, patterns); - populateStdToLLVMNonMemoryConversionPatterns(converter, patterns); - populateStdToLLVMMemoryConversionPatterns(converter, patterns); -} - namespace { /// A pass converting MLIR operations into the LLVM IR dialect. struct LLVMLoweringPass : public ConvertStandardToLLVMBase { @@ -2752,7 +1437,6 @@ this->useBarePtrCallConv = useBarePtrCallConv; this->emitCWrappers = emitCWrappers; this->indexBitwidth = indexBitwidth; - this->useAlignedAlloc = useAlignedAlloc; this->dataLayout = dataLayout.getStringRepresentation(); } @@ -2782,9 +1466,6 @@ options.emitCWrappers = emitCWrappers; if (indexBitwidth != kDeriveIndexBitwidthFromDataLayout) options.overrideIndexBitwidth(indexBitwidth); - options.allocLowering = - (useAlignedAlloc ? LowerToLLVMOptions::AllocLowering::AlignedAlloc - : LowerToLLVMOptions::AllocLowering::Malloc); options.dataLayout = llvm::DataLayout(this->dataLayout); LLVMTypeConverter typeConverter(&getContext(), options, @@ -2802,54 +1483,6 @@ }; } // end namespace -Value AllocLikeOpLLVMLowering::createAligned( - ConversionPatternRewriter &rewriter, Location loc, Value input, - Value alignment) { - Value one = createIndexAttrConstant(rewriter, loc, alignment.getType(), 1); - Value bump = rewriter.create(loc, alignment, one); - Value bumped = rewriter.create(loc, input, bump); - Value mod = rewriter.create(loc, bumped, alignment); - return rewriter.create(loc, bumped, mod); -} - -LogicalResult AllocLikeOpLLVMLowering::matchAndRewrite( - Operation *op, ArrayRef operands, - ConversionPatternRewriter &rewriter) const { - MemRefType memRefType = getMemRefResultType(op); - if (!isConvertibleAndHasIdentityMaps(memRefType)) - return rewriter.notifyMatchFailure(op, "incompatible memref type"); - auto loc = op->getLoc(); - - // Get actual sizes of the memref as values: static sizes are constant - // values and dynamic sizes are passed to 'alloc' as operands. In case of - // zero-dimensional memref, assume a scalar (size 1). - SmallVector sizes; - SmallVector strides; - Value sizeBytes; - this->getMemRefDescriptorSizes(loc, memRefType, operands, rewriter, sizes, - strides, sizeBytes); - - // Allocate the underlying buffer. - Value allocatedPtr; - Value alignedPtr; - std::tie(allocatedPtr, alignedPtr) = - this->allocateBuffer(rewriter, loc, sizeBytes, op); - - // Create the MemRef descriptor. - auto memRefDescriptor = this->createMemRefDescriptor( - loc, memRefType, allocatedPtr, alignedPtr, sizes, strides, rewriter); - - // Return the final value of the descriptor. - rewriter.replaceOp(op, {memRefDescriptor}); - return success(); -} - -mlir::LLVMConversionTarget::LLVMConversionTarget(MLIRContext &ctx) - : ConversionTarget(ctx) { - this->addLegalDialect(); - this->addIllegalOp(); -} - std::unique_ptr> mlir::createLowerToLLVMPass() { return std::make_unique(); } diff --git a/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt b/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt --- a/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt +++ b/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt @@ -18,9 +18,9 @@ MLIRArmSVETransforms MLIRAMX MLIRAMXTransforms + MLIRLLVMCommonConversion MLIRLLVMIR MLIRMemRef - MLIRStandardToLLVM MLIRTargetLLVMIRExport MLIRTransforms MLIRVector diff --git a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp --- a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp +++ b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp @@ -8,8 +8,7 @@ #include "mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" +#include "mlir/Conversion/LLVMCommon/VectorPattern.h" #include "mlir/Dialect/LLVMIR/FunctionCallUtils.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/MemRef/IR/MemRef.h" diff --git a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp --- a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp +++ b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp @@ -10,8 +10,8 @@ #include "../PassDetail.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/TypeConverter.h" #include "mlir/Dialect/AMX/AMXDialect.h" #include "mlir/Dialect/AMX/Transforms.h" #include "mlir/Dialect/ArmNeon/ArmNeonDialect.h" diff --git a/mlir/lib/Conversion/VectorToROCDL/CMakeLists.txt b/mlir/lib/Conversion/VectorToROCDL/CMakeLists.txt --- a/mlir/lib/Conversion/VectorToROCDL/CMakeLists.txt +++ b/mlir/lib/Conversion/VectorToROCDL/CMakeLists.txt @@ -13,6 +13,8 @@ LINK_LIBS PUBLIC MLIRROCDLIR + MLIRLLVMCommonConversion + MLIRMemRefToLLVM MLIRStandardToLLVM MLIRVector MLIRTransforms diff --git a/mlir/lib/Conversion/VectorToROCDL/VectorToROCDL.cpp b/mlir/lib/Conversion/VectorToROCDL/VectorToROCDL.cpp --- a/mlir/lib/Conversion/VectorToROCDL/VectorToROCDL.cpp +++ b/mlir/lib/Conversion/VectorToROCDL/VectorToROCDL.cpp @@ -14,8 +14,10 @@ #include "mlir/Conversion/VectorToROCDL/VectorToROCDL.h" #include "../PassDetail.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Dialect/GPU/GPUDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/LLVMIR/ROCDLDialect.h" @@ -161,6 +163,7 @@ RewritePatternSet patterns(&getContext()); populateVectorToROCDLConversionPatterns(converter, patterns); + populateMemRefToLLVMConversionPatterns(converter, patterns); populateStdToLLVMConversionPatterns(converter, patterns); LLVMConversionTarget target(getContext()); diff --git a/mlir/lib/Dialect/AMX/Transforms/CMakeLists.txt b/mlir/lib/Dialect/AMX/Transforms/CMakeLists.txt --- a/mlir/lib/Dialect/AMX/Transforms/CMakeLists.txt +++ b/mlir/lib/Dialect/AMX/Transforms/CMakeLists.txt @@ -7,6 +7,6 @@ LINK_LIBS PUBLIC MLIRAMX MLIRIR + MLIRLLVMCommonConversion MLIRLLVMIR - MLIRStandardToLLVM ) diff --git a/mlir/lib/Dialect/AMX/Transforms/LegalizeForLLVMExport.cpp b/mlir/lib/Dialect/AMX/Transforms/LegalizeForLLVMExport.cpp --- a/mlir/lib/Dialect/AMX/Transforms/LegalizeForLLVMExport.cpp +++ b/mlir/lib/Dialect/AMX/Transforms/LegalizeForLLVMExport.cpp @@ -8,7 +8,8 @@ #include "mlir/Dialect/AMX/Transforms.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/AMX/AMXDialect.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/StandardOps/IR/Ops.h" diff --git a/mlir/lib/Dialect/ArmSVE/Transforms/CMakeLists.txt b/mlir/lib/Dialect/ArmSVE/Transforms/CMakeLists.txt --- a/mlir/lib/Dialect/ArmSVE/Transforms/CMakeLists.txt +++ b/mlir/lib/Dialect/ArmSVE/Transforms/CMakeLists.txt @@ -7,6 +7,6 @@ LINK_LIBS PUBLIC MLIRArmSVE MLIRIR + MLIRLLVMCommonConversion MLIRLLVMIR - MLIRStandardToLLVM ) diff --git a/mlir/lib/Dialect/ArmSVE/Transforms/LegalizeForLLVMExport.cpp b/mlir/lib/Dialect/ArmSVE/Transforms/LegalizeForLLVMExport.cpp --- a/mlir/lib/Dialect/ArmSVE/Transforms/LegalizeForLLVMExport.cpp +++ b/mlir/lib/Dialect/ArmSVE/Transforms/LegalizeForLLVMExport.cpp @@ -6,7 +6,8 @@ // //===----------------------------------------------------------------------===// -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/ArmSVE/ArmSVEDialect.h" #include "mlir/Dialect/ArmSVE/Transforms.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" diff --git a/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp b/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp --- a/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp +++ b/mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp @@ -1236,9 +1236,11 @@ static LogicalResult verify(AddressOfOp op) { auto global = op.getGlobal(); auto function = op.getFunction(); - if (!global && !function) + if (!global && !function) { + op->getParentOfType().dump(); return op.emitOpError( "must reference a global defined by 'llvm.mlir.global' or 'llvm.func'"); + } if (global && LLVM::LLVMPointerType::get(global.getType(), global.addr_space()) != diff --git a/mlir/lib/Dialect/X86Vector/Transforms/CMakeLists.txt b/mlir/lib/Dialect/X86Vector/Transforms/CMakeLists.txt --- a/mlir/lib/Dialect/X86Vector/Transforms/CMakeLists.txt +++ b/mlir/lib/Dialect/X86Vector/Transforms/CMakeLists.txt @@ -7,6 +7,6 @@ LINK_LIBS PUBLIC MLIRX86Vector MLIRIR + MLIRLLVMCommonConversion MLIRLLVMIR - MLIRStandardToLLVM ) diff --git a/mlir/lib/Dialect/X86Vector/Transforms/LegalizeForLLVMExport.cpp b/mlir/lib/Dialect/X86Vector/Transforms/LegalizeForLLVMExport.cpp --- a/mlir/lib/Dialect/X86Vector/Transforms/LegalizeForLLVMExport.cpp +++ b/mlir/lib/Dialect/X86Vector/Transforms/LegalizeForLLVMExport.cpp @@ -8,7 +8,8 @@ #include "mlir/Dialect/X86Vector/Transforms.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" +#include "mlir/Conversion/LLVMCommon/ConversionTarget.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/StandardOps/IR/Ops.h" #include "mlir/Dialect/X86Vector/X86VectorDialect.h" diff --git a/mlir/test/Conversion/StandardToLLVM/convert-alloca-scope.mlir b/mlir/test/Conversion/MemRefToLLVM/convert-alloca-scope.mlir rename from mlir/test/Conversion/StandardToLLVM/convert-alloca-scope.mlir rename to mlir/test/Conversion/MemRefToLLVM/convert-alloca-scope.mlir --- a/mlir/test/Conversion/StandardToLLVM/convert-alloca-scope.mlir +++ b/mlir/test/Conversion/MemRefToLLVM/convert-alloca-scope.mlir @@ -1,6 +1,6 @@ -// RUN: mlir-opt -convert-std-to-llvm %s | FileCheck %s +// RUN: mlir-opt -convert-memref-to-llvm %s | FileCheck %s -// CHECK-LABEL: llvm.func @empty +// CHECK-LABEL: @empty func @empty() { // CHECK: llvm.intr.stacksave // CHECK: llvm.br @@ -9,11 +9,10 @@ } // CHECK: llvm.intr.stackrestore // CHECK: llvm.br - // CHECK: llvm.return return } -// CHECK-LABEL: llvm.func @returns_nothing +// CHECK-LABEL: @returns_nothing func @returns_nothing(%b: f32) { %a = constant 10.0 : f32 // CHECK: llvm.intr.stacksave @@ -22,11 +21,10 @@ memref.alloca_scope.return } // CHECK: llvm.intr.stackrestore - // CHECK: llvm.return return } -// CHECK-LABEL: llvm.func @returns_one_value +// CHECK-LABEL: @returns_one_value func @returns_one_value(%b: f32) -> f32 { %a = constant 10.0 : f32 // CHECK: llvm.intr.stacksave @@ -35,11 +33,10 @@ memref.alloca_scope.return %c: f32 } // CHECK: llvm.intr.stackrestore - // CHECK: llvm.return return %result : f32 } -// CHECK-LABEL: llvm.func @returns_multiple_values +// CHECK-LABEL: @returns_multiple_values func @returns_multiple_values(%b: f32) -> f32 { %a = constant 10.0 : f32 // CHECK: llvm.intr.stacksave @@ -49,7 +46,6 @@ memref.alloca_scope.return %c, %d: f32, f32 } // CHECK: llvm.intr.stackrestore - // CHECK: llvm.return %result = std.addf %result1, %result2 : f32 return %result : f32 } diff --git a/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir b/mlir/test/Conversion/MemRefToLLVM/convert-dynamic-memref-ops.mlir rename from mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir rename to mlir/test/Conversion/MemRefToLLVM/convert-dynamic-memref-ops.mlir --- a/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir +++ b/mlir/test/Conversion/MemRefToLLVM/convert-dynamic-memref-ops.mlir @@ -1,33 +1,11 @@ -// RUN: mlir-opt -split-input-file -convert-std-to-llvm %s | FileCheck %s -// RUN: mlir-opt -split-input-file -convert-std-to-llvm='use-aligned-alloc=1' %s | FileCheck %s --check-prefix=ALIGNED-ALLOC - -// CHECK-LABEL: func @check_strided_memref_arguments( -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -func @check_strided_memref_arguments(%static: memref<10x20xf32, affine_map<(i,j)->(20 * i + j + 1)>>, - %dynamic : memref(M * i + j + 1)>>, - %mixed : memref<10x?xf32, affine_map<(i,j)[M]->(M * i + j + 1)>>) { - return -} - -// CHECK-LABEL: func @check_arguments -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -func @check_arguments(%static: memref<10x20xf32>, %dynamic : memref, %mixed : memref<10x?xf32>) { - return -} +// RUN: mlir-opt -split-input-file -convert-memref-to-llvm %s | FileCheck %s +// RUN: mlir-opt -split-input-file -convert-memref-to-llvm='use-aligned-alloc=1' %s | FileCheck %s --check-prefix=ALIGNED-ALLOC // CHECK-LABEL: func @mixed_alloc( -// CHECK: %[[M:.*]]: i64, %[[N:.*]]: i64) -> !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> { +// CHECK: %[[Marg:.*]]: index, %[[Narg:.*]]: index) func @mixed_alloc(%arg0: index, %arg1: index) -> memref { +// CHECK: %[[M:.*]] = llvm.mlir.cast %[[Marg]] +// CHECK: %[[N:.*]] = llvm.mlir.cast %[[Narg]] // CHECK: %[[c42:.*]] = llvm.mlir.constant(42 : index) : i64 // CHECK-NEXT: %[[one:.*]] = llvm.mlir.constant(1 : index) : i64 // CHECK-NEXT: %[[st0:.*]] = llvm.mul %[[N]], %[[c42]] : i64 @@ -49,23 +27,27 @@ // CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> // CHECK-NEXT: llvm.insertvalue %[[one]], %{{.*}}[4, 2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> %0 = memref.alloc(%arg0, %arg1) : memref -// CHECK-NEXT: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> return %0 : memref } +// ----- + // CHECK-LABEL: func @mixed_dealloc func @mixed_dealloc(%arg0: memref) { // CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> // CHECK-NEXT: %[[ptri8:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr // CHECK-NEXT: llvm.call @free(%[[ptri8]]) : (!llvm.ptr) -> () memref.dealloc %arg0 : memref -// CHECK-NEXT: llvm.return return } +// ----- + // CHECK-LABEL: func @dynamic_alloc( -// CHECK: %[[M:.*]]: i64, %[[N:.*]]: i64) -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { +// CHECK: %[[Marg:.*]]: index, %[[Narg:.*]]: index) func @dynamic_alloc(%arg0: index, %arg1: index) -> memref { +// CHECK: %[[M:.*]] = llvm.mlir.cast %[[Marg]] +// CHECK: %[[N:.*]] = llvm.mlir.cast %[[Narg]] // CHECK-NEXT: %[[one:.*]] = llvm.mlir.constant(1 : index) : i64 // CHECK-NEXT: %[[sz:.*]] = llvm.mul %[[N]], %[[M]] : i64 // CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr @@ -83,15 +65,16 @@ // CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: llvm.insertvalue %[[one]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> %0 = memref.alloc(%arg0, %arg1) : memref -// CHECK-NEXT: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> return %0 : memref } // ----- // CHECK-LABEL: func @dynamic_alloca -// CHECK: %[[M:.*]]: i64, %[[N:.*]]: i64) -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { +// CHECK: %[[Marg:.*]]: index, %[[Narg:.*]]: index) func @dynamic_alloca(%arg0: index, %arg1: index) -> memref { +// CHECK: %[[M:.*]] = llvm.mlir.cast %[[Marg]] +// CHECK: %[[N:.*]] = llvm.mlir.cast %[[Narg]] // CHECK-NEXT: %[[st1:.*]] = llvm.mlir.constant(1 : index) : i64 // CHECK-NEXT: %[[num_elems:.*]] = llvm.mul %[[N]], %[[M]] : i64 // CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr @@ -120,6 +103,8 @@ return %0 : memref } +// ----- + // CHECK-LABEL: func @dynamic_dealloc func @dynamic_dealloc(%arg0: memref) { // CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> @@ -129,8 +114,10 @@ return } -// CHECK-LABEL: func @stdlib_aligned_alloc({{.*}}) -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { -// ALIGNED-ALLOC-LABEL: func @stdlib_aligned_alloc({{.*}}) -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { +// ----- + +// CHECK-LABEL: func @stdlib_aligned_alloc({{.*}}) +// ALIGNED-ALLOC-LABEL: func @stdlib_aligned_alloc({{.*}}) func @stdlib_aligned_alloc(%N : index) -> memref<32x18xf32> { // ALIGNED-ALLOC-NEXT: %[[sz1:.*]] = llvm.mlir.constant(32 : index) : i64 // ALIGNED-ALLOC-NEXT: %[[sz2:.*]] = llvm.mlir.constant(18 : index) : i64 @@ -175,12 +162,13 @@ return %0 : memref<32x18xf32> } +// ----- + // CHECK-LABEL: func @mixed_load( -// CHECK-COUNT-2: !llvm.ptr, -// CHECK-COUNT-5: {{%[a-zA-Z0-9]*}}: i64 -// CHECK: %[[I:.*]]: i64, -// CHECK: %[[J:.*]]: i64) +// CHECK: %{{.*}}, %[[Iarg:.*]]: index, %[[Jarg:.*]]: index) func @mixed_load(%mixed : memref<42x?xf32>, %i : index, %j : index) { +// CHECK: %[[I:.*]] = llvm.mlir.cast %[[Iarg]] +// CHECK: %[[J:.*]] = llvm.mlir.cast %[[Jarg]] // CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[st0:.*]] = llvm.extractvalue %[[ld]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 @@ -191,17 +179,13 @@ return } +// ----- + // CHECK-LABEL: func @dynamic_load( -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[J:[a-zA-Z0-9]*]]: i64 +// CHECK: %{{.*}}, %[[Iarg:.*]]: index, %[[Jarg:.*]]: index) func @dynamic_load(%dynamic : memref, %i : index, %j : index) { +// CHECK: %[[I:.*]] = llvm.mlir.cast %[[Iarg]] +// CHECK: %[[J:.*]] = llvm.mlir.cast %[[Jarg]] // CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[st0:.*]] = llvm.extractvalue %[[ld]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 @@ -212,17 +196,13 @@ return } +// ----- + // CHECK-LABEL: func @prefetch -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[J:[a-zA-Z0-9]*]]: i64 +// CHECK: %{{.*}}, %[[Iarg:.*]]: index, %[[Jarg:.*]]: index) func @prefetch(%A : memref, %i : index, %j : index) { +// CHECK: %[[I:.*]] = llvm.mlir.cast %[[Iarg]] +// CHECK: %[[J:.*]] = llvm.mlir.cast %[[Jarg]] // CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[st0:.*]] = llvm.extractvalue %[[ld]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 @@ -246,17 +226,13 @@ return } +// ----- + // CHECK-LABEL: func @dynamic_store -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[J:[a-zA-Z0-9]*]]: i64 +// CHECK: %{{.*}}, %[[Iarg:.*]]: index, %[[Jarg:.*]]: index func @dynamic_store(%dynamic : memref, %i : index, %j : index, %val : f32) { +// CHECK: %[[I:.*]] = llvm.mlir.cast %[[Iarg]] +// CHECK: %[[J:.*]] = llvm.mlir.cast %[[Jarg]] // CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[st0:.*]] = llvm.extractvalue %[[ld]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 @@ -267,17 +243,13 @@ return } +// ----- + // CHECK-LABEL: func @mixed_store -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[J:[a-zA-Z0-9]*]]: i64 +// CHECK: %{{.*}}, %[[Iarg:.*]]: index, %[[Jarg:.*]]: index func @mixed_store(%mixed : memref<42x?xf32>, %i : index, %j : index, %val : f32) { +// CHECK: %[[I:.*]] = llvm.mlir.cast %[[Iarg]] +// CHECK: %[[J:.*]] = llvm.mlir.cast %[[Jarg]] // CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[st0:.*]] = llvm.extractvalue %[[ld]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> // CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 @@ -288,6 +260,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_static_to_dynamic func @memref_cast_static_to_dynamic(%static : memref<10x42xf32>) { // CHECK-NOT: llvm.bitcast @@ -295,6 +269,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_static_to_mixed func @memref_cast_static_to_mixed(%static : memref<10x42xf32>) { // CHECK-NOT: llvm.bitcast @@ -302,6 +278,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_dynamic_to_static func @memref_cast_dynamic_to_static(%dynamic : memref) { // CHECK-NOT: llvm.bitcast @@ -309,6 +287,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_dynamic_to_mixed func @memref_cast_dynamic_to_mixed(%dynamic : memref) { // CHECK-NOT: llvm.bitcast @@ -316,6 +296,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_mixed_to_dynamic func @memref_cast_mixed_to_dynamic(%mixed : memref<42x?xf32>) { // CHECK-NOT: llvm.bitcast @@ -323,6 +305,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_mixed_to_static func @memref_cast_mixed_to_static(%mixed : memref<42x?xf32>) { // CHECK-NOT: llvm.bitcast @@ -330,6 +314,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_mixed_to_mixed func @memref_cast_mixed_to_mixed(%mixed : memref<42x?xf32>) { // CHECK-NOT: llvm.bitcast @@ -337,6 +323,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_ranked_to_unranked func @memref_cast_ranked_to_unranked(%arg : memref<42x2x?xf32>) { // CHECK-DAG: %[[c:.*]] = llvm.mlir.constant(1 : index) : i64 @@ -351,6 +339,8 @@ return } +// ----- + // CHECK-LABEL: func @memref_cast_unranked_to_ranked func @memref_cast_unranked_to_ranked(%arg : memref<*xf32>) { // CHECK: %[[p:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(i64, ptr)> @@ -359,49 +349,47 @@ return } +// ----- + // CHECK-LABEL: func @mixed_memref_dim func @mixed_memref_dim(%mixed : memref<42x?x?x13x?xf32>) { // CHECK: llvm.mlir.constant(42 : index) : i64 %c0 = constant 0 : index %0 = memref.dim %mixed, %c0 : memref<42x?x?x13x?xf32> -// CHECK: llvm.extractvalue %[[ld:.*]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> +// CHECK: llvm.extractvalue %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> %c1 = constant 1 : index %1 = memref.dim %mixed, %c1 : memref<42x?x?x13x?xf32> -// CHECK: llvm.extractvalue %[[ld]][3, 2] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> +// CHECK: llvm.extractvalue %{{.*}}[3, 2] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> %c2 = constant 2 : index %2 = memref.dim %mixed, %c2 : memref<42x?x?x13x?xf32> // CHECK: llvm.mlir.constant(13 : index) : i64 %c3 = constant 3 : index %3 = memref.dim %mixed, %c3 : memref<42x?x?x13x?xf32> -// CHECK: llvm.extractvalue %[[ld]][3, 4] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> +// CHECK: llvm.extractvalue %{{.*}}[3, 4] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> %c4 = constant 4 : index %4 = memref.dim %mixed, %c4 : memref<42x?x?x13x?xf32> return } +// ----- + // CHECK-LABEL: @memref_dim_with_dyn_index -// CHECK-SAME: %[[ALLOC_PTR:.*]]: !llvm.ptr, %[[ALIGN_PTR:.*]]: !llvm.ptr, %[[OFFSET:.*]]: i64, %[[SIZE0:.*]]: i64, %[[SIZE1:.*]]: i64, %[[STRIDE0:.*]]: i64, %[[STRIDE1:.*]]: i64, %[[IDX:.*]]: i64) -> i64 +// CHECK: %{{.*}}, %[[IDXarg:.*]]: index func @memref_dim_with_dyn_index(%arg : memref<3x?xf32>, %idx : index) -> index { - // CHECK-NEXT: %[[DESCR0:.*]] = llvm.mlir.undef : [[DESCR_TY:!llvm.struct<\(ptr, ptr, i64, array<2 x i64>, array<2 x i64>\)>]] - // CHECK-NEXT: %[[DESCR1:.*]] = llvm.insertvalue %[[ALLOC_PTR]], %[[DESCR0]][0] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR2:.*]] = llvm.insertvalue %[[ALIGN_PTR]], %[[DESCR1]][1] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR3:.*]] = llvm.insertvalue %[[OFFSET]], %[[DESCR2]][2] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR4:.*]] = llvm.insertvalue %[[SIZE0]], %[[DESCR3]][3, 0] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR5:.*]] = llvm.insertvalue %[[STRIDE0]], %[[DESCR4]][4, 0] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR6:.*]] = llvm.insertvalue %[[SIZE1]], %[[DESCR5]][3, 1] : [[DESCR_TY]] - // CHECK-NEXT: %[[DESCR7:.*]] = llvm.insertvalue %[[STRIDE1]], %[[DESCR6]][4, 1] : [[DESCR_TY]] // CHECK-DAG: %[[C0:.*]] = llvm.mlir.constant(0 : index) : i64 // CHECK-DAG: %[[C1:.*]] = llvm.mlir.constant(1 : index) : i64 - // CHECK-DAG: %[[SIZES:.*]] = llvm.extractvalue %[[DESCR7]][3] : [[DESCR_TY]] + // CHECK-DAG: %[[SIZES:.*]] = llvm.extractvalue %{{.*}}[3] : ![[DESCR_TY:.*]] // CHECK-DAG: %[[SIZES_PTR:.*]] = llvm.alloca %[[C1]] x !llvm.array<2 x i64> : (i64) -> !llvm.ptr> // CHECK-DAG: llvm.store %[[SIZES]], %[[SIZES_PTR]] : !llvm.ptr> + // CHECK-DAG: %[[IDX:.*]] = llvm.mlir.cast %[[IDXarg]] // CHECK-DAG: %[[RESULT_PTR:.*]] = llvm.getelementptr %[[SIZES_PTR]][%[[C0]], %[[IDX]]] : (!llvm.ptr>, i64, i64) -> !llvm.ptr // CHECK-DAG: %[[RESULT:.*]] = llvm.load %[[RESULT_PTR]] : !llvm.ptr - // CHECK-DAG: llvm.return %[[RESULT]] : i64 %result = memref.dim %arg, %idx : memref<3x?xf32> return %result : index } +// ----- + // CHECK-LABEL: @memref_reinterpret_cast_ranked_to_static_shape func @memref_reinterpret_cast_ranked_to_static_shape(%input : memref<2x3xf32>) { %output = memref.reinterpret_cast %input to @@ -409,7 +397,8 @@ : memref<2x3xf32> to memref<6x1xf32> return } -// CHECK: [[INPUT:%.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : [[TY:!.*]] +// CHECK: [[INPUT:%.*]] = llvm.mlir.cast %{{.*}} : +// CHECK: to [[TY:!.*]] // CHECK: [[OUT_0:%.*]] = llvm.mlir.undef : [[TY]] // CHECK: [[BASE_PTR:%.*]] = llvm.extractvalue [[INPUT]][0] : [[TY]] // CHECK: [[ALIGNED_PTR:%.*]] = llvm.extractvalue [[INPUT]][1] : [[TY]] @@ -426,6 +415,8 @@ // CHECK: [[STRIDE_1:%.*]] = llvm.mlir.constant(1 : index) : i64 // CHECK: [[OUT_7:%.*]] = llvm.insertvalue [[STRIDE_1]], [[OUT_6]][4, 1] : [[TY]] +// ----- + // CHECK-LABEL: @memref_reinterpret_cast_unranked_to_dynamic_shape func @memref_reinterpret_cast_unranked_to_dynamic_shape(%offset: index, %size_0 : index, @@ -439,10 +430,15 @@ : memref<*xf32> to memref return } -// CHECK-SAME: ([[OFFSET:%[a-z,0-9]+]]: i64, -// CHECK-SAME: [[SIZE_0:%[a-z,0-9]+]]: i64, [[SIZE_1:%[a-z,0-9]+]]: i64, -// CHECK-SAME: [[STRIDE_0:%[a-z,0-9]+]]: i64, [[STRIDE_1:%[a-z,0-9]+]]: i64, -// CHECK: [[INPUT:%.*]] = llvm.insertvalue {{.*}}[1] : !llvm.struct<(i64, ptr)> +// CHECK-SAME: ([[OFFSETarg:%[a-z,0-9]+]]: index, +// CHECK-SAME: [[SIZE_0arg:%[a-z,0-9]+]]: index, [[SIZE_1arg:%[a-z,0-9]+]]: index, +// CHECK-SAME: [[STRIDE_0arg:%[a-z,0-9]+]]: index, [[STRIDE_1arg:%[a-z,0-9]+]]: index, +// CHECK: [[INPUT:%.*]] = llvm.mlir.cast +// CHECK: [[OFFSET:%.*]] = llvm.mlir.cast [[OFFSETarg]] +// CHECK: [[SIZE_0:%.*]] = llvm.mlir.cast [[SIZE_0arg]] +// CHECK: [[SIZE_1:%.*]] = llvm.mlir.cast [[SIZE_1arg]] +// CHECK: [[STRIDE_0:%.*]] = llvm.mlir.cast [[STRIDE_0arg]] +// CHECK: [[STRIDE_1:%.*]] = llvm.mlir.cast [[STRIDE_1arg]] // CHECK: [[OUT_0:%.*]] = llvm.mlir.undef : [[TY:!.*]] // CHECK: [[DESCRIPTOR:%.*]] = llvm.extractvalue [[INPUT]][1] : !llvm.struct<(i64, ptr)> // CHECK: [[BASE_PTR_PTR:%.*]] = llvm.bitcast [[DESCRIPTOR]] : !llvm.ptr to !llvm.ptr> @@ -460,14 +456,16 @@ // CHECK: [[OUT_6:%.*]] = llvm.insertvalue [[SIZE_1]], [[OUT_5]][3, 1] : [[TY]] // CHECK: [[OUT_7:%.*]] = llvm.insertvalue [[STRIDE_1]], [[OUT_6]][4, 1] : [[TY]] +// ----- + // CHECK-LABEL: @memref_reshape func @memref_reshape(%input : memref<2x3xf32>, %shape : memref) { %output = memref.reshape %input(%shape) : (memref<2x3xf32>, memref) -> memref<*xf32> return } -// CHECK: [[INPUT:%.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : [[INPUT_TY:!.*]] -// CHECK: [[SHAPE:%.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : [[SHAPE_TY:!.*]] +// CHECK: [[INPUT:%.*]] = llvm.mlir.cast %{{.*}} to [[INPUT_TY:!.*]] +// CHECK: [[SHAPE:%.*]] = llvm.mlir.cast %{{.*}} to [[SHAPE_TY:!.*]] // CHECK: [[RANK:%.*]] = llvm.extractvalue [[SHAPE]][3, 0] : [[SHAPE_TY]] // CHECK: [[UNRANKED_OUT_O:%.*]] = llvm.mlir.undef : !llvm.struct<(i64, ptr)> // CHECK: [[UNRANKED_OUT_1:%.*]] = llvm.insertvalue [[RANK]], [[UNRANKED_OUT_O]][0] : !llvm.struct<(i64, ptr)> @@ -528,7 +526,7 @@ // CHECK: llvm.br ^bb1([[STRIDE_COND]], [[UPDATE_STRIDE]] : i64, i64) // CHECK: ^bb3: -// CHECK: llvm.return +// CHECK: return // ----- diff --git a/mlir/test/Conversion/MemRefToLLVM/convert-static-memref-ops.mlir b/mlir/test/Conversion/MemRefToLLVM/convert-static-memref-ops.mlir new file mode 100644 --- /dev/null +++ b/mlir/test/Conversion/MemRefToLLVM/convert-static-memref-ops.mlir @@ -0,0 +1,221 @@ +// RUN: mlir-opt -convert-memref-to-llvm -split-input-file %s | FileCheck %s + +// ----- + +// CHECK-LABEL: func @zero_d_alloc() +func @zero_d_alloc() -> memref { +// CHECK: %[[one:.*]] = llvm.mlir.constant(1 : index) : i64 +// CHECK: %[[null:.*]] = llvm.mlir.null : !llvm.ptr +// CHECK: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[one]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 +// CHECK: llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr +// CHECK: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr +// CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64)> +// CHECK: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: llvm.insertvalue %[[ptr]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 +// CHECK: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: llvm.mlir.cast %{{.*}} + + %0 = memref.alloc() : memref + return %0 : memref +} + +// ----- + +// CHECK-LABEL: func @zero_d_dealloc +func @zero_d_dealloc(%arg0: memref) { +// CHECK: llvm.mlir.cast +// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr +// CHECK: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () + + memref.dealloc %arg0 : memref + return +} + +// ----- + +// CHECK-LABEL: func @aligned_1d_alloc( +func @aligned_1d_alloc() -> memref<42xf32> { +// CHECK: %[[sz1:.*]] = llvm.mlir.constant(42 : index) : i64 +// CHECK: %[[st1:.*]] = llvm.mlir.constant(1 : index) : i64 +// CHECK: %[[null:.*]] = llvm.mlir.null : !llvm.ptr +// CHECK: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[sz1]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 +// CHECK: %[[alignment:.*]] = llvm.mlir.constant(8 : index) : i64 +// CHECK: %[[allocsize:.*]] = llvm.add %[[size_bytes]], %[[alignment]] : i64 +// CHECK: %[[allocated:.*]] = llvm.call @malloc(%[[allocsize]]) : (i64) -> !llvm.ptr +// CHECK: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr +// CHECK: %[[allocatedAsInt:.*]] = llvm.ptrtoint %[[ptr]] : !llvm.ptr to i64 +// CHECK: %[[one_1:.*]] = llvm.mlir.constant(1 : index) : i64 +// CHECK: %[[bump:.*]] = llvm.sub %[[alignment]], %[[one_1]] : i64 +// CHECK: %[[bumped:.*]] = llvm.add %[[allocatedAsInt]], %[[bump]] : i64 +// CHECK: %[[mod:.*]] = llvm.urem %[[bumped]], %[[alignment]] : i64 +// CHECK: %[[aligned:.*]] = llvm.sub %[[bumped]], %[[mod]] : i64 +// CHECK: %[[alignedBitCast:.*]] = llvm.inttoptr %[[aligned]] : i64 to !llvm.ptr +// CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> +// CHECK: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> +// CHECK: llvm.insertvalue %[[alignedBitCast]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> +// CHECK: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 +// CHECK: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + %0 = memref.alloc() {alignment = 8} : memref<42xf32> + return %0 : memref<42xf32> +} + +// ----- + +// CHECK-LABEL: func @static_alloc() +func @static_alloc() -> memref<32x18xf32> { +// CHECK: %[[num_elems:.*]] = llvm.mlir.constant(576 : index) : i64 +// CHECK: %[[null:.*]] = llvm.mlir.null : !llvm.ptr +// CHECK: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 +// CHECK: %[[allocated:.*]] = llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr +// CHECK: llvm.bitcast %[[allocated]] : !llvm.ptr to !llvm.ptr + %0 = memref.alloc() : memref<32x18xf32> + return %0 : memref<32x18xf32> +} + +// ----- + +// CHECK-LABEL: func @static_alloca() +func @static_alloca() -> memref<32x18xf32> { +// CHECK: %[[sz1:.*]] = llvm.mlir.constant(32 : index) : i64 +// CHECK: %[[sz2:.*]] = llvm.mlir.constant(18 : index) : i64 +// CHECK: %[[st2:.*]] = llvm.mlir.constant(1 : index) : i64 +// CHECK: %[[num_elems:.*]] = llvm.mlir.constant(576 : index) : i64 +// CHECK: %[[null:.*]] = llvm.mlir.null : !llvm.ptr +// CHECK: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 +// CHECK: %[[allocated:.*]] = llvm.alloca %[[size_bytes]] x f32 : (i64) -> !llvm.ptr + %0 = memref.alloca() : memref<32x18xf32> + + // Test with explicitly specified alignment. llvm.alloca takes care of the + // alignment. The same pointer is thus used for allocation and aligned + // accesses. + // CHECK: %[[alloca_aligned:.*]] = llvm.alloca %{{.*}} x f32 {alignment = 32 : i64} : (i64) -> !llvm.ptr + // CHECK: %[[desc:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[desc1:.*]] = llvm.insertvalue %[[alloca_aligned]], %[[desc]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[alloca_aligned]], %[[desc1]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + memref.alloca() {alignment = 32} : memref<32x18xf32> + return %0 : memref<32x18xf32> +} + +// ----- + +// CHECK-LABEL: func @static_dealloc +func @static_dealloc(%static: memref<10x8xf32>) { +// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// CHECK: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr +// CHECK: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () + memref.dealloc %static : memref<10x8xf32> + return +} + +// ----- + +// CHECK-LABEL: func @zero_d_load +func @zero_d_load(%arg0: memref) -> f32 { +// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: %{{.*}} = llvm.load %[[ptr]] : !llvm.ptr + %0 = memref.load %arg0[] : memref + return %0 : f32 +} + +// ----- + +// CHECK-LABEL: func @static_load +// CHECK: %[[MEMREF:.*]]: memref<10x42xf32>, +// CHECK: %[[I:.*]]: index, +// CHECK: %[[J:.*]]: index) +func @static_load(%static : memref<10x42xf32>, %i : index, %j : index) { +// CHECK: %[[II:.*]] = llvm.mlir.cast %[[I]] +// CHECK: %[[JJ:.*]] = llvm.mlir.cast %[[J]] +// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// CHECK: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 +// CHECK: %[[offI:.*]] = llvm.mul %[[II]], %[[st0]] : i64 +// CHECK: %[[off1:.*]] = llvm.add %[[offI]], %[[JJ]] : i64 +// CHECK: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: llvm.load %[[addr]] : !llvm.ptr + %0 = memref.load %static[%i, %j] : memref<10x42xf32> + return +} + +// ----- + +// CHECK-LABEL: func @zero_d_store +func @zero_d_store(%arg0: memref, %arg1: f32) { +// CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64)> +// CHECK: llvm.store %{{.*}}, %[[ptr]] : !llvm.ptr + memref.store %arg1, %arg0[] : memref + return +} + +// ----- + +// CHECK-LABEL: func @static_store +// CHECK: %[[MEMREF:.*]]: memref<10x42xf32>, +// CHECK-SAME: %[[I:.*]]: index, %[[J:.*]]: index, +func @static_store(%static : memref<10x42xf32>, %i : index, %j : index, %val : f32) { +// CHECK: %[[II:.*]] = llvm.mlir.cast %[[I]] +// CHECK: %[[JJ:.*]] = llvm.mlir.cast %[[J]] +// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// CHECK: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 +// CHECK: %[[offI:.*]] = llvm.mul %[[II]], %[[st0]] : i64 +// CHECK: %[[off1:.*]] = llvm.add %[[offI]], %[[JJ]] : i64 +// CHECK: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr +// CHECK: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr + + memref.store %val, %static[%i, %j] : memref<10x42xf32> + return +} + +// ----- + +// CHECK-LABEL: func @static_memref_dim +func @static_memref_dim(%static : memref<42x32x15x13x27xf32>) { +// CHECK: llvm.mlir.constant(42 : index) : i64 + %c0 = constant 0 : index + %0 = memref.dim %static, %c0 : memref<42x32x15x13x27xf32> +// CHECK: llvm.mlir.constant(32 : index) : i64 + %c1 = constant 1 : index + %1 = memref.dim %static, %c1 : memref<42x32x15x13x27xf32> +// CHECK: llvm.mlir.constant(15 : index) : i64 + %c2 = constant 2 : index + %2 = memref.dim %static, %c2 : memref<42x32x15x13x27xf32> +// CHECK: llvm.mlir.constant(13 : index) : i64 + %c3 = constant 3 : index + %3 = memref.dim %static, %c3 : memref<42x32x15x13x27xf32> +// CHECK: llvm.mlir.constant(27 : index) : i64 + %c4 = constant 4 : index + %4 = memref.dim %static, %c4 : memref<42x32x15x13x27xf32> + return +} + +// ----- + +// Check that consistent types are emitted in address arithemic in presence of +// a data layout specification. +module attributes { dlti.dl_spec = #dlti.dl_spec<#dlti.dl_entry> } { + func @address() { + %c1 = constant 1 : index + %0 = memref.alloc(%c1) : memref> + // CHECK: %[[CST_S:.*]] = constant 1 : index + // CHECK: %[[CST:.*]] = llvm.mlir.cast + // CHECK: llvm.mlir.null + // CHECK: llvm.getelementptr %{{.*}}[[CST]] + // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 + // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 + // CHECK: llvm.add %{{.*}} : i32 + // CHECK: llvm.call @malloc(%{{.*}}) : (i32) -> !llvm.ptr + // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 + // CHECK: llvm.sub {{.*}} : i32 + // CHECK: llvm.add {{.*}} : i32 + // CHECK: llvm.urem {{.*}} : i32 + // CHECK: llvm.sub {{.*}} : i32 + // CHECK: llvm.inttoptr %{{.*}} : i32 to !llvm.ptr + return + } +} + diff --git a/mlir/test/Conversion/MemRefToLLVM/memref-to-llvm.mlir b/mlir/test/Conversion/MemRefToLLVM/memref-to-llvm.mlir new file mode 100644 --- /dev/null +++ b/mlir/test/Conversion/MemRefToLLVM/memref-to-llvm.mlir @@ -0,0 +1,700 @@ +// RUN: mlir-opt -convert-memref-to-llvm %s -split-input-file | FileCheck %s +// RUN: mlir-opt -convert-memref-to-llvm='index-bitwidth=32' %s -split-input-file | FileCheck --check-prefix=CHECK32 %s + + +// CHECK-LABEL: func @view( +// CHECK: %[[ARG0F:.*]]: index, %[[ARG1F:.*]]: index, %[[ARG2F:.*]]: index +func @view(%arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK: llvm.mlir.constant(2048 : index) : i64 + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + %0 = memref.alloc() : memref<2048xi8> + + // Test two dynamic sizes. + // CHECK: %[[ARG2:.*]] = llvm.mlir.cast %[[ARG2F:.*]] + // CHECK: %[[ARG0:.*]] = llvm.mlir.cast %[[ARG0F:.*]] + // CHECK: %[[ARG1:.*]] = llvm.mlir.cast %[[ARG1F:.*]] + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BASE_PTR:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: %[[SHIFTED_BASE_PTR:.*]] = llvm.getelementptr %[[BASE_PTR]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr + // CHECK: %[[CAST_SHIFTED_BASE_PTR:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR]] : !llvm.ptr to !llvm.ptr + // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C0:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[C0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[ARG1]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(1 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[ARG0]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mul %{{.*}}, %[[ARG1]] + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %1 = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref + + // Test one dynamic size. + // CHECK: %[[ARG2:.*]] = llvm.mlir.cast %[[ARG2F:.*]] + // CHECK: %[[ARG1:.*]] = llvm.mlir.cast %[[ARG1F:.*]] + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BASE_PTR_2:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: %[[SHIFTED_BASE_PTR_2:.*]] = llvm.getelementptr %[[BASE_PTR_2]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr + // CHECK: %[[CAST_SHIFTED_BASE_PTR_2:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_2]] : !llvm.ptr to !llvm.ptr + // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_2]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C0_2:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[C0_2]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[ARG1]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(1 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(4 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mul %{{.*}}, %[[ARG1]] + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %3 = memref.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32> + + // Test static sizes. + // CHECK: %[[ARG2:.*]] = llvm.mlir.cast %[[ARG2F:.*]] + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BASE_PTR_3:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: %[[SHIFTED_BASE_PTR_3:.*]] = llvm.getelementptr %[[BASE_PTR_3]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr + // CHECK: %[[CAST_SHIFTED_BASE_PTR_3:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_3]] : !llvm.ptr to !llvm.ptr + // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_3]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C0_3:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[C0_3]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(4 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(1 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(64 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(4 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %5 = memref.view %0[%arg2][] : memref<2048xi8> to memref<64x4xf32> + + // Test view memory space. + // CHECK: llvm.mlir.constant(2048 : index) : i64 + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + %6 = memref.alloc() : memref<2048xi8, 4> + + // CHECK: %[[ARG2:.*]] = llvm.mlir.cast %[[ARG2F:.*]] + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BASE_PTR_4:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: %[[SHIFTED_BASE_PTR_4:.*]] = llvm.getelementptr %[[BASE_PTR_4]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr + // CHECK: %[[CAST_SHIFTED_BASE_PTR_4:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_4]] : !llvm.ptr to !llvm.ptr + // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_4]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C0_4:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[C0_4]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(4 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(1 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(64 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.mlir.constant(4 : index) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %7 = memref.view %6[%arg2][] : memref<2048xi8, 4> to memref<64x4xf32, 4> + + return +} + +// ----- + +// CHECK-LABEL: func @subview( +// CHECK: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG2f:.*]]: index) +// CHECK32-LABEL: func @subview( +// CHECK32: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK32: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG2f:.*]]: index) +func @subview(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + + // CHECK: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0c:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1c:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0c:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1c:.*]] = llvm.mlir.cast %[[ARG1f]] + + // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i64 + // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 + // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i64 + // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 + // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1c]], %[[STRIDE1]] : i64 + // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0c]], %[[STRIDE0]] : i64 + // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i32 + // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 + // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i32 + // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1c]], %[[STRIDE1]] : i32 + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0c]], %[[STRIDE0]] : i32 + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + + %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] : + memref<64x4xf32, offset: 0, strides: [4, 1]> + to memref + return +} + +// ----- + +// CHECK-LABEL: func @subview_non_zero_addrspace( +// CHECK: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG2f:.*]]: index) +// CHECK32-LABEL: func @subview_non_zero_addrspace( +// CHECK32: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK32: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG2f:.*]]: index) +func @subview_non_zero_addrspace(%0 : memref<64x4xf32, offset: 0, strides: [4, 1], 3>, %arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + + // CHECK: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0c:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1c:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0c:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1c:.*]] = llvm.mlir.cast %[[ARG1f]] + + // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i64 + // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 + // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i64 + // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 + // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1c]], %[[STRIDE1]] : i64 + // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0c]], %[[STRIDE0]] : i64 + // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i32 + // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 + // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i32 + // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1c]], %[[STRIDE1]] : i32 + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0c]], %[[STRIDE0]] : i32 + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + + %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] : + memref<64x4xf32, offset: 0, strides: [4, 1], 3> + to memref + return +} + +// ----- + +// CHECK-LABEL: func @subview_const_size( +// CHECK-SAME: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK-SAME: %[[ARG0f:[a-zA-Z0-9]*]]: index +// CHECK-SAME: %[[ARG1f:[a-zA-Z0-9]*]]: index +// CHECK-SAME: %[[ARG2f:[a-zA-Z0-9]*]]: index +// CHECK32-LABEL: func @subview_const_size( +// CHECK32-SAME: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK32-SAME: %[[ARG0f:[a-zA-Z0-9]*]]: index +// CHECK32-SAME: %[[ARG1f:[a-zA-Z0-9]*]]: index +// CHECK32-SAME: %[[ARG2f:[a-zA-Z0-9]*]]: index +func @subview_const_size(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + + // CHECK: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + + // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i64 + // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 + // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i64 + // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 + // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) + // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1b]], %[[STRIDE1]] : i64 + // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[CST2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) + // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0b]], %[[STRIDE0]] : i64 + // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[CST4]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i32 + // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 + // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i32 + // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) + // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1b]], %[[STRIDE1]] : i32 + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[CST2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) + // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0b]], %[[STRIDE0]] : i32 + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST4]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + %1 = memref.subview %0[%arg0, %arg1][4, 2][%arg0, %arg1] : + memref<64x4xf32, offset: 0, strides: [4, 1]> + to memref<4x2xf32, offset: ?, strides: [?, ?]> + return +} + +// ----- + +// CHECK-LABEL: func @subview_const_stride( +// CHECK-SAME: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK-SAME: %[[ARG0f:[a-zA-Z0-9]*]]: index +// CHECK-SAME: %[[ARG1f:[a-zA-Z0-9]*]]: index +// CHECK-SAME: %[[ARG2f:[a-zA-Z0-9]*]]: index +// CHECK32-LABEL: func @subview_const_stride( +// CHECK32-SAME: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK32-SAME: %[[ARG0f:[a-zA-Z0-9]*]]: index +// CHECK32-SAME: %[[ARG1f:[a-zA-Z0-9]*]]: index +// CHECK32-SAME: %[[ARG2f:[a-zA-Z0-9]*]]: index +func @subview_const_stride(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + + // CHECK: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0a:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1a:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG0b:.*]] = llvm.mlir.cast %[[ARG0f]] + // CHECK32: %[[ARG1b:.*]] = llvm.mlir.cast %[[ARG1f]] + + // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i64 + // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 + // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i64 + // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 + // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) + // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[CST2]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) + // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0a]], %[[STRIDE0]] : i32 + // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 + // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1a]], %[[STRIDE1]] : i32 + // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1b]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST2]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0b]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][1, 2] : + memref<64x4xf32, offset: 0, strides: [4, 1]> + to memref + return +} + +// ----- + +// CHECK-LABEL: func @subview_const_stride_and_offset( +// CHECK32-LABEL: func @subview_const_stride_and_offset( +func @subview_const_stride_and_offset(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>) { + // The last "insertvalue" that populates the memref descriptor from the function arguments. + // CHECK: %[[MEMREF:.*]] = llvm.mlir.cast + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast + + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST8:.*]] = llvm.mlir.constant(8 : index) + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[CST8]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST3:.*]] = llvm.mlir.constant(3 : i64) + // CHECK32: %[[CST1:.*]] = llvm.mlir.constant(1 : i64) + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[CST3]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST62:.*]] = llvm.mlir.constant(62 : i64) + // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + %1 = memref.subview %0[0, 8][62, 3][1, 1] : + memref<64x4xf32, offset: 0, strides: [4, 1]> + to memref<62x3xf32, offset: 8, strides: [4, 1]> + return +} + +// ----- + +// CHECK-LABEL: func @subview_mixed_static_dynamic( +// CHECK: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK: %[[ARG2f:.*]]: index) +// CHECK32-LABEL: func @subview_mixed_static_dynamic( +// CHECK32: %[[MEM:.*]]: memref<{{.*}}>, +// CHECK32: %[[ARG0f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG1f:[a-zA-Z0-9]*]]: index, +// CHECK32: %[[ARG2f:.*]]: index) +func @subview_mixed_static_dynamic(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { + // CHECK32: %[[MEMREF:.*]] = llvm.mlir.cast %[[MEM]] + // CHECK32: %[[ARG1:.*]] = llvm.mlir.cast %[[ARG1f]] + // CHECK32: %[[ARG2:.*]] = llvm.mlir.cast %[[ARG2f]] + // CHECK32: %[[ARG0:.*]] = llvm.mlir.cast %[[ARG0f]] + + // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr + // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[OFFM1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE0]] : i32 + // CHECK32: %[[OFFA1:.*]] = llvm.add %[[OFF]], %[[OFFM1]] : i32 + // CHECK32: %[[CST8:.*]] = llvm.mlir.constant(8 : i64) : i32 + // CHECK32: %[[OFFM2:.*]] = llvm.mul %[[CST8]], %[[STRIDE1]] : i32 + // CHECK32: %[[OFFA2:.*]] = llvm.add %[[OFFA1]], %[[OFFM2]] : i32 + // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFFA2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + + // CHECK32: %[[CST1:.*]] = llvm.mlir.constant(1 : i64) : i32 + // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: %[[CST62:.*]] = llvm.mlir.constant(62 : i64) : i32 + // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 + // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + // CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> + %1 = memref.subview %0[%arg1, 8][62, %arg2][%arg0, 1] : + memref<64x4xf32, offset: 0, strides: [4, 1]> + to memref<62x?xf32, offset: ?, strides: [?, 1]> + return +} + +// ----- + +// CHECK-LABEL: func @subview_leading_operands( +func @subview_leading_operands(%0 : memref<5x3xf32>, %1: memref<5x?xf32>) { + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Alloc ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Aligned ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Offset + // CHECK: %[[C6:.*]] = llvm.mlir.constant(6 : index) : i64 + // CHECK: llvm.insertvalue %[[C6:.*]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Sizes and strides @rank 1: both static. + // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : i64) : i64 + // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 + // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Sizes and strides @rank 0: both static extracted from type. + // CHECK: %[[C3_2:.*]] = llvm.mlir.constant(3 : i64) : i64 + // CHECK: %[[C3_3:.*]] = llvm.mlir.constant(3 : i64) : i64 + // CHECK: llvm.insertvalue %[[C3_2]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[C3_3]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %2 = memref.subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]> + + return +} + +// ----- + +// CHECK-LABEL: func @subview_leading_operands_dynamic( +func @subview_leading_operands_dynamic(%0 : memref<5x?xf32>) { + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Alloc ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Aligned ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Extract strides + // CHECK: %[[ST0:.*]] = llvm.extractvalue %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[ST1:.*]] = llvm.extractvalue %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Compute and insert offset from 2 + dynamic value. + // CHECK: %[[OFF:.*]] = llvm.extractvalue %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C2:.*]] = llvm.mlir.constant(2 : i64) : i64 + // CHECK: %[[MUL:.*]] = llvm.mul %[[C2]], %[[ST0]] : i64 + // CHECK: %[[NEW_OFF:.*]] = llvm.add %[[OFF]], %[[MUL]] : i64 + // CHECK: llvm.insertvalue %[[NEW_OFF]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Sizes and strides @rank 1: static stride 1, dynamic size unchanged from source memref. + // CHECK: %[[SZ1:.*]] = llvm.extractvalue %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Sizes and strides @rank 0: both static. + // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : i64) : i64 + // CHECK: %[[C1_2:.*]] = llvm.mlir.constant(1 : i64) : i64 + // CHECK: %[[MUL:.*]] = llvm.mul %[[C1_2]], %[[ST0]] : i64 + // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[MUL]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + %1 = memref.subview %0[2][3][1]: memref<5x?xf32> to memref<3x?xf32, offset: ?, strides: [?, 1]> + return +} + +// ----- + +// CHECK-LABEL: func @subview_rank_reducing_leading_operands( +func @subview_rank_reducing_leading_operands(%0 : memref<5x3xf32>) { + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // Alloc ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // Aligned ptr + // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // Extract strides + // CHECK: %[[ST0:.*]] = llvm.extractvalue %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[ST1:.*]] = llvm.extractvalue %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // Offset + // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : index) : i64 + // CHECK: llvm.insertvalue %[[C3:.*]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // Sizes and strides @rank 0: both static. + // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : i64) : i64 + // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 + // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + %1 = memref.subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]> + + return +} + +// ----- + +// CHECK-LABEL: func @assume_alignment +func @assume_alignment(%0 : memref<4x4xf16>) { + // CHECK: %[[PTR:.*]] = llvm.extractvalue %[[MEMREF:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK-NEXT: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK-NEXT: %[[MASK:.*]] = llvm.mlir.constant(15 : index) : i64 + // CHECK-NEXT: %[[INT:.*]] = llvm.ptrtoint %[[PTR]] : !llvm.ptr to i64 + // CHECK-NEXT: %[[MASKED_PTR:.*]] = llvm.and %[[INT]], %[[MASK:.*]] : i64 + // CHECK-NEXT: %[[CONDITION:.*]] = llvm.icmp "eq" %[[MASKED_PTR]], %[[ZERO]] : i64 + // CHECK-NEXT: "llvm.intr.assume"(%[[CONDITION]]) : (i1) -> () + memref.assume_alignment %0, 16 : memref<4x4xf16> + return +} + +// ----- + +// CHECK-LABEL: func @dim_of_unranked +// CHECK32-LABEL: func @dim_of_unranked +func @dim_of_unranked(%unranked: memref<*xi32>) -> index { + %c0 = constant 0 : index + %dim = memref.dim %unranked, %c0 : memref<*xi32> + return %dim : index +} +// CHECK: %[[UNRANKED_DESC:.*]] = llvm.mlir.cast + +// CHECK: %[[RANKED_DESC:.*]] = llvm.extractvalue %[[UNRANKED_DESC]][1] +// CHECK-SAME: : !llvm.struct<(i64, ptr)> + +// CHECK: %[[ZERO_D_DESC:.*]] = llvm.bitcast %[[RANKED_DESC]] +// CHECK-SAME: : !llvm.ptr to !llvm.ptr, ptr, i64)>> + +// CHECK: %[[C2_i32:.*]] = llvm.mlir.constant(2 : i32) : i32 +// CHECK: %[[C0_:.*]] = llvm.mlir.constant(0 : index) : i64 + +// CHECK: %[[OFFSET_PTR:.*]] = llvm.getelementptr %[[ZERO_D_DESC]]{{\[}} +// CHECK-SAME: %[[C0_]], %[[C2_i32]]] : (!llvm.ptr, ptr, +// CHECK-SAME: i64)>>, i64, i32) -> !llvm.ptr + +// CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : index) : i64 +// CHECK: %[[INDEX_INC:.*]] = llvm.add %[[C1]], %{{.*}} : i64 + +// CHECK: %[[SIZE_PTR:.*]] = llvm.getelementptr %[[OFFSET_PTR]]{{\[}} +// CHECK-SAME: %[[INDEX_INC]]] : (!llvm.ptr, i64) -> !llvm.ptr + +// CHECK: %[[SIZE:.*]] = llvm.load %[[SIZE_PTR]] : !llvm.ptr + +// CHECK32: %[[SIZE:.*]] = llvm.load %{{.*}} : !llvm.ptr + +// ----- + +// CHECK-LABEL: func @address_space( +func @address_space(%arg0 : memref<32xf32, affine_map<(d0) -> (d0)>, 7>) { + %0 = memref.alloc() : memref<32xf32, affine_map<(d0) -> (d0)>, 5> + %1 = constant 7 : index + // CHECK: llvm.load %{{.*}} : !llvm.ptr + %2 = memref.load %0[%1] : memref<32xf32, affine_map<(d0) -> (d0)>, 5> + std.return +} + +// ----- + +// CHECK-LABEL: func @transpose +// CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.extractvalue {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[3, 2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.extractvalue {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.extractvalue {{.*}}[3, 2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +// CHECK: llvm.insertvalue {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> +func @transpose(%arg0: memref) { + %0 = memref.transpose %arg0 (i, j, k) -> (k, i, j) : memref to memref (d2 * s1 + s0 + d0 * s2 + d1)>> + return +} + +// ----- + +// CHECK: llvm.mlir.global external @gv0() : !llvm.array<2 x f32> +memref.global @gv0 : memref<2xf32> = uninitialized + +// CHECK: llvm.mlir.global private @gv1() : !llvm.array<2 x f32> +memref.global "private" @gv1 : memref<2xf32> + +// CHECK: llvm.mlir.global external @gv2(dense<{{\[\[}}0.000000e+00, 1.000000e+00, 2.000000e+00], [3.000000e+00, 4.000000e+00, 5.000000e+00]]> : tensor<2x3xf32>) : !llvm.array<2 x array<3 x f32>> +memref.global @gv2 : memref<2x3xf32> = dense<[[0.0, 1.0, 2.0], [3.0, 4.0, 5.0]]> + +// Test 1D memref. +// CHECK-LABEL: func @get_gv0_memref +func @get_gv0_memref() { + %0 = memref.get_global @gv0 : memref<2xf32> + // CHECK: %[[DIM:.*]] = llvm.mlir.constant(2 : index) : i64 + // CHECK: %[[STRIDE:.*]] = llvm.mlir.constant(1 : index) : i64 + // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv0 : !llvm.ptr> + // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]], %[[ZERO]]] : (!llvm.ptr>, i64, i64) -> !llvm.ptr + // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 + // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: llvm.insertvalue %[[DIM]], {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // CHECK: llvm.insertvalue %[[STRIDE]], {{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + return +} + +// Test 2D memref. +// CHECK-LABEL: func @get_gv2_memref +func @get_gv2_memref() { + // CHECK: %[[DIM0:.*]] = llvm.mlir.constant(2 : index) : i64 + // CHECK: %[[DIM1:.*]] = llvm.mlir.constant(3 : index) : i64 + // CHECK: %[[STRIDE1:.*]] = llvm.mlir.constant(1 : index) : i64 + // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv2 : !llvm.ptr>> + // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]], %[[ZERO]], %[[ZERO]]] : (!llvm.ptr>>, i64, i64, i64) -> !llvm.ptr + // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 + // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DIM0]], {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DIM1]], {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[DIM1]], {{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + // CHECK: llvm.insertvalue %[[STRIDE1]], {{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + + %0 = memref.get_global @gv2 : memref<2x3xf32> + return +} + +// Test scalar memref. +// CHECK: llvm.mlir.global external @gv3(1.000000e+00 : f32) : f32 +memref.global @gv3 : memref = dense<1.0> + +// CHECK-LABEL: func @get_gv3_memref +func @get_gv3_memref() { + // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv3 : !llvm.ptr + // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]]] : (!llvm.ptr, i64) -> !llvm.ptr + // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 + // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr + // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64)> + // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> + // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> + // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 + // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64)> + %0 = memref.get_global @gv3 : memref + return +} + diff --git a/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir b/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir --- a/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir +++ b/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir @@ -1,5 +1,5 @@ -// RUN: mlir-opt -convert-std-to-llvm='emit-c-wrappers=1' %s | FileCheck %s -// RUN: mlir-opt -convert-std-to-llvm %s | FileCheck %s --check-prefix=EMIT_C_ATTRIBUTE +// RUN: mlir-opt -convert-memref-to-llvm -convert-std-to-llvm='emit-c-wrappers=1' %s | FileCheck %s +// RUN: mlir-opt -convert-memref-to-llvm -convert-std-to-llvm %s | FileCheck %s --check-prefix=EMIT_C_ATTRIBUTE // This tests the default memref calling convention and the emission of C // wrappers. We don't need to separate runs because the wrapper-emission diff --git a/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir b/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir --- a/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir +++ b/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir @@ -4,8 +4,6 @@ // When expanding the memref to multiple arguments, argument attributes are replicated. // CHECK-COUNT-7: {dialect.a = true, dialect.b = 4 : i64} func @check_attributes(%static: memref<10x20xf32> {dialect.a = true, dialect.b = 4 : i64 }) { - %c0 = constant 0 : index - %0 = memref.load %static[%c0, %c0]: memref<10x20xf32> return } diff --git a/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir b/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir deleted file mode 100644 --- a/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir +++ /dev/null @@ -1,450 +0,0 @@ -// RUN: mlir-opt -convert-std-to-llvm -split-input-file %s | FileCheck %s -// RUN: mlir-opt -convert-std-to-llvm='use-bare-ptr-memref-call-conv=1' -split-input-file %s | FileCheck %s --check-prefix=BAREPTR - -// BAREPTR-LABEL: func @check_noalias -// BAREPTR-SAME: %{{.*}}: !llvm.ptr {llvm.noalias}, %{{.*}}: !llvm.ptr {llvm.noalias} -func @check_noalias(%static : memref<2xf32> {llvm.noalias}, %other : memref<2xf32> {llvm.noalias}) { - return -} - -// ----- - -// CHECK-LABEL: func @check_static_return -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-LABEL: func @check_static_return -// BAREPTR-SAME: (%[[arg:.*]]: !llvm.ptr) -> !llvm.ptr { -func @check_static_return(%static : memref<32x18xf32>) -> memref<32x18xf32> { -// CHECK: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - -// BAREPTR: %[[udf:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[base0:.*]] = llvm.insertvalue %[[arg]], %[[udf]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[aligned:.*]] = llvm.insertvalue %[[arg]], %[[base0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val0:.*]] = llvm.mlir.constant(0 : index) : i64 -// BAREPTR-NEXT: %[[ins0:.*]] = llvm.insertvalue %[[val0]], %[[aligned]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val1:.*]] = llvm.mlir.constant(32 : index) : i64 -// BAREPTR-NEXT: %[[ins1:.*]] = llvm.insertvalue %[[val1]], %[[ins0]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val2:.*]] = llvm.mlir.constant(18 : index) : i64 -// BAREPTR-NEXT: %[[ins2:.*]] = llvm.insertvalue %[[val2]], %[[ins1]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val3:.*]] = llvm.mlir.constant(18 : index) : i64 -// BAREPTR-NEXT: %[[ins3:.*]] = llvm.insertvalue %[[val3]], %[[ins2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val4:.*]] = llvm.mlir.constant(1 : index) : i64 -// BAREPTR-NEXT: %[[ins4:.*]] = llvm.insertvalue %[[val4]], %[[ins3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[base1:.*]] = llvm.extractvalue %[[ins4]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: llvm.return %[[base1]] : !llvm.ptr - return %static : memref<32x18xf32> -} - -// ----- - -// CHECK-LABEL: func @check_static_return_with_offset -// CHECK-COUNT-2: !llvm.ptr -// CHECK-COUNT-5: i64 -// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-LABEL: func @check_static_return_with_offset -// BAREPTR-SAME: (%[[arg:.*]]: !llvm.ptr) -> !llvm.ptr { -func @check_static_return_with_offset(%static : memref<32x18xf32, offset:7, strides:[22,1]>) -> memref<32x18xf32, offset:7, strides:[22,1]> { -// CHECK: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - -// BAREPTR: %[[udf:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[base0:.*]] = llvm.insertvalue %[[arg]], %[[udf]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[aligned:.*]] = llvm.insertvalue %[[arg]], %[[base0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val0:.*]] = llvm.mlir.constant(7 : index) : i64 -// BAREPTR-NEXT: %[[ins0:.*]] = llvm.insertvalue %[[val0]], %[[aligned]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val1:.*]] = llvm.mlir.constant(32 : index) : i64 -// BAREPTR-NEXT: %[[ins1:.*]] = llvm.insertvalue %[[val1]], %[[ins0]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val2:.*]] = llvm.mlir.constant(22 : index) : i64 -// BAREPTR-NEXT: %[[ins2:.*]] = llvm.insertvalue %[[val2]], %[[ins1]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val3:.*]] = llvm.mlir.constant(18 : index) : i64 -// BAREPTR-NEXT: %[[ins3:.*]] = llvm.insertvalue %[[val3]], %[[ins2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[val4:.*]] = llvm.mlir.constant(1 : index) : i64 -// BAREPTR-NEXT: %[[ins4:.*]] = llvm.insertvalue %[[val4]], %[[ins3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[base1:.*]] = llvm.extractvalue %[[ins4]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: llvm.return %[[base1]] : !llvm.ptr - return %static : memref<32x18xf32, offset:7, strides:[22,1]> -} - -// ----- - -// CHECK-LABEL: func @zero_d_alloc() -> !llvm.struct<(ptr, ptr, i64)> { -// BAREPTR-LABEL: func @zero_d_alloc() -> !llvm.ptr { -func @zero_d_alloc() -> memref { -// CHECK-NEXT: %[[one:.*]] = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// CHECK-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[one]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// CHECK-NEXT: llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr -// CHECK-NEXT: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr -// CHECK-NEXT: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64)> - -// BAREPTR-NEXT: %[[one:.*]] = llvm.mlir.constant(1 : index) : i64 -// BAREPTR-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// BAREPTR-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[one]]] : (!llvm.ptr, i64) -> !llvm.ptr -// BAREPTR-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// BAREPTR-NEXT: llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr -// BAREPTR-NEXT: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr -// BAREPTR-NEXT: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 -// BAREPTR-NEXT: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64)> - %0 = memref.alloc() : memref - return %0 : memref -} - -// ----- - -// CHECK-LABEL: func @zero_d_dealloc -// BAREPTR-LABEL: func @zero_d_dealloc(%{{.*}}: !llvm.ptr) { -func @zero_d_dealloc(%arg0: memref) { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr -// CHECK-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr -// BAREPTR-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () - memref.dealloc %arg0 : memref - return -} - -// ----- - -// CHECK-LABEL: func @aligned_1d_alloc( -// BAREPTR-LABEL: func @aligned_1d_alloc( -func @aligned_1d_alloc() -> memref<42xf32> { -// CHECK-NEXT: %[[sz1:.*]] = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: %[[st1:.*]] = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// CHECK-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[sz1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// CHECK-NEXT: %[[alignment:.*]] = llvm.mlir.constant(8 : index) : i64 -// CHECK-NEXT: %[[allocsize:.*]] = llvm.add %[[size_bytes]], %[[alignment]] : i64 -// CHECK-NEXT: %[[allocated:.*]] = llvm.call @malloc(%[[allocsize]]) : (i64) -> !llvm.ptr -// CHECK-NEXT: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr -// CHECK-NEXT: %[[allocatedAsInt:.*]] = llvm.ptrtoint %[[ptr]] : !llvm.ptr to i64 -// CHECK-NEXT: %[[one_1:.*]] = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: %[[bump:.*]] = llvm.sub %[[alignment]], %[[one_1]] : i64 -// CHECK-NEXT: %[[bumped:.*]] = llvm.add %[[allocatedAsInt]], %[[bump]] : i64 -// CHECK-NEXT: %[[mod:.*]] = llvm.urem %[[bumped]], %[[alignment]] : i64 -// CHECK-NEXT: %[[aligned:.*]] = llvm.sub %[[bumped]], %[[mod]] : i64 -// CHECK-NEXT: %[[alignedBitCast:.*]] = llvm.inttoptr %[[aligned]] : i64 to !llvm.ptr -// CHECK-NEXT: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// CHECK-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// CHECK-NEXT: llvm.insertvalue %[[alignedBitCast]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// CHECK-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - -// BAREPTR-NEXT: %[[sz1:.*]] = llvm.mlir.constant(42 : index) : i64 -// BAREPTR-NEXT: %[[st1:.*]] = llvm.mlir.constant(1 : index) : i64 -// BAREPTR-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// BAREPTR-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[sz1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// BAREPTR-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// BAREPTR-NEXT: %[[alignment:.*]] = llvm.mlir.constant(8 : index) : i64 -// BAREPTR-NEXT: %[[allocsize:.*]] = llvm.add %[[size_bytes]], %[[alignment]] : i64 -// BAREPTR-NEXT: %[[allocated:.*]] = llvm.call @malloc(%[[allocsize]]) : (i64) -> !llvm.ptr -// BAREPTR-NEXT: %[[ptr:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr -// BAREPTR-NEXT: %[[allocatedAsInt:.*]] = llvm.ptrtoint %[[ptr]] : !llvm.ptr to i64 -// BAREPTR-NEXT: %[[one_2:.*]] = llvm.mlir.constant(1 : index) : i64 -// BAREPTR-NEXT: %[[bump:.*]] = llvm.sub %[[alignment]], %[[one_2]] : i64 -// BAREPTR-NEXT: %[[bumped:.*]] = llvm.add %[[allocatedAsInt]], %[[bump]] : i64 -// BAREPTR-NEXT: %[[mod:.*]] = llvm.urem %[[bumped]], %[[alignment]] : i64 -// BAREPTR-NEXT: %[[aligned:.*]] = llvm.sub %[[bumped]], %[[mod]] : i64 -// BAREPTR-NEXT: %[[alignedBitCast:.*]] = llvm.inttoptr %[[aligned]] : i64 to !llvm.ptr -// BAREPTR-NEXT: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// BAREPTR-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// BAREPTR-NEXT: llvm.insertvalue %[[alignedBitCast]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// BAREPTR-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 -// BAREPTR-NEXT: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - %0 = memref.alloc() {alignment = 8} : memref<42xf32> - return %0 : memref<42xf32> -} - -// ----- - -// CHECK-LABEL: func @static_alloc() -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { -// BAREPTR-LABEL: func @static_alloc() -> !llvm.ptr { -func @static_alloc() -> memref<32x18xf32> { -// CHECK: %[[num_elems:.*]] = llvm.mlir.constant(576 : index) : i64 -// CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// CHECK-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// CHECK-NEXT: %[[allocated:.*]] = llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr -// CHECK-NEXT: llvm.bitcast %[[allocated]] : !llvm.ptr to !llvm.ptr - -// BAREPTR: %[[num_elems:.*]] = llvm.mlir.constant(576 : index) : i64 -// BAREPTR-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// BAREPTR-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr, i64) -> !llvm.ptr -// BAREPTR-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// BAREPTR-NEXT: %[[allocated:.*]] = llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr -// BAREPTR-NEXT: llvm.bitcast %[[allocated]] : !llvm.ptr to !llvm.ptr - %0 = memref.alloc() : memref<32x18xf32> - return %0 : memref<32x18xf32> -} - -// ----- - -// CHECK-LABEL: func @static_alloca() -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> { -func @static_alloca() -> memref<32x18xf32> { -// CHECK-NEXT: %[[sz1:.*]] = llvm.mlir.constant(32 : index) : i64 -// CHECK-NEXT: %[[sz2:.*]] = llvm.mlir.constant(18 : index) : i64 -// CHECK-NEXT: %[[st2:.*]] = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: %[[num_elems:.*]] = llvm.mlir.constant(576 : index) : i64 -// CHECK-NEXT: %[[null:.*]] = llvm.mlir.null : !llvm.ptr -// CHECK-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr to i64 -// CHECK-NEXT: %[[allocated:.*]] = llvm.alloca %[[size_bytes]] x f32 : (i64) -> !llvm.ptr - %0 = memref.alloca() : memref<32x18xf32> - - // Test with explicitly specified alignment. llvm.alloca takes care of the - // alignment. The same pointer is thus used for allocation and aligned - // accesses. - // CHECK: %[[alloca_aligned:.*]] = llvm.alloca %{{.*}} x f32 {alignment = 32 : i64} : (i64) -> !llvm.ptr - // CHECK: %[[desc:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[desc1:.*]] = llvm.insertvalue %[[alloca_aligned]], %[[desc]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[alloca_aligned]], %[[desc1]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - memref.alloca() {alignment = 32} : memref<32x18xf32> - return %0 : memref<32x18xf32> -} - -// ----- - -// CHECK-LABEL: func @static_dealloc -// BAREPTR-LABEL: func @static_dealloc(%{{.*}}: !llvm.ptr) { -func @static_dealloc(%static: memref<10x8xf32>) { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// CHECK-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr -// CHECK-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr to !llvm.ptr -// BAREPTR-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr) -> () - memref.dealloc %static : memref<10x8xf32> - return -} - -// ----- - -// CHECK-LABEL: func @zero_d_load -// BAREPTR-LABEL: func @zero_d_load(%{{.*}}: !llvm.ptr) -> f32 -func @zero_d_load(%arg0: memref) -> f32 { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: %{{.*}} = llvm.load %[[ptr]] : !llvm.ptr - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: llvm.load %[[ptr:.*]] : !llvm.ptr - %0 = memref.load %arg0[] : memref - return %0 : f32 -} - -// ----- - -// CHECK-LABEL: func @static_load( -// CHECK-COUNT-2: !llvm.ptr, -// CHECK-COUNT-5: {{%[a-zA-Z0-9]*}}: i64 -// CHECK: %[[I:.*]]: i64, -// CHECK: %[[J:.*]]: i64) -// BAREPTR-LABEL: func @static_load -// BAREPTR-SAME: (%[[A:.*]]: !llvm.ptr, %[[I:.*]]: i64, %[[J:.*]]: i64) { -func @static_load(%static : memref<10x42xf32>, %i : index, %j : index) { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// CHECK-NEXT: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 -// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64 -// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: llvm.load %[[addr]] : !llvm.ptr - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 -// BAREPTR-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 -// BAREPTR-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64 -// BAREPTR-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// BAREPTR-NEXT: llvm.load %[[addr]] : !llvm.ptr - %0 = memref.load %static[%i, %j] : memref<10x42xf32> - return -} - -// ----- - -// CHECK-LABEL: func @zero_d_store -// BAREPTR-LABEL: func @zero_d_store -// BAREPTR-SAME: (%[[A:.*]]: !llvm.ptr, %[[val:.*]]: f32) -func @zero_d_store(%arg0: memref, %arg1: f32) { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %[[ld:.*]][1] : !llvm.struct<(ptr, ptr, i64)> -// CHECK-NEXT: llvm.store %{{.*}}, %[[ptr]] : !llvm.ptr - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> -// BAREPTR-NEXT: llvm.store %[[val]], %[[ptr]] : !llvm.ptr - memref.store %arg1, %arg0[] : memref - return -} - -// ----- - -// CHECK-LABEL: func @static_store -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[J:[a-zA-Z0-9]*]]: i64 -// BAREPTR-LABEL: func @static_store -// BAREPTR-SAME: %[[A:.*]]: !llvm.ptr -// BAREPTR-SAME: %[[I:[a-zA-Z0-9]*]]: i64 -// BAREPTR-SAME: %[[J:[a-zA-Z0-9]*]]: i64 -func @static_store(%static : memref<10x42xf32>, %i : index, %j : index, %val : f32) { -// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// CHECK-NEXT: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 -// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64 -// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// CHECK-NEXT: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr - -// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> -// BAREPTR-NEXT: %[[st0:.*]] = llvm.mlir.constant(42 : index) : i64 -// BAREPTR-NEXT: %[[offI:.*]] = llvm.mul %[[I]], %[[st0]] : i64 -// BAREPTR-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64 -// BAREPTR-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr, i64) -> !llvm.ptr -// BAREPTR-NEXT: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr - memref.store %val, %static[%i, %j] : memref<10x42xf32> - return -} - -// ----- - -// CHECK-LABEL: func @static_memref_dim -// BAREPTR-LABEL: func @static_memref_dim(%{{.*}}: !llvm.ptr) { -func @static_memref_dim(%static : memref<42x32x15x13x27xf32>) { -// CHECK: llvm.mlir.constant(42 : index) : i64 -// BAREPTR: llvm.insertvalue %{{.*}}, %{{.*}}[4, 4] : !llvm.struct<(ptr, ptr, i64, array<5 x i64>, array<5 x i64>)> -// BAREPTR: llvm.mlir.constant(42 : index) : i64 - %c0 = constant 0 : index - %0 = memref.dim %static, %c0 : memref<42x32x15x13x27xf32> -// CHECK: llvm.mlir.constant(32 : index) : i64 -// BAREPTR: llvm.mlir.constant(32 : index) : i64 - %c1 = constant 1 : index - %1 = memref.dim %static, %c1 : memref<42x32x15x13x27xf32> -// CHECK: llvm.mlir.constant(15 : index) : i64 -// BAREPTR: llvm.mlir.constant(15 : index) : i64 - %c2 = constant 2 : index - %2 = memref.dim %static, %c2 : memref<42x32x15x13x27xf32> -// CHECK: llvm.mlir.constant(13 : index) : i64 -// BAREPTR: llvm.mlir.constant(13 : index) : i64 - %c3 = constant 3 : index - %3 = memref.dim %static, %c3 : memref<42x32x15x13x27xf32> -// CHECK: llvm.mlir.constant(27 : index) : i64 -// BAREPTR: llvm.mlir.constant(27 : index) : i64 - %c4 = constant 4 : index - %4 = memref.dim %static, %c4 : memref<42x32x15x13x27xf32> - return -} - -// ----- - -// BAREPTR: llvm.func @foo(!llvm.ptr) -> !llvm.ptr -func private @foo(memref<10xi8>) -> memref<20xi8> - -// BAREPTR-LABEL: func @check_memref_func_call -// BAREPTR-SAME: %[[in:.*]]: !llvm.ptr) -> !llvm.ptr -func @check_memref_func_call(%in : memref<10xi8>) -> memref<20xi8> { - // BAREPTR: %[[inDesc:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] - // BAREPTR-NEXT: %[[barePtr:.*]] = llvm.extractvalue %[[inDesc]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[call:.*]] = llvm.call @foo(%[[barePtr]]) : (!llvm.ptr) -> !llvm.ptr - // BAREPTR-NEXT: %[[desc0:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[desc1:.*]] = llvm.insertvalue %[[call]], %[[desc0]][0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[desc2:.*]] = llvm.insertvalue %[[call]], %[[desc1]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 - // BAREPTR-NEXT: %[[desc4:.*]] = llvm.insertvalue %[[c0]], %[[desc2]][2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[c20:.*]] = llvm.mlir.constant(20 : index) : i64 - // BAREPTR-NEXT: %[[desc6:.*]] = llvm.insertvalue %[[c20]], %[[desc4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: %[[c1:.*]] = llvm.mlir.constant(1 : index) : i64 - // BAREPTR-NEXT: %[[outDesc:.*]] = llvm.insertvalue %[[c1]], %[[desc6]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - %res = call @foo(%in) : (memref<10xi8>) -> (memref<20xi8>) - // BAREPTR-NEXT: %[[res:.*]] = llvm.extractvalue %[[outDesc]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // BAREPTR-NEXT: llvm.return %[[res]] : !llvm.ptr - return %res : memref<20xi8> -} - -// ----- - -// BAREPTR: llvm.func @goo(f32) -> f32 -func private @goo(f32) -> f32 - -// BAREPTR-LABEL: func @check_scalar_func_call -// BAREPTR-SAME: %[[in:.*]]: f32) -func @check_scalar_func_call(%in : f32) { - // BAREPTR-NEXT: %[[call:.*]] = llvm.call @goo(%[[in]]) : (f32) -> f32 - %res = call @goo(%in) : (f32) -> (f32) - return -} - -// ----- - -// Unranked memrefs are currently not supported in the bare-ptr calling -// convention. Check that the conversion to the LLVM-IR dialect doesn't happen -// in the presence of unranked memrefs when using such a calling convention. - -// BAREPTR: func private @hoo(memref<*xi8>) -> memref<*xi8> -func private @hoo(memref<*xi8>) -> memref<*xi8> - -// BAREPTR-LABEL: func @check_unranked_memref_func_call(%{{.*}}: memref<*xi8>) -> memref<*xi8> -func @check_unranked_memref_func_call(%in: memref<*xi8>) -> memref<*xi8> { - // BAREPTR-NEXT: call @hoo(%{{.*}}) : (memref<*xi8>) -> memref<*xi8> - %res = call @hoo(%in) : (memref<*xi8>) -> memref<*xi8> - // BAREPTR-NEXT: return %{{.*}} : memref<*xi8> - return %res : memref<*xi8> -} - -// ----- - -// Check that consistent types are emitted in address arithemic in presence of -// a data layout specification. -module attributes { dlti.dl_spec = #dlti.dl_spec<#dlti.dl_entry> } { - func @address() { - %c1 = constant 1 : index - %0 = memref.alloc(%c1) : memref> - // CHECK: %[[CST:.*]] = llvm.mlir.constant(1 : index) : i32 - // CHECK: llvm.mlir.null - // CHECK: llvm.getelementptr %{{.*}}[[CST]] - // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 - // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 - // CHECK: llvm.add %{{.*}} : i32 - // CHECK: llvm.call @malloc(%{{.*}}) : (i32) -> !llvm.ptr - // CHECK: llvm.ptrtoint %{{.*}} : !llvm.ptr<{{.*}}> to i32 - // CHECK: llvm.sub {{.*}} : i32 - // CHECK: llvm.add {{.*}} : i32 - // CHECK: llvm.urem {{.*}} : i32 - // CHECK: llvm.sub {{.*}} : i32 - // CHECK: llvm.inttoptr %{{.*}} : i32 to !llvm.ptr - return - } -} - -// ----- - -// Should not convert memrefs with unsupported types in any convention. - -// CHECK: @unsupported_memref_element_type -// CHECK-SAME: memref< -// CHECK-NOT: !llvm.struct -// BAREPTR: @unsupported_memref_element_type -// BAREPTR-SAME: memref< -// BAREPTR-NOT: !llvm.ptr -func private @unsupported_memref_element_type() -> memref<42 x !test.memref_element> - -// CHECK: @unsupported_unranked_memref_element_type -// CHECK-SAME: memref< -// CHECK-NOT: !llvm.struct -// BAREPTR: @unsupported_unranked_memref_element_type -// BAREPTR-SAME: memref< -// BAREPTR-NOT: !llvm.ptr -func private @unsupported_unranked_memref_element_type() -> memref<* x !test.memref_element> - diff --git a/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir b/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir deleted file mode 100644 --- a/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir +++ /dev/null @@ -1,1455 +0,0 @@ -// RUN: mlir-opt -convert-std-to-llvm %s -split-input-file | FileCheck %s -// RUN: mlir-opt -convert-std-to-llvm='index-bitwidth=32' %s -split-input-file | FileCheck --check-prefix=CHECK32 %s - -// CHECK-LABEL: func @empty() { -// CHECK-NEXT: llvm.return -// CHECK-NEXT: } -func @empty() { -^bb0: - return -} - -// CHECK-LABEL: llvm.func @body(i64) -func private @body(index) - -// CHECK-LABEL: func @simple_loop() { -// CHECK32-LABEL: func @simple_loop() { -func @simple_loop() { -^bb0: -// CHECK-NEXT: llvm.br ^bb1 -// CHECK32-NEXT: llvm.br ^bb1 - br ^bb1 - -// CHECK-NEXT: ^bb1: // pred: ^bb0 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK32-NEXT: ^bb1: // pred: ^bb0 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i32 -// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) -^bb1: // pred: ^bb0 - %c1 = constant 1 : index - %c42 = constant 42 : index - br ^bb2(%c1 : index) - -// CHECK: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb3 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 -// CHECK32: ^bb2({{.*}}: i32): // 2 preds: ^bb1, ^bb3 -// CHECK32-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i32 -// CHECK32-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 -^bb2(%0: index): // 2 preds: ^bb1, ^bb3 - %1 = cmpi slt, %0, %c42 : index - cond_br %1, ^bb3, ^bb4 - -// CHECK: ^bb3: // pred: ^bb2 -// CHECK-NEXT: llvm.call @body({{.*}}) : (i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK32: ^bb3: // pred: ^bb2 -// CHECK32-NEXT: llvm.call @body({{.*}}) : (i32) -> () -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 -// CHECK32-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i32 -// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) -^bb3: // pred: ^bb2 - call @body(%0) : (index) -> () - %c1_0 = constant 1 : index - %2 = addi %0, %c1_0 : index - br ^bb2(%2 : index) - -// CHECK: ^bb4: // pred: ^bb2 -// CHECK-NEXT: llvm.return -^bb4: // pred: ^bb2 - return -} - -// CHECK-LABEL: func @simple_caller() { -// CHECK-NEXT: llvm.call @simple_loop() : () -> () -// CHECK-NEXT: llvm.return -// CHECK-NEXT: } -func @simple_caller() { -^bb0: - call @simple_loop() : () -> () - return -} - -// Check that function call attributes persist during conversion. -// CHECK-LABEL: @call_with_attributes -func @call_with_attributes() { - // CHECK: llvm.call @simple_loop() {baz = [1, 2, 3, 4], foo = "bar"} : () -> () - call @simple_loop() {foo="bar", baz=[1,2,3,4]} : () -> () - return -} - -// CHECK-LABEL: func @ml_caller() { -// CHECK-NEXT: llvm.call @simple_loop() : () -> () -// CHECK-NEXT: llvm.call @more_imperfectly_nested_loops() : () -> () -// CHECK-NEXT: llvm.return -// CHECK-NEXT: } -func @ml_caller() { -^bb0: - call @simple_loop() : () -> () - call @more_imperfectly_nested_loops() : () -> () - return -} - -// CHECK-LABEL: llvm.func @body_args(i64) -> i64 -// CHECK32-LABEL: llvm.func @body_args(i32) -> i32 -func private @body_args(index) -> index -// CHECK-LABEL: llvm.func @other(i64, i32) -> i32 -// CHECK32-LABEL: llvm.func @other(i32, i32) -> i32 -func private @other(index, i32) -> i32 - -// CHECK-LABEL: func @func_args(%arg0: i32, %arg1: i32) -> i32 { -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : i32) : i32 -// CHECK-NEXT: llvm.br ^bb1 -// CHECK32-LABEL: func @func_args(%arg0: i32, %arg1: i32) -> i32 { -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : i32) : i32 -// CHECK32-NEXT: llvm.br ^bb1 -func @func_args(i32, i32) -> i32 { -^bb0(%arg0: i32, %arg1: i32): - %c0_i32 = constant 0 : i32 - br ^bb1 - -// CHECK-NEXT: ^bb1: // pred: ^bb0 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK32-NEXT: ^bb1: // pred: ^bb0 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i32 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i32 -// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) -^bb1: // pred: ^bb0 - %c0 = constant 0 : index - %c42 = constant 42 : index - br ^bb2(%c0 : index) - -// CHECK-NEXT: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb3 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 -// CHECK32-NEXT: ^bb2({{.*}}: i32): // 2 preds: ^bb1, ^bb3 -// CHECK32-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i32 -// CHECK32-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 -^bb2(%0: index): // 2 preds: ^bb1, ^bb3 - %1 = cmpi slt, %0, %c42 : index - cond_br %1, ^bb3, ^bb4 - -// CHECK-NEXT: ^bb3: // pred: ^bb2 -// CHECK-NEXT: {{.*}} = llvm.call @body_args({{.*}}) : (i64) -> i64 -// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg0) : (i64, i32) -> i32 -// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i64, i32) -> i32 -// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg1) : (i64, i32) -> i32 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK32-NEXT: ^bb3: // pred: ^bb2 -// CHECK32-NEXT: {{.*}} = llvm.call @body_args({{.*}}) : (i32) -> i32 -// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg0) : (i32, i32) -> i32 -// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i32, i32) -> i32 -// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg1) : (i32, i32) -> i32 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 -// CHECK32-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i32 -// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) -^bb3: // pred: ^bb2 - %2 = call @body_args(%0) : (index) -> index - %3 = call @other(%2, %arg0) : (index, i32) -> i32 - %4 = call @other(%2, %3) : (index, i32) -> i32 - %5 = call @other(%2, %arg1) : (index, i32) -> i32 - %c1 = constant 1 : index - %6 = addi %0, %c1 : index - br ^bb2(%6 : index) - -// CHECK-NEXT: ^bb4: // pred: ^bb2 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i64, i32) -> i32 -// CHECK-NEXT: llvm.return {{.*}} : i32 -// CHECK32-NEXT: ^bb4: // pred: ^bb2 -// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i32 -// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i32, i32) -> i32 -// CHECK32-NEXT: llvm.return {{.*}} : i32 -^bb4: // pred: ^bb2 - %c0_0 = constant 0 : index - %7 = call @other(%c0_0, %c0_i32) : (index, i32) -> i32 - return %7 : i32 -} - -// CHECK-LABEL: llvm.func @pre(i64) -// CHECK32-LABEL: llvm.func @pre(i32) -func private @pre(index) - -// CHECK-LABEL: llvm.func @body2(i64, i64) -// CHECK32-LABEL: llvm.func @body2(i32, i32) -func private @body2(index, index) - -// CHECK-LABEL: llvm.func @post(i64) -// CHECK32-LABEL: llvm.func @post(i32) -func private @post(index) - -// CHECK-LABEL: func @imperfectly_nested_loops() { -// CHECK-NEXT: llvm.br ^bb1 -func @imperfectly_nested_loops() { -^bb0: - br ^bb1 - -// CHECK-NEXT: ^bb1: // pred: ^bb0 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -^bb1: // pred: ^bb0 - %c0 = constant 0 : index - %c42 = constant 42 : index - br ^bb2(%c0 : index) - -// CHECK-NEXT: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb7 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb8 -^bb2(%0: index): // 2 preds: ^bb1, ^bb7 - %1 = cmpi slt, %0, %c42 : index - cond_br %1, ^bb3, ^bb8 - -// CHECK-NEXT: ^bb3: -// CHECK-NEXT: llvm.call @pre({{.*}}) : (i64) -> () -// CHECK-NEXT: llvm.br ^bb4 -^bb3: // pred: ^bb2 - call @pre(%0) : (index) -> () - br ^bb4 - -// CHECK-NEXT: ^bb4: // pred: ^bb3 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(7 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(56 : index) : i64 -// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) -^bb4: // pred: ^bb3 - %c7 = constant 7 : index - %c56 = constant 56 : index - br ^bb5(%c7 : index) - -// CHECK-NEXT: ^bb5({{.*}}: i64): // 2 preds: ^bb4, ^bb6 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb6, ^bb7 -^bb5(%2: index): // 2 preds: ^bb4, ^bb6 - %3 = cmpi slt, %2, %c56 : index - cond_br %3, ^bb6, ^bb7 - -// CHECK-NEXT: ^bb6: // pred: ^bb5 -// CHECK-NEXT: llvm.call @body2({{.*}}, {{.*}}) : (i64, i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(2 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) -^bb6: // pred: ^bb5 - call @body2(%0, %2) : (index, index) -> () - %c2 = constant 2 : index - %4 = addi %2, %c2 : index - br ^bb5(%4 : index) - -// CHECK-NEXT: ^bb7: // pred: ^bb5 -// CHECK-NEXT: llvm.call @post({{.*}}) : (i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -^bb7: // pred: ^bb5 - call @post(%0) : (index) -> () - %c1 = constant 1 : index - %5 = addi %0, %c1 : index - br ^bb2(%5 : index) - -// CHECK-NEXT: ^bb8: // pred: ^bb2 -// CHECK-NEXT: llvm.return -^bb8: // pred: ^bb2 - return -} - -// CHECK-LABEL: llvm.func @mid(i64) -func private @mid(index) - -// CHECK-LABEL: llvm.func @body3(i64, i64) -func private @body3(index, index) - -// A complete function transformation check. -// CHECK-LABEL: func @more_imperfectly_nested_loops() { -// CHECK-NEXT: llvm.br ^bb1 -// CHECK-NEXT:^bb1: // pred: ^bb0 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK-NEXT:^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb11 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb12 -// CHECK-NEXT:^bb3: // pred: ^bb2 -// CHECK-NEXT: llvm.call @pre({{.*}}) : (i64) -> () -// CHECK-NEXT: llvm.br ^bb4 -// CHECK-NEXT:^bb4: // pred: ^bb3 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(7 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(56 : index) : i64 -// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) -// CHECK-NEXT:^bb5({{.*}}: i64): // 2 preds: ^bb4, ^bb6 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb6, ^bb7 -// CHECK-NEXT:^bb6: // pred: ^bb5 -// CHECK-NEXT: llvm.call @body2({{.*}}, {{.*}}) : (i64, i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(2 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) -// CHECK-NEXT:^bb7: // pred: ^bb5 -// CHECK-NEXT: llvm.call @mid({{.*}}) : (i64) -> () -// CHECK-NEXT: llvm.br ^bb8 -// CHECK-NEXT:^bb8: // pred: ^bb7 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(18 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(37 : index) : i64 -// CHECK-NEXT: llvm.br ^bb9({{.*}} : i64) -// CHECK-NEXT:^bb9({{.*}}: i64): // 2 preds: ^bb8, ^bb10 -// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb10, ^bb11 -// CHECK-NEXT:^bb10: // pred: ^bb9 -// CHECK-NEXT: llvm.call @body3({{.*}}, {{.*}}) : (i64, i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(3 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb9({{.*}} : i64) -// CHECK-NEXT:^bb11: // pred: ^bb9 -// CHECK-NEXT: llvm.call @post({{.*}}) : (i64) -> () -// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 -// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) -// CHECK-NEXT:^bb12: // pred: ^bb2 -// CHECK-NEXT: llvm.return -// CHECK-NEXT: } -func @more_imperfectly_nested_loops() { -^bb0: - br ^bb1 -^bb1: // pred: ^bb0 - %c0 = constant 0 : index - %c42 = constant 42 : index - br ^bb2(%c0 : index) -^bb2(%0: index): // 2 preds: ^bb1, ^bb11 - %1 = cmpi slt, %0, %c42 : index - cond_br %1, ^bb3, ^bb12 -^bb3: // pred: ^bb2 - call @pre(%0) : (index) -> () - br ^bb4 -^bb4: // pred: ^bb3 - %c7 = constant 7 : index - %c56 = constant 56 : index - br ^bb5(%c7 : index) -^bb5(%2: index): // 2 preds: ^bb4, ^bb6 - %3 = cmpi slt, %2, %c56 : index - cond_br %3, ^bb6, ^bb7 -^bb6: // pred: ^bb5 - call @body2(%0, %2) : (index, index) -> () - %c2 = constant 2 : index - %4 = addi %2, %c2 : index - br ^bb5(%4 : index) -^bb7: // pred: ^bb5 - call @mid(%0) : (index) -> () - br ^bb8 -^bb8: // pred: ^bb7 - %c18 = constant 18 : index - %c37 = constant 37 : index - br ^bb9(%c18 : index) -^bb9(%5: index): // 2 preds: ^bb8, ^bb10 - %6 = cmpi slt, %5, %c37 : index - cond_br %6, ^bb10, ^bb11 -^bb10: // pred: ^bb9 - call @body3(%0, %5) : (index, index) -> () - %c3 = constant 3 : index - %7 = addi %5, %c3 : index - br ^bb9(%7 : index) -^bb11: // pred: ^bb9 - call @post(%0) : (index) -> () - %c1 = constant 1 : index - %8 = addi %0, %c1 : index - br ^bb2(%8 : index) -^bb12: // pred: ^bb2 - return -} - -// CHECK-LABEL: llvm.func @get_i64() -> i64 -func private @get_i64() -> (i64) -// CHECK-LABEL: llvm.func @get_f32() -> f32 -func private @get_f32() -> (f32) -// CHECK-LABEL: llvm.func @get_c16() -> !llvm.struct<(f16, f16)> -func private @get_c16() -> (complex) -// CHECK-LABEL: llvm.func @get_c32() -> !llvm.struct<(f32, f32)> -func private @get_c32() -> (complex) -// CHECK-LABEL: llvm.func @get_c64() -> !llvm.struct<(f64, f64)> -func private @get_c64() -> (complex) -// CHECK-LABEL: llvm.func @get_memref() -> !llvm.struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)> -// CHECK32-LABEL: llvm.func @get_memref() -> !llvm.struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)> -func private @get_memref() -> (memref<42x?x10x?xf32>) - -// CHECK-LABEL: llvm.func @multireturn() -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> { -// CHECK32-LABEL: llvm.func @multireturn() -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> { -func @multireturn() -> (i64, f32, memref<42x?x10x?xf32>) { -^bb0: -// CHECK-NEXT: {{.*}} = llvm.call @get_i64() : () -> i64 -// CHECK-NEXT: {{.*}} = llvm.call @get_f32() : () -> f32 -// CHECK-NEXT: {{.*}} = llvm.call @get_memref() : () -> !llvm.struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)> -// CHECK32-NEXT: {{.*}} = llvm.call @get_i64() : () -> i64 -// CHECK32-NEXT: {{.*}} = llvm.call @get_f32() : () -> f32 -// CHECK32-NEXT: {{.*}} = llvm.call @get_memref() : () -> !llvm.struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)> - %0 = call @get_i64() : () -> (i64) - %1 = call @get_f32() : () -> (f32) - %2 = call @get_memref() : () -> (memref<42x?x10x?xf32>) -// CHECK-NEXT: {{.*}} = llvm.mlir.undef : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: llvm.return {{.*}} : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK32-NEXT: {{.*}} = llvm.mlir.undef : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: llvm.return {{.*}} : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> - return %0, %1, %2 : i64, f32, memref<42x?x10x?xf32> -} - - -// CHECK-LABEL: llvm.func @multireturn_caller() { -// CHECK32-LABEL: llvm.func @multireturn_caller() { -func @multireturn_caller() { -^bb0: -// CHECK-NEXT: {{.*}} = llvm.call @multireturn() : () -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> -// CHECK32-NEXT: {{.*}} = llvm.call @multireturn() : () -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> -// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> - %0:3 = call @multireturn() : () -> (i64, f32, memref<42x?x10x?xf32>) - %1 = constant 42 : i64 -// CHECK: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 - %2 = addi %0#0, %1 : i64 - %3 = constant 42.0 : f32 -// CHECK: {{.*}} = llvm.fadd {{.*}}, {{.*}} : f32 - %4 = addf %0#1, %3 : f32 - %5 = constant 0 : index - return -} - -// CHECK-LABEL: llvm.func @vector_ops(%arg0: vector<4xf32>, %arg1: vector<4xi1>, %arg2: vector<4xi64>, %arg3: vector<4xi64>) -> vector<4xf32> { -func @vector_ops(%arg0: vector<4xf32>, %arg1: vector<4xi1>, %arg2: vector<4xi64>, %arg3: vector<4xi64>) -> vector<4xf32> { -// CHECK-NEXT: %0 = llvm.mlir.constant(dense<4.200000e+01> : vector<4xf32>) : vector<4xf32> - %0 = constant dense<42.> : vector<4xf32> -// CHECK-NEXT: %1 = llvm.fadd %arg0, %0 : vector<4xf32> - %1 = addf %arg0, %0 : vector<4xf32> -// CHECK-NEXT: %2 = llvm.sdiv %arg2, %arg2 : vector<4xi64> - %3 = divi_signed %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %3 = llvm.udiv %arg2, %arg2 : vector<4xi64> - %4 = divi_unsigned %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %4 = llvm.srem %arg2, %arg2 : vector<4xi64> - %5 = remi_signed %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %5 = llvm.urem %arg2, %arg2 : vector<4xi64> - %6 = remi_unsigned %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %6 = llvm.fdiv %arg0, %0 : vector<4xf32> - %7 = divf %arg0, %0 : vector<4xf32> -// CHECK-NEXT: %7 = llvm.frem %arg0, %0 : vector<4xf32> - %8 = remf %arg0, %0 : vector<4xf32> -// CHECK-NEXT: %8 = llvm.and %arg2, %arg3 : vector<4xi64> - %9 = and %arg2, %arg3 : vector<4xi64> -// CHECK-NEXT: %9 = llvm.or %arg2, %arg3 : vector<4xi64> - %10 = or %arg2, %arg3 : vector<4xi64> -// CHECK-NEXT: %10 = llvm.xor %arg2, %arg3 : vector<4xi64> - %11 = xor %arg2, %arg3 : vector<4xi64> -// CHECK-NEXT: %11 = llvm.shl %arg2, %arg2 : vector<4xi64> - %12 = shift_left %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %12 = llvm.ashr %arg2, %arg2 : vector<4xi64> - %13 = shift_right_signed %arg2, %arg2 : vector<4xi64> -// CHECK-NEXT: %13 = llvm.lshr %arg2, %arg2 : vector<4xi64> - %14 = shift_right_unsigned %arg2, %arg2 : vector<4xi64> - return %1 : vector<4xf32> -} - -// CHECK-LABEL: @ops -func @ops(f32, f32, i32, i32, f64) -> (f32, i32) { -^bb0(%arg0: f32, %arg1: f32, %arg2: i32, %arg3: i32, %arg4: f64): -// CHECK-NEXT: %0 = llvm.fsub %arg0, %arg1 : f32 - %0 = subf %arg0, %arg1: f32 -// CHECK-NEXT: %1 = llvm.sub %arg2, %arg3 : i32 - %1 = subi %arg2, %arg3: i32 -// CHECK-NEXT: %2 = llvm.icmp "slt" %arg2, %1 : i32 - %2 = cmpi slt, %arg2, %1 : i32 -// CHECK-NEXT: %3 = llvm.sdiv %arg2, %arg3 : i32 - %3 = divi_signed %arg2, %arg3 : i32 -// CHECK-NEXT: %4 = llvm.udiv %arg2, %arg3 : i32 - %4 = divi_unsigned %arg2, %arg3 : i32 -// CHECK-NEXT: %5 = llvm.srem %arg2, %arg3 : i32 - %5 = remi_signed %arg2, %arg3 : i32 -// CHECK-NEXT: %6 = llvm.urem %arg2, %arg3 : i32 - %6 = remi_unsigned %arg2, %arg3 : i32 -// CHECK-NEXT: %7 = llvm.select %2, %arg2, %arg3 : i1, i32 - %7 = select %2, %arg2, %arg3 : i32 -// CHECK-NEXT: %8 = llvm.fdiv %arg0, %arg1 : f32 - %8 = divf %arg0, %arg1 : f32 -// CHECK-NEXT: %9 = llvm.frem %arg0, %arg1 : f32 - %9 = remf %arg0, %arg1 : f32 -// CHECK-NEXT: %10 = llvm.and %arg2, %arg3 : i32 - %10 = and %arg2, %arg3 : i32 -// CHECK-NEXT: %11 = llvm.or %arg2, %arg3 : i32 - %11 = or %arg2, %arg3 : i32 -// CHECK-NEXT: %12 = llvm.xor %arg2, %arg3 : i32 - %12 = xor %arg2, %arg3 : i32 -// CHECK-NEXT: %13 = "llvm.intr.exp"(%arg0) : (f32) -> f32 - %13 = math.exp %arg0 : f32 -// CHECK-NEXT: %14 = "llvm.intr.exp2"(%arg0) : (f32) -> f32 - %14 = math.exp2 %arg0 : f32 -// CHECK-NEXT: %15 = llvm.mlir.constant(7.900000e-01 : f64) : f64 - %15 = constant 7.9e-01 : f64 -// CHECK-NEXT: %16 = llvm.shl %arg2, %arg3 : i32 - %16 = shift_left %arg2, %arg3 : i32 -// CHECK-NEXT: %17 = llvm.ashr %arg2, %arg3 : i32 - %17 = shift_right_signed %arg2, %arg3 : i32 -// CHECK-NEXT: %18 = llvm.lshr %arg2, %arg3 : i32 - %18 = shift_right_unsigned %arg2, %arg3 : i32 -// CHECK-NEXT: %{{[0-9]+}} = "llvm.intr.sqrt"(%arg0) : (f32) -> f32 - %19 = math.sqrt %arg0 : f32 -// CHECK-NEXT: %{{[0-9]+}} = "llvm.intr.sqrt"(%arg4) : (f64) -> f64 - %20 = math.sqrt %arg4 : f64 - return %0, %4 : f32, i32 -} - -// Checking conversion of index types to integers using i1, assuming no target -// system would have a 1-bit address space. Otherwise, we would have had to -// make this test dependent on the pointer size on the target system. -// CHECK-LABEL: @index_cast -func @index_cast(%arg0: index, %arg1: i1) { -// CHECK-NEXT: = llvm.trunc %arg0 : i{{.*}} to i1 - %0 = index_cast %arg0: index to i1 -// CHECK-NEXT: = llvm.sext %arg1 : i1 to i{{.*}} - %1 = index_cast %arg1: i1 to index - return -} - -// CHECK-LABEL: @vector_index_cast -func @vector_index_cast(%arg0: vector<2xindex>, %arg1: vector<2xi1>) { -// CHECK-NEXT: = llvm.trunc %{{.*}} : vector<2xi{{.*}}> to vector<2xi1> - %0 = index_cast %arg0: vector<2xindex> to vector<2xi1> -// CHECK-NEXT: = llvm.sext %{{.*}} : vector<2xi1> to vector<2xi{{.*}}> - %1 = index_cast %arg1: vector<2xi1> to vector<2xindex> - return -} - -// Checking conversion of signed integer types to floating point. -// CHECK-LABEL: @sitofp -func @sitofp(%arg0 : i32, %arg1 : i64) { -// CHECK-NEXT: = llvm.sitofp {{.*}} : i32 to f32 - %0 = sitofp %arg0: i32 to f32 -// CHECK-NEXT: = llvm.sitofp {{.*}} : i32 to f64 - %1 = sitofp %arg0: i32 to f64 -// CHECK-NEXT: = llvm.sitofp {{.*}} : i64 to f32 - %2 = sitofp %arg1: i64 to f32 -// CHECK-NEXT: = llvm.sitofp {{.*}} : i64 to f64 - %3 = sitofp %arg1: i64 to f64 - return -} - -// Checking conversion of integer vectors to floating point vector types. -// CHECK-LABEL: @sitofp_vector -func @sitofp_vector(%arg0 : vector<2xi16>, %arg1 : vector<2xi32>, %arg2 : vector<2xi64>) { -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi16> to vector<2xf32> - %0 = sitofp %arg0: vector<2xi16> to vector<2xf32> -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi16> to vector<2xf64> - %1 = sitofp %arg0: vector<2xi16> to vector<2xf64> -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi32> to vector<2xf32> - %2 = sitofp %arg1: vector<2xi32> to vector<2xf32> -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi32> to vector<2xf64> - %3 = sitofp %arg1: vector<2xi32> to vector<2xf64> -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi64> to vector<2xf32> - %4 = sitofp %arg2: vector<2xi64> to vector<2xf32> -// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi64> to vector<2xf64> - %5 = sitofp %arg2: vector<2xi64> to vector<2xf64> - return -} - -// Checking conversion of unsigned integer types to floating point. -// CHECK-LABEL: @uitofp -func @uitofp(%arg0 : i32, %arg1 : i64) { -// CHECK-NEXT: = llvm.uitofp {{.*}} : i32 to f32 - %0 = uitofp %arg0: i32 to f32 -// CHECK-NEXT: = llvm.uitofp {{.*}} : i32 to f64 - %1 = uitofp %arg0: i32 to f64 -// CHECK-NEXT: = llvm.uitofp {{.*}} : i64 to f32 - %2 = uitofp %arg1: i64 to f32 -// CHECK-NEXT: = llvm.uitofp {{.*}} : i64 to f64 - %3 = uitofp %arg1: i64 to f64 - return -} - -// Checking conversion of integer types to floating point. -// CHECK-LABEL: @fpext -func @fpext(%arg0 : f16, %arg1 : f32) { -// CHECK-NEXT: = llvm.fpext {{.*}} : f16 to f32 - %0 = fpext %arg0: f16 to f32 -// CHECK-NEXT: = llvm.fpext {{.*}} : f16 to f64 - %1 = fpext %arg0: f16 to f64 -// CHECK-NEXT: = llvm.fpext {{.*}} : f32 to f64 - %2 = fpext %arg1: f32 to f64 - return -} - -// Checking conversion of integer types to floating point. -// CHECK-LABEL: @fpext -func @fpext_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>) { -// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf16> to vector<2xf32> - %0 = fpext %arg0: vector<2xf16> to vector<2xf32> -// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf16> to vector<2xf64> - %1 = fpext %arg0: vector<2xf16> to vector<2xf64> -// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf32> to vector<2xf64> - %2 = fpext %arg1: vector<2xf32> to vector<2xf64> - return -} - -// Checking conversion of floating point to integer types. -// CHECK-LABEL: @fptosi -func @fptosi(%arg0 : f32, %arg1 : f64) { -// CHECK-NEXT: = llvm.fptosi {{.*}} : f32 to i32 - %0 = fptosi %arg0: f32 to i32 -// CHECK-NEXT: = llvm.fptosi {{.*}} : f32 to i64 - %1 = fptosi %arg0: f32 to i64 -// CHECK-NEXT: = llvm.fptosi {{.*}} : f64 to i32 - %2 = fptosi %arg1: f64 to i32 -// CHECK-NEXT: = llvm.fptosi {{.*}} : f64 to i64 - %3 = fptosi %arg1: f64 to i64 - return -} - -// Checking conversion of floating point vectors to integer vector types. -// CHECK-LABEL: @fptosi_vector -func @fptosi_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>, %arg2 : vector<2xf64>) { -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf16> to vector<2xi32> - %0 = fptosi %arg0: vector<2xf16> to vector<2xi32> -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf16> to vector<2xi64> - %1 = fptosi %arg0: vector<2xf16> to vector<2xi64> -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf32> to vector<2xi32> - %2 = fptosi %arg1: vector<2xf32> to vector<2xi32> -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf32> to vector<2xi64> - %3 = fptosi %arg1: vector<2xf32> to vector<2xi64> -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf64> to vector<2xi32> - %4 = fptosi %arg2: vector<2xf64> to vector<2xi32> -// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf64> to vector<2xi64> - %5 = fptosi %arg2: vector<2xf64> to vector<2xi64> - return -} - -// Checking conversion of floating point to integer types. -// CHECK-LABEL: @fptoui -func @fptoui(%arg0 : f32, %arg1 : f64) { -// CHECK-NEXT: = llvm.fptoui {{.*}} : f32 to i32 - %0 = fptoui %arg0: f32 to i32 -// CHECK-NEXT: = llvm.fptoui {{.*}} : f32 to i64 - %1 = fptoui %arg0: f32 to i64 -// CHECK-NEXT: = llvm.fptoui {{.*}} : f64 to i32 - %2 = fptoui %arg1: f64 to i32 -// CHECK-NEXT: = llvm.fptoui {{.*}} : f64 to i64 - %3 = fptoui %arg1: f64 to i64 - return -} - -// Checking conversion of floating point vectors to integer vector types. -// CHECK-LABEL: @fptoui_vector -func @fptoui_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>, %arg2 : vector<2xf64>) { -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf16> to vector<2xi32> - %0 = fptoui %arg0: vector<2xf16> to vector<2xi32> -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf16> to vector<2xi64> - %1 = fptoui %arg0: vector<2xf16> to vector<2xi64> -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf32> to vector<2xi32> - %2 = fptoui %arg1: vector<2xf32> to vector<2xi32> -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf32> to vector<2xi64> - %3 = fptoui %arg1: vector<2xf32> to vector<2xi64> -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf64> to vector<2xi32> - %4 = fptoui %arg2: vector<2xf64> to vector<2xi32> -// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf64> to vector<2xi64> - %5 = fptoui %arg2: vector<2xf64> to vector<2xi64> - return -} - -// Checking conversion of integer vectors to floating point vector types. -// CHECK-LABEL: @uitofp_vector -func @uitofp_vector(%arg0 : vector<2xi16>, %arg1 : vector<2xi32>, %arg2 : vector<2xi64>) { -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi16> to vector<2xf32> - %0 = uitofp %arg0: vector<2xi16> to vector<2xf32> -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi16> to vector<2xf64> - %1 = uitofp %arg0: vector<2xi16> to vector<2xf64> -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi32> to vector<2xf32> - %2 = uitofp %arg1: vector<2xi32> to vector<2xf32> -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi32> to vector<2xf64> - %3 = uitofp %arg1: vector<2xi32> to vector<2xf64> -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi64> to vector<2xf32> - %4 = uitofp %arg2: vector<2xi64> to vector<2xf32> -// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi64> to vector<2xf64> - %5 = uitofp %arg2: vector<2xi64> to vector<2xf64> - return -} - -// Checking conversion of integer types to floating point. -// CHECK-LABEL: @fptrunc -func @fptrunc(%arg0 : f32, %arg1 : f64) { -// CHECK-NEXT: = llvm.fptrunc {{.*}} : f32 to f16 - %0 = fptrunc %arg0: f32 to f16 -// CHECK-NEXT: = llvm.fptrunc {{.*}} : f64 to f16 - %1 = fptrunc %arg1: f64 to f16 -// CHECK-NEXT: = llvm.fptrunc {{.*}} : f64 to f32 - %2 = fptrunc %arg1: f64 to f32 - return -} - -// Checking conversion of integer types to floating point. -// CHECK-LABEL: @fptrunc -func @fptrunc_vector(%arg0 : vector<2xf32>, %arg1 : vector<2xf64>) { -// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf32> to vector<2xf16> - %0 = fptrunc %arg0: vector<2xf32> to vector<2xf16> -// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf64> to vector<2xf16> - %1 = fptrunc %arg1: vector<2xf64> to vector<2xf16> -// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf64> to vector<2xf32> - %2 = fptrunc %arg1: vector<2xf64> to vector<2xf32> - return -} - -// Check sign and zero extension and truncation of integers. -// CHECK-LABEL: @integer_extension_and_truncation -func @integer_extension_and_truncation(%arg0 : i3) { -// CHECK-NEXT: = llvm.sext %arg0 : i3 to i6 - %0 = sexti %arg0 : i3 to i6 -// CHECK-NEXT: = llvm.zext %arg0 : i3 to i6 - %1 = zexti %arg0 : i3 to i6 -// CHECK-NEXT: = llvm.trunc %arg0 : i3 to i2 - %2 = trunci %arg0 : i3 to i2 - return -} - -// CHECK-LABEL: @dfs_block_order -func @dfs_block_order(%arg0: i32) -> (i32) { -// CHECK-NEXT: %[[CST:.*]] = llvm.mlir.constant(42 : i32) : i32 - %0 = constant 42 : i32 -// CHECK-NEXT: llvm.br ^bb2 - br ^bb2 - -// CHECK-NEXT: ^bb1: -// CHECK-NEXT: %[[ADD:.*]] = llvm.add %arg0, %[[CST]] : i32 -// CHECK-NEXT: llvm.return %[[ADD]] : i32 -^bb1: - %2 = addi %arg0, %0 : i32 - return %2 : i32 - -// CHECK-NEXT: ^bb2: -^bb2: -// CHECK-NEXT: llvm.br ^bb1 - br ^bb1 -} - -// CHECK-LABEL: func @fcmp(%arg0: f32, %arg1: f32) { -func @fcmp(f32, f32) -> () { -^bb0(%arg0: f32, %arg1: f32): - // CHECK: llvm.fcmp "oeq" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ogt" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "oge" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "olt" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ole" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "one" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ord" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ueq" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ugt" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "uge" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ult" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "ule" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "une" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.fcmp "uno" %arg0, %arg1 : f32 - // CHECK-NEXT: llvm.return - %1 = cmpf oeq, %arg0, %arg1 : f32 - %2 = cmpf ogt, %arg0, %arg1 : f32 - %3 = cmpf oge, %arg0, %arg1 : f32 - %4 = cmpf olt, %arg0, %arg1 : f32 - %5 = cmpf ole, %arg0, %arg1 : f32 - %6 = cmpf one, %arg0, %arg1 : f32 - %7 = cmpf ord, %arg0, %arg1 : f32 - %8 = cmpf ueq, %arg0, %arg1 : f32 - %9 = cmpf ugt, %arg0, %arg1 : f32 - %10 = cmpf uge, %arg0, %arg1 : f32 - %11 = cmpf ult, %arg0, %arg1 : f32 - %12 = cmpf ule, %arg0, %arg1 : f32 - %13 = cmpf une, %arg0, %arg1 : f32 - %14 = cmpf uno, %arg0, %arg1 : f32 - - return -} - -// CHECK-LABEL: @splat -// CHECK-SAME: %[[A:arg[0-9]+]]: vector<4xf32> -// CHECK-SAME: %[[ELT:arg[0-9]+]]: f32 -func @splat(%a: vector<4xf32>, %b: f32) -> vector<4xf32> { - %vb = splat %b : vector<4xf32> - %r = mulf %a, %vb : vector<4xf32> - return %r : vector<4xf32> -} -// CHECK-NEXT: %[[UNDEF:[0-9]+]] = llvm.mlir.undef : vector<4xf32> -// CHECK-NEXT: %[[ZERO:[0-9]+]] = llvm.mlir.constant(0 : i32) : i32 -// CHECK-NEXT: %[[V:[0-9]+]] = llvm.insertelement %[[ELT]], %[[UNDEF]][%[[ZERO]] : i32] : vector<4xf32> -// CHECK-NEXT: %[[SPLAT:[0-9]+]] = llvm.shufflevector %[[V]], %[[UNDEF]] [0 : i32, 0 : i32, 0 : i32, 0 : i32] -// CHECK-NEXT: %[[SCALE:[0-9]+]] = llvm.fmul %[[A]], %[[SPLAT]] : vector<4xf32> -// CHECK-NEXT: llvm.return %[[SCALE]] : vector<4xf32> - -// CHECK-LABEL: func @view( -// CHECK: %[[ARG0:.*]]: i64, %[[ARG1:.*]]: i64, %[[ARG2:.*]]: i64 -func @view(%arg0 : index, %arg1 : index, %arg2 : index) { - // CHECK: llvm.mlir.constant(2048 : index) : i64 - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - %0 = memref.alloc() : memref<2048xi8> - - // Test two dynamic sizes. - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BASE_PTR:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: %[[SHIFTED_BASE_PTR:.*]] = llvm.getelementptr %[[BASE_PTR]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr - // CHECK: %[[CAST_SHIFTED_BASE_PTR:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR]] : !llvm.ptr to !llvm.ptr - // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C0:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[C0]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[ARG1]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(1 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[ARG0]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mul %{{.*}}, %[[ARG1]] - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %1 = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref - - // Test one dynamic size. - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BASE_PTR_2:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: %[[SHIFTED_BASE_PTR_2:.*]] = llvm.getelementptr %[[BASE_PTR_2]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr - // CHECK: %[[CAST_SHIFTED_BASE_PTR_2:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_2]] : !llvm.ptr to !llvm.ptr - // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_2]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C0_2:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[C0_2]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[ARG1]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(1 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(4 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mul %{{.*}}, %[[ARG1]] - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %3 = memref.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32> - - // Test static sizes. - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BASE_PTR_3:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: %[[SHIFTED_BASE_PTR_3:.*]] = llvm.getelementptr %[[BASE_PTR_3]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr - // CHECK: %[[CAST_SHIFTED_BASE_PTR_3:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_3]] : !llvm.ptr to !llvm.ptr - // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_3]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C0_3:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[C0_3]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(4 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(1 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(64 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(4 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %5 = memref.view %0[%arg2][] : memref<2048xi8> to memref<64x4xf32> - - // Test view memory space. - // CHECK: llvm.mlir.constant(2048 : index) : i64 - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - %6 = memref.alloc() : memref<2048xi8, 4> - - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BASE_PTR_4:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: %[[SHIFTED_BASE_PTR_4:.*]] = llvm.getelementptr %[[BASE_PTR_4]][%[[ARG2]]] : (!llvm.ptr, i64) -> !llvm.ptr - // CHECK: %[[CAST_SHIFTED_BASE_PTR_4:.*]] = llvm.bitcast %[[SHIFTED_BASE_PTR_4]] : !llvm.ptr to !llvm.ptr - // CHECK: llvm.insertvalue %[[CAST_SHIFTED_BASE_PTR_4]], %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C0_4:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[C0_4]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(4 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(1 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(64 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.mlir.constant(4 : index) : i64 - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %7 = memref.view %6[%arg2][] : memref<2048xi8, 4> to memref<64x4xf32, 4> - - return -} - -// CHECK-LABEL: func @subview( -// CHECK-COUNT-2: !llvm.ptr, -// CHECK-COUNT-5: {{%[a-zA-Z0-9]*}}: i64, -// CHECK: %[[ARG0:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG1:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG2:.*]]: i64) -// CHECK32-LABEL: func @subview( -// CHECK32-COUNT-2: !llvm.ptr, -// CHECK32-COUNT-5: {{%[a-zA-Z0-9]*}}: i32, -// CHECK32: %[[ARG0:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG1:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG2:.*]]: i32) -func @subview(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i64 - // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 - // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i64 - // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 - // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i64 - // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i64 - // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 - // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 - // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i32 - // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i32 - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - - %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] : - memref<64x4xf32, offset: 0, strides: [4, 1]> - to memref - return -} - -// CHECK-LABEL: func @subview_non_zero_addrspace( -// CHECK-COUNT-2: !llvm.ptr, -// CHECK-COUNT-5: {{%[a-zA-Z0-9]*}}: i64, -// CHECK: %[[ARG0:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG1:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG2:.*]]: i64) -// CHECK32-LABEL: func @subview_non_zero_addrspace( -// CHECK32-COUNT-2: !llvm.ptr, -// CHECK32-COUNT-5: {{%[a-zA-Z0-9]*}}: i32, -// CHECK32: %[[ARG0:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG1:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG2:.*]]: i32) -func @subview_non_zero_addrspace(%0 : memref<64x4xf32, offset: 0, strides: [4, 1], 3>, %arg0 : index, %arg1 : index, %arg2 : index) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i64 - // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 - // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i64 - // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 - // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i64 - // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i64 - // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 - // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 - // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i32 - // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE1]] : i32 - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG1]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - - %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] : - memref<64x4xf32, offset: 0, strides: [4, 1], 3> - to memref - return -} - -// CHECK-LABEL: func @subview_const_size( -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG7:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG8:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG9:[a-zA-Z0-9]*]]: i64 -// CHECK32-LABEL: func @subview_const_size( -// CHECK32-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK32-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK32-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG7:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG8:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG9:[a-zA-Z0-9]*]]: i32 -func @subview_const_size(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i64 - // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 - // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i64 - // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 - // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) - // CHECK: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i64 - // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[CST2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) - // CHECK: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i64 - // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[CST4]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i32 - // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 - // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i32 - // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) - // CHECK32: %[[DESCSTRIDE1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i32 - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[CST2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[DESCSTRIDE1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) - // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i32 - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST4]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - %1 = memref.subview %0[%arg0, %arg1][4, 2][%arg0, %arg1] : - memref<64x4xf32, offset: 0, strides: [4, 1]> - to memref<4x2xf32, offset: ?, strides: [?, ?]> - return -} - -// CHECK-LABEL: func @subview_const_stride( -// CHECK-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG7:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG8:[a-zA-Z0-9]*]]: i64 -// CHECK-SAME: %[[ARG9:[a-zA-Z0-9]*]]: i64 -// CHECK32-LABEL: func @subview_const_stride( -// CHECK32-SAME: %[[ARG0:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK32-SAME: %[[ARG1:[a-zA-Z0-9]*]]: !llvm.ptr, -// CHECK32-SAME: %[[ARG2:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG3:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG4:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG5:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG6:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG7:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG8:[a-zA-Z0-9]*]]: i32 -// CHECK32-SAME: %[[ARG9:[a-zA-Z0-9]*]]: i32 -func @subview_const_stride(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFFINC:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i64 - // CHECK: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i64 - // CHECK: %[[OFFINC1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i64 - // CHECK: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i64 - // CHECK: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) - // CHECK: %[[DESC3:.*]] = llvm.insertvalue %[[ARG8]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[DESC4:.*]] = llvm.insertvalue %[[CST2]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) - // CHECK: %[[DESC5:.*]] = llvm.insertvalue %[[ARG7]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFFINC:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i32 - // CHECK32: %[[OFF1:.*]] = llvm.add %[[OFF]], %[[OFFINC]] : i32 - // CHECK32: %[[OFFINC1:.*]] = llvm.mul %[[ARG8]], %[[STRIDE1]] : i32 - // CHECK32: %[[OFF2:.*]] = llvm.add %[[OFF1]], %[[OFFINC1]] : i32 - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFF2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST2:.*]] = llvm.mlir.constant(2 : i64) - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG8]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST2]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG7]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][1, 2] : - memref<64x4xf32, offset: 0, strides: [4, 1]> - to memref - return -} - -// CHECK-LABEL: func @subview_const_stride_and_offset( -// CHECK32-LABEL: func @subview_const_stride_and_offset( -func @subview_const_stride_and_offset(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST8:.*]] = llvm.mlir.constant(8 : index) - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[CST8]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST3:.*]] = llvm.mlir.constant(3 : i64) - // CHECK32: %[[CST1:.*]] = llvm.mlir.constant(1 : i64) - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[CST3]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST62:.*]] = llvm.mlir.constant(62 : i64) - // CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64) - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - %1 = memref.subview %0[0, 8][62, 3][1, 1] : - memref<64x4xf32, offset: 0, strides: [4, 1]> - to memref<62x3xf32, offset: 8, strides: [4, 1]> - return -} - -// CHECK-LABEL: func @subview_mixed_static_dynamic( -// CHECK-COUNT-2: !llvm.ptr, -// CHECK-COUNT-5: {{%[a-zA-Z0-9]*}}: i64, -// CHECK: %[[ARG0:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG1:[a-zA-Z0-9]*]]: i64, -// CHECK: %[[ARG2:.*]]: i64) -// CHECK32-LABEL: func @subview_mixed_static_dynamic( -// CHECK32-COUNT-2: !llvm.ptr, -// CHECK32-COUNT-5: {{%[a-zA-Z0-9]*}}: i32, -// CHECK32: %[[ARG0:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG1:[a-zA-Z0-9]*]]: i32, -// CHECK32: %[[ARG2:.*]]: i32) -func @subview_mixed_static_dynamic(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index, %arg1 : index, %arg2 : index) { - // The last "insertvalue" that populates the memref descriptor from the function arguments. - // CHECK: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - // CHECK32: %[[MEMREF:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 1] - - // CHECK32: %[[DESC:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST0:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC0:.*]] = llvm.insertvalue %[[BITCAST0]], %[[DESC]][0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[BITCAST1:.*]] = llvm.bitcast %{{.*}} : !llvm.ptr to !llvm.ptr - // CHECK32: %[[DESC1:.*]] = llvm.insertvalue %[[BITCAST1]], %[[DESC0]][1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE0:.*]] = llvm.extractvalue %[[MEMREF]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[STRIDE1:.*]] = llvm.extractvalue %[[MEMREF]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFF:.*]] = llvm.extractvalue %[[MEMREF]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[OFFM1:.*]] = llvm.mul %[[ARG1]], %[[STRIDE0]] : i32 - // CHECK32: %[[OFFA1:.*]] = llvm.add %[[OFF]], %[[OFFM1]] : i32 - // CHECK32: %[[CST8:.*]] = llvm.mlir.constant(8 : i64) : i32 - // CHECK32: %[[OFFM2:.*]] = llvm.mul %[[CST8]], %[[STRIDE1]] : i32 - // CHECK32: %[[OFFA2:.*]] = llvm.add %[[OFFA1]], %[[OFFM2]] : i32 - // CHECK32: %[[DESC2:.*]] = llvm.insertvalue %[[OFFA2]], %[[DESC1]][2] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - - // CHECK32: %[[CST1:.*]] = llvm.mlir.constant(1 : i64) : i32 - // CHECK32: %[[DESC3:.*]] = llvm.insertvalue %[[ARG2]], %[[DESC2]][3, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[DESC4:.*]] = llvm.insertvalue %[[CST1]], %[[DESC3]][4, 1] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: %[[CST62:.*]] = llvm.mlir.constant(62 : i64) : i32 - // CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32 - // CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - // CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr, ptr, i32, array<2 x i32>, array<2 x i32>)> - %1 = memref.subview %0[%arg1, 8][62, %arg2][%arg0, 1] : - memref<64x4xf32, offset: 0, strides: [4, 1]> - to memref<62x?xf32, offset: ?, strides: [?, 1]> - return -} - -// CHECK-LABEL: func @subview_leading_operands( -func @subview_leading_operands(%0 : memref<5x3xf32>, %1: memref<5x?xf32>) { - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Alloc ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Aligned ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Offset - // CHECK: %[[C6:.*]] = llvm.mlir.constant(6 : index) : i64 - // CHECK: llvm.insertvalue %[[C6:.*]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Sizes and strides @rank 1: both static. - // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : index) : i64 - // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 - // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Sizes and strides @rank 0: both static extracted from type. - // CHECK: %[[C3_2:.*]] = llvm.mlir.constant(3 : i64) : i64 - // CHECK: %[[C3_3:.*]] = llvm.mlir.constant(3 : i64) : i64 - // CHECK: llvm.insertvalue %[[C3_2]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[C3_3]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %2 = memref.subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]> - - return -} - -// CHECK-LABEL: func @subview_leading_operands_dynamic( -func @subview_leading_operands_dynamic(%0 : memref<5x?xf32>) { - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Alloc ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Aligned ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Extract strides - // CHECK: %[[ST0:.*]] = llvm.extractvalue %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[ST1:.*]] = llvm.extractvalue %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Compute and insert offset from 2 + dynamic value. - // CHECK: %[[OFF:.*]] = llvm.extractvalue %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C2:.*]] = llvm.mlir.constant(2 : i64) : i64 - // CHECK: %[[MUL:.*]] = llvm.mul %[[C2]], %[[ST0]] : i64 - // CHECK: %[[NEW_OFF:.*]] = llvm.add %[[OFF]], %[[MUL]] : i64 - // CHECK: llvm.insertvalue %[[NEW_OFF]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Sizes and strides @rank 1: static stride 1, dynamic size unchanged from source memref. - // CHECK: %[[SZ1:.*]] = llvm.extractvalue %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 - // CHECK: llvm.insertvalue %[[SZ1]], %{{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Sizes and strides @rank 0: both static. - // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : i64) : i64 - // CHECK: %[[C1_2:.*]] = llvm.mlir.constant(1 : i64) : i64 - // CHECK: %[[MUL:.*]] = llvm.mul %[[C1_2]], %[[ST0]] : i64 - // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[MUL]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - %1 = memref.subview %0[2][3][1]: memref<5x?xf32> to memref<3x?xf32, offset: ?, strides: [?, 1]> - return -} - -// CHECK-LABEL: func @subview_rank_reducing_leading_operands( -func @subview_rank_reducing_leading_operands(%0 : memref<5x3xf32>) { - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Alloc ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Aligned ptr - // CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Extract strides - // CHECK: %[[ST0:.*]] = llvm.extractvalue %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[ST1:.*]] = llvm.extractvalue %{{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // Offset - // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : index) : i64 - // CHECK: llvm.insertvalue %[[C3:.*]], %{{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // Sizes and strides @rank 0: both static. - // CHECK: %[[C3:.*]] = llvm.mlir.constant(3 : index) : i64 - // CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64 - // CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - %1 = memref.subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]> - - return -} - - -// ----- - -// CHECK-LABEL: func @atomic_rmw -func @atomic_rmw(%I : memref<10xi32>, %ival : i32, %F : memref<10xf32>, %fval : f32, %i : index) { - atomic_rmw assign %fval, %F[%i] : (f32, memref<10xf32>) -> f32 - // CHECK: llvm.atomicrmw xchg %{{.*}}, %{{.*}} acq_rel - atomic_rmw addi %ival, %I[%i] : (i32, memref<10xi32>) -> i32 - // CHECK: llvm.atomicrmw add %{{.*}}, %{{.*}} acq_rel - atomic_rmw maxs %ival, %I[%i] : (i32, memref<10xi32>) -> i32 - // CHECK: llvm.atomicrmw max %{{.*}}, %{{.*}} acq_rel - atomic_rmw mins %ival, %I[%i] : (i32, memref<10xi32>) -> i32 - // CHECK: llvm.atomicrmw min %{{.*}}, %{{.*}} acq_rel - atomic_rmw maxu %ival, %I[%i] : (i32, memref<10xi32>) -> i32 - // CHECK: llvm.atomicrmw umax %{{.*}}, %{{.*}} acq_rel - atomic_rmw minu %ival, %I[%i] : (i32, memref<10xi32>) -> i32 - // CHECK: llvm.atomicrmw umin %{{.*}}, %{{.*}} acq_rel - atomic_rmw addf %fval, %F[%i] : (f32, memref<10xf32>) -> f32 - // CHECK: llvm.atomicrmw fadd %{{.*}}, %{{.*}} acq_rel - return -} - -// ----- - -// CHECK-LABEL: func @generic_atomic_rmw -func @generic_atomic_rmw(%I : memref<10xi32>, %i : index) -> i32 { - %x = generic_atomic_rmw %I[%i] : memref<10xi32> { - ^bb0(%old_value : i32): - %c1 = constant 1 : i32 - atomic_yield %c1 : i32 - } - // CHECK: [[init:%.*]] = llvm.load %{{.*}} : !llvm.ptr - // CHECK-NEXT: llvm.br ^bb1([[init]] : i32) - // CHECK-NEXT: ^bb1([[loaded:%.*]]: i32): - // CHECK-NEXT: [[c1:%.*]] = llvm.mlir.constant(1 : i32) - // CHECK-NEXT: [[pair:%.*]] = llvm.cmpxchg %{{.*}}, [[loaded]], [[c1]] - // CHECK-SAME: acq_rel monotonic : i32 - // CHECK-NEXT: [[new:%.*]] = llvm.extractvalue [[pair]][0] - // CHECK-NEXT: [[ok:%.*]] = llvm.extractvalue [[pair]][1] - // CHECK-NEXT: llvm.cond_br [[ok]], ^bb2, ^bb1([[new]] : i32) - // CHECK-NEXT: ^bb2: - %c2 = constant 2 : i32 - %add = addi %c2, %x : i32 - return %add : i32 - // CHECK-NEXT: [[c2:%.*]] = llvm.mlir.constant(2 : i32) - // CHECK-NEXT: [[add:%.*]] = llvm.add [[c2]], [[new]] : i32 - // CHECK-NEXT: llvm.return [[add]] -} - -// ----- - -// CHECK-LABEL: func @assume_alignment -func @assume_alignment(%0 : memref<4x4xf16>) { - // CHECK: %[[PTR:.*]] = llvm.extractvalue %[[MEMREF:.*]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK-NEXT: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK-NEXT: %[[MASK:.*]] = llvm.mlir.constant(15 : index) : i64 - // CHECK-NEXT: %[[INT:.*]] = llvm.ptrtoint %[[PTR]] : !llvm.ptr to i64 - // CHECK-NEXT: %[[MASKED_PTR:.*]] = llvm.and %[[INT]], %[[MASK:.*]] : i64 - // CHECK-NEXT: %[[CONDITION:.*]] = llvm.icmp "eq" %[[MASKED_PTR]], %[[ZERO]] : i64 - // CHECK-NEXT: "llvm.intr.assume"(%[[CONDITION]]) : (i1) -> () - memref.assume_alignment %0, 16 : memref<4x4xf16> - return -} - -// ----- - -// CHECK-LABEL: func @memref_index -// CHECK-SAME: %arg0: !llvm.ptr, %arg1: !llvm.ptr, -// CHECK-SAME: %arg2: i64, %arg3: i64, %arg4: i64) -// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> -// CHECK32-LABEL: func @memref_index -// CHECK32-SAME: %arg0: !llvm.ptr, %arg1: !llvm.ptr, -// CHECK32-SAME: %arg2: i32, %arg3: i32, %arg4: i32) -// CHECK32-SAME: -> !llvm.struct<(ptr, ptr, i32, array<1 x i32>, array<1 x i32>)> -func @memref_index(%arg0: memref<32xindex>) -> memref<32xindex> { - return %arg0 : memref<32xindex> -} - -// ----- - -// CHECK-LABEL: func @rank_of_unranked -// CHECK32-LABEL: func @rank_of_unranked -func @rank_of_unranked(%unranked: memref<*xi32>) { - %rank = rank %unranked : memref<*xi32> - return -} -// CHECK-NEXT: llvm.mlir.undef -// CHECK-NEXT: llvm.insertvalue -// CHECK-NEXT: llvm.insertvalue -// CHECK-NEXT: llvm.extractvalue %{{.*}}[0] : !llvm.struct<(i64, ptr)> -// CHECK32: llvm.extractvalue %{{.*}}[0] : !llvm.struct<(i32, ptr)> - -// CHECK-LABEL: func @rank_of_ranked -// CHECK32-LABEL: func @rank_of_ranked -func @rank_of_ranked(%ranked: memref) { - %rank = rank %ranked : memref - return -} -// CHECK: llvm.mlir.constant(1 : index) : i64 -// CHECK32: llvm.mlir.constant(1 : index) : i32 - -// ----- - -// CHECK-LABEL: func @dim_of_unranked -// CHECK32-LABEL: func @dim_of_unranked -func @dim_of_unranked(%unranked: memref<*xi32>) -> index { - %c0 = constant 0 : index - %dim = memref.dim %unranked, %c0 : memref<*xi32> - return %dim : index -} -// CHECK-NEXT: llvm.mlir.undef : !llvm.struct<(i64, ptr)> -// CHECK-NEXT: llvm.insertvalue -// CHECK-NEXT: %[[UNRANKED_DESC:.*]] = llvm.insertvalue -// CHECK-NEXT: %[[C0:.*]] = llvm.mlir.constant(0 : index) : i64 - -// CHECK-NEXT: %[[RANKED_DESC:.*]] = llvm.extractvalue %[[UNRANKED_DESC]][1] -// CHECK-SAME: : !llvm.struct<(i64, ptr)> - -// CHECK-NEXT: %[[ZERO_D_DESC:.*]] = llvm.bitcast %[[RANKED_DESC]] -// CHECK-SAME: : !llvm.ptr to !llvm.ptr, ptr, i64)>> - -// CHECK-NEXT: %[[C2_i32:.*]] = llvm.mlir.constant(2 : i32) : i32 -// CHECK-NEXT: %[[C0_:.*]] = llvm.mlir.constant(0 : index) : i64 - -// CHECK-NEXT: %[[OFFSET_PTR:.*]] = llvm.getelementptr %[[ZERO_D_DESC]]{{\[}} -// CHECK-SAME: %[[C0_]], %[[C2_i32]]] : (!llvm.ptr, ptr, -// CHECK-SAME: i64)>>, i64, i32) -> !llvm.ptr - -// CHECK-NEXT: %[[C1:.*]] = llvm.mlir.constant(1 : index) : i64 -// CHECK-NEXT: %[[INDEX_INC:.*]] = llvm.add %[[C1]], %[[C0]] : i64 - -// CHECK-NEXT: %[[SIZE_PTR:.*]] = llvm.getelementptr %[[OFFSET_PTR]]{{\[}} -// CHECK-SAME: %[[INDEX_INC]]] : (!llvm.ptr, i64) -> !llvm.ptr - -// CHECK-NEXT: %[[SIZE:.*]] = llvm.load %[[SIZE_PTR]] : !llvm.ptr -// CHECK-NEXT: llvm.return %[[SIZE]] : i64 - -// CHECK32: %[[SIZE:.*]] = llvm.load %{{.*}} : !llvm.ptr -// CHECK32-NEXT: llvm.return %[[SIZE]] : i32 - -// ----- - -func @switchLower(%arg0: i32, %arg1 : i32, %arg2 : i32) -> i32 { - switch %arg0 : i32, [ - default: ^bb2, - 115: ^bb1 - ] - ^bb1: - llvm.return %arg1 : i32 - ^bb2: - llvm.return %arg2 : i32 -} - -// CHECK: llvm.switch %arg0, ^[[bb2:.+]] [ -// CHECK-NEXT: 115: ^[[bb1:.+]] -// CHECK-NEXT: ] -// CHECK: ^[[bb1]]: -// CHECK: llvm.return %arg1 : i32 -// CHECK: ^[[bb2]]: -// CHECK: llvm.return %arg2 : i32 diff --git a/mlir/test/Conversion/StandardToLLVM/func-memref.mlir b/mlir/test/Conversion/StandardToLLVM/func-memref.mlir new file mode 100644 --- /dev/null +++ b/mlir/test/Conversion/StandardToLLVM/func-memref.mlir @@ -0,0 +1,184 @@ +// RUN: mlir-opt -convert-std-to-llvm -split-input-file %s | FileCheck %s +// RUN: mlir-opt -convert-std-to-llvm='use-bare-ptr-memref-call-conv=1' -split-input-file %s | FileCheck %s --check-prefix=BAREPTR + +// BAREPTR-LABEL: func @check_noalias +// BAREPTR-SAME: %{{.*}}: !llvm.ptr {llvm.noalias}, %{{.*}}: !llvm.ptr {llvm.noalias} +func @check_noalias(%static : memref<2xf32> {llvm.noalias}, %other : memref<2xf32> {llvm.noalias}) { + return +} + +// ----- + +// CHECK-LABEL: func @check_strided_memref_arguments( +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +func @check_strided_memref_arguments(%static: memref<10x20xf32, affine_map<(i,j)->(20 * i + j + 1)>>, + %dynamic : memref(M * i + j + 1)>>, + %mixed : memref<10x?xf32, affine_map<(i,j)[M]->(M * i + j + 1)>>) { + return +} + +// ----- + +// CHECK-LABEL: func @memref_index +// CHECK-SAME: %arg0: !llvm.ptr, %arg1: !llvm.ptr, +// CHECK-SAME: %arg2: i64, %arg3: i64, %arg4: i64) +// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> +// CHECK32-LABEL: func @memref_index +// CHECK32-SAME: %arg0: !llvm.ptr, %arg1: !llvm.ptr, +// CHECK32-SAME: %arg2: i32, %arg3: i32, %arg4: i32) +// CHECK32-SAME: -> !llvm.struct<(ptr, ptr, i32, array<1 x i32>, array<1 x i32>)> +func @memref_index(%arg0: memref<32xindex>) -> memref<32xindex> { + return %arg0 : memref<32xindex> +} + +// ----- + +// CHECK-LABEL: func @check_arguments +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +func @check_arguments(%static: memref<10x20xf32>, %dynamic : memref, %mixed : memref<10x?xf32>) { + return +} + +// ----- + +// CHECK-LABEL: func @check_static_return +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-LABEL: func @check_static_return +// BAREPTR-SAME: (%[[arg:.*]]: !llvm.ptr) -> !llvm.ptr { +func @check_static_return(%static : memref<32x18xf32>) -> memref<32x18xf32> { +// CHECK: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + +// BAREPTR: %[[udf:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[base0:.*]] = llvm.insertvalue %[[arg]], %[[udf]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[aligned:.*]] = llvm.insertvalue %[[arg]], %[[base0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val0:.*]] = llvm.mlir.constant(0 : index) : i64 +// BAREPTR-NEXT: %[[ins0:.*]] = llvm.insertvalue %[[val0]], %[[aligned]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val1:.*]] = llvm.mlir.constant(32 : index) : i64 +// BAREPTR-NEXT: %[[ins1:.*]] = llvm.insertvalue %[[val1]], %[[ins0]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val2:.*]] = llvm.mlir.constant(18 : index) : i64 +// BAREPTR-NEXT: %[[ins2:.*]] = llvm.insertvalue %[[val2]], %[[ins1]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val3:.*]] = llvm.mlir.constant(18 : index) : i64 +// BAREPTR-NEXT: %[[ins3:.*]] = llvm.insertvalue %[[val3]], %[[ins2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val4:.*]] = llvm.mlir.constant(1 : index) : i64 +// BAREPTR-NEXT: %[[ins4:.*]] = llvm.insertvalue %[[val4]], %[[ins3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[base1:.*]] = llvm.extractvalue %[[ins4]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: llvm.return %[[base1]] : !llvm.ptr + return %static : memref<32x18xf32> +} + +// ----- + +// CHECK-LABEL: func @check_static_return_with_offset +// CHECK-COUNT-2: !llvm.ptr +// CHECK-COUNT-5: i64 +// CHECK-SAME: -> !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-LABEL: func @check_static_return_with_offset +// BAREPTR-SAME: (%[[arg:.*]]: !llvm.ptr) -> !llvm.ptr { +func @check_static_return_with_offset(%static : memref<32x18xf32, offset:7, strides:[22,1]>) -> memref<32x18xf32, offset:7, strides:[22,1]> { +// CHECK: llvm.return %{{.*}} : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> + +// BAREPTR: %[[udf:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[base0:.*]] = llvm.insertvalue %[[arg]], %[[udf]][0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[aligned:.*]] = llvm.insertvalue %[[arg]], %[[base0]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val0:.*]] = llvm.mlir.constant(7 : index) : i64 +// BAREPTR-NEXT: %[[ins0:.*]] = llvm.insertvalue %[[val0]], %[[aligned]][2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val1:.*]] = llvm.mlir.constant(32 : index) : i64 +// BAREPTR-NEXT: %[[ins1:.*]] = llvm.insertvalue %[[val1]], %[[ins0]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val2:.*]] = llvm.mlir.constant(22 : index) : i64 +// BAREPTR-NEXT: %[[ins2:.*]] = llvm.insertvalue %[[val2]], %[[ins1]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val3:.*]] = llvm.mlir.constant(18 : index) : i64 +// BAREPTR-NEXT: %[[ins3:.*]] = llvm.insertvalue %[[val3]], %[[ins2]][3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[val4:.*]] = llvm.mlir.constant(1 : index) : i64 +// BAREPTR-NEXT: %[[ins4:.*]] = llvm.insertvalue %[[val4]], %[[ins3]][4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: %[[base1:.*]] = llvm.extractvalue %[[ins4]][1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> +// BAREPTR-NEXT: llvm.return %[[base1]] : !llvm.ptr + return %static : memref<32x18xf32, offset:7, strides:[22,1]> +} + +// ----- + +// BAREPTR: llvm.func @foo(!llvm.ptr) -> !llvm.ptr +func private @foo(memref<10xi8>) -> memref<20xi8> + +// BAREPTR-LABEL: func @check_memref_func_call +// BAREPTR-SAME: %[[in:.*]]: !llvm.ptr) -> !llvm.ptr +func @check_memref_func_call(%in : memref<10xi8>) -> memref<20xi8> { + // BAREPTR: %[[inDesc:.*]] = llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] + // BAREPTR-NEXT: %[[barePtr:.*]] = llvm.extractvalue %[[inDesc]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[call:.*]] = llvm.call @foo(%[[barePtr]]) : (!llvm.ptr) -> !llvm.ptr + // BAREPTR-NEXT: %[[desc0:.*]] = llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[desc1:.*]] = llvm.insertvalue %[[call]], %[[desc0]][0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[desc2:.*]] = llvm.insertvalue %[[call]], %[[desc1]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64 + // BAREPTR-NEXT: %[[desc4:.*]] = llvm.insertvalue %[[c0]], %[[desc2]][2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[c20:.*]] = llvm.mlir.constant(20 : index) : i64 + // BAREPTR-NEXT: %[[desc6:.*]] = llvm.insertvalue %[[c20]], %[[desc4]][3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: %[[c1:.*]] = llvm.mlir.constant(1 : index) : i64 + // BAREPTR-NEXT: %[[outDesc:.*]] = llvm.insertvalue %[[c1]], %[[desc6]][4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + %res = call @foo(%in) : (memref<10xi8>) -> (memref<20xi8>) + // BAREPTR-NEXT: %[[res:.*]] = llvm.extractvalue %[[outDesc]][1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> + // BAREPTR-NEXT: llvm.return %[[res]] : !llvm.ptr + return %res : memref<20xi8> +} + +// ----- + +// Unranked memrefs are currently not supported in the bare-ptr calling +// convention. Check that the conversion to the LLVM-IR dialect doesn't happen +// in the presence of unranked memrefs when using such a calling convention. + +// BAREPTR: func private @hoo(memref<*xi8>) -> memref<*xi8> +func private @hoo(memref<*xi8>) -> memref<*xi8> + +// BAREPTR-LABEL: func @check_unranked_memref_func_call(%{{.*}}: memref<*xi8>) -> memref<*xi8> +func @check_unranked_memref_func_call(%in: memref<*xi8>) -> memref<*xi8> { + // BAREPTR-NEXT: call @hoo(%{{.*}}) : (memref<*xi8>) -> memref<*xi8> + %res = call @hoo(%in) : (memref<*xi8>) -> memref<*xi8> + // BAREPTR-NEXT: return %{{.*}} : memref<*xi8> + return %res : memref<*xi8> +} + +// ----- + +// Should not convert memrefs with unsupported types in any convention. + +// CHECK: @unsupported_memref_element_type +// CHECK-SAME: memref< +// CHECK-NOT: !llvm.struct +// BAREPTR: @unsupported_memref_element_type +// BAREPTR-SAME: memref< +// BAREPTR-NOT: !llvm.ptr +func private @unsupported_memref_element_type() -> memref<42 x !test.memref_element> + +// CHECK: @unsupported_unranked_memref_element_type +// CHECK-SAME: memref< +// CHECK-NOT: !llvm.struct +// BAREPTR: @unsupported_unranked_memref_element_type +// BAREPTR-SAME: memref< +// BAREPTR-NOT: !llvm.ptr +func private @unsupported_unranked_memref_element_type() -> memref<* x !test.memref_element> + +// ----- + +// BAREPTR: llvm.func @goo(f32) -> f32 +func private @goo(f32) -> f32 + +// BAREPTR-LABEL: func @check_scalar_func_call +// BAREPTR-SAME: %[[in:.*]]: f32) +func @check_scalar_func_call(%in : f32) { + // BAREPTR-NEXT: %[[call:.*]] = llvm.call @goo(%[[in]]) : (f32) -> f32 + %res = call @goo(%in) : (f32) -> (f32) + return +} diff --git a/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir b/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir --- a/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir +++ b/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir @@ -1,15 +1,865 @@ -// RUN: mlir-opt -allow-unregistered-dialect %s -convert-std-to-llvm -split-input-file -verify-diagnostics | FileCheck %s - -// CHECK-LABEL: func @address_space( -// CHECK-SAME: !llvm.ptr -func @address_space(%arg0 : memref<32xf32, affine_map<(d0) -> (d0)>, 7>) { - %0 = memref.alloc() : memref<32xf32, affine_map<(d0) -> (d0)>, 5> - %1 = constant 7 : index - // CHECK: llvm.load %{{.*}} : !llvm.ptr - %2 = memref.load %0[%1] : memref<32xf32, affine_map<(d0) -> (d0)>, 5> - std.return +// RUN: mlir-opt -convert-std-to-llvm %s -split-input-file | FileCheck %s +// RUN: mlir-opt -convert-std-to-llvm='index-bitwidth=32' %s -split-input-file | FileCheck --check-prefix=CHECK32 %s + +// CHECK-LABEL: func @empty() { +// CHECK-NEXT: llvm.return +// CHECK-NEXT: } +func @empty() { +^bb0: + return +} + +// CHECK-LABEL: llvm.func @body(i64) +func private @body(index) + +// CHECK-LABEL: func @simple_loop() { +// CHECK32-LABEL: func @simple_loop() { +func @simple_loop() { +^bb0: +// CHECK-NEXT: llvm.br ^bb1 +// CHECK32-NEXT: llvm.br ^bb1 + br ^bb1 + +// CHECK-NEXT: ^bb1: // pred: ^bb0 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK32-NEXT: ^bb1: // pred: ^bb0 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i32 +// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) +^bb1: // pred: ^bb0 + %c1 = constant 1 : index + %c42 = constant 42 : index + br ^bb2(%c1 : index) + +// CHECK: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb3 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 +// CHECK32: ^bb2({{.*}}: i32): // 2 preds: ^bb1, ^bb3 +// CHECK32-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i32 +// CHECK32-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 +^bb2(%0: index): // 2 preds: ^bb1, ^bb3 + %1 = cmpi slt, %0, %c42 : index + cond_br %1, ^bb3, ^bb4 + +// CHECK: ^bb3: // pred: ^bb2 +// CHECK-NEXT: llvm.call @body({{.*}}) : (i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK32: ^bb3: // pred: ^bb2 +// CHECK32-NEXT: llvm.call @body({{.*}}) : (i32) -> () +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 +// CHECK32-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i32 +// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) +^bb3: // pred: ^bb2 + call @body(%0) : (index) -> () + %c1_0 = constant 1 : index + %2 = addi %0, %c1_0 : index + br ^bb2(%2 : index) + +// CHECK: ^bb4: // pred: ^bb2 +// CHECK-NEXT: llvm.return +^bb4: // pred: ^bb2 + return +} + +// CHECK-LABEL: func @simple_caller() { +// CHECK-NEXT: llvm.call @simple_loop() : () -> () +// CHECK-NEXT: llvm.return +// CHECK-NEXT: } +func @simple_caller() { +^bb0: + call @simple_loop() : () -> () + return +} + +// Check that function call attributes persist during conversion. +// CHECK-LABEL: @call_with_attributes +func @call_with_attributes() { + // CHECK: llvm.call @simple_loop() {baz = [1, 2, 3, 4], foo = "bar"} : () -> () + call @simple_loop() {foo="bar", baz=[1,2,3,4]} : () -> () + return +} + +// CHECK-LABEL: func @ml_caller() { +// CHECK-NEXT: llvm.call @simple_loop() : () -> () +// CHECK-NEXT: llvm.call @more_imperfectly_nested_loops() : () -> () +// CHECK-NEXT: llvm.return +// CHECK-NEXT: } +func @ml_caller() { +^bb0: + call @simple_loop() : () -> () + call @more_imperfectly_nested_loops() : () -> () + return +} + +// CHECK-LABEL: llvm.func @body_args(i64) -> i64 +// CHECK32-LABEL: llvm.func @body_args(i32) -> i32 +func private @body_args(index) -> index +// CHECK-LABEL: llvm.func @other(i64, i32) -> i32 +// CHECK32-LABEL: llvm.func @other(i32, i32) -> i32 +func private @other(index, i32) -> i32 + +// CHECK-LABEL: func @func_args(%arg0: i32, %arg1: i32) -> i32 { +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : i32) : i32 +// CHECK-NEXT: llvm.br ^bb1 +// CHECK32-LABEL: func @func_args(%arg0: i32, %arg1: i32) -> i32 { +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : i32) : i32 +// CHECK32-NEXT: llvm.br ^bb1 +func @func_args(i32, i32) -> i32 { +^bb0(%arg0: i32, %arg1: i32): + %c0_i32 = constant 0 : i32 + br ^bb1 + +// CHECK-NEXT: ^bb1: // pred: ^bb0 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK32-NEXT: ^bb1: // pred: ^bb0 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i32 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i32 +// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) +^bb1: // pred: ^bb0 + %c0 = constant 0 : index + %c42 = constant 42 : index + br ^bb2(%c0 : index) + +// CHECK-NEXT: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb3 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 +// CHECK32-NEXT: ^bb2({{.*}}: i32): // 2 preds: ^bb1, ^bb3 +// CHECK32-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i32 +// CHECK32-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb4 +^bb2(%0: index): // 2 preds: ^bb1, ^bb3 + %1 = cmpi slt, %0, %c42 : index + cond_br %1, ^bb3, ^bb4 + +// CHECK-NEXT: ^bb3: // pred: ^bb2 +// CHECK-NEXT: {{.*}} = llvm.call @body_args({{.*}}) : (i64) -> i64 +// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg0) : (i64, i32) -> i32 +// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i64, i32) -> i32 +// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg1) : (i64, i32) -> i32 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK32-NEXT: ^bb3: // pred: ^bb2 +// CHECK32-NEXT: {{.*}} = llvm.call @body_args({{.*}}) : (i32) -> i32 +// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg0) : (i32, i32) -> i32 +// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i32, i32) -> i32 +// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, %arg1) : (i32, i32) -> i32 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i32 +// CHECK32-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i32 +// CHECK32-NEXT: llvm.br ^bb2({{.*}} : i32) +^bb3: // pred: ^bb2 + %2 = call @body_args(%0) : (index) -> index + %3 = call @other(%2, %arg0) : (index, i32) -> i32 + %4 = call @other(%2, %3) : (index, i32) -> i32 + %5 = call @other(%2, %arg1) : (index, i32) -> i32 + %c1 = constant 1 : index + %6 = addi %0, %c1 : index + br ^bb2(%6 : index) + +// CHECK-NEXT: ^bb4: // pred: ^bb2 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i64, i32) -> i32 +// CHECK-NEXT: llvm.return {{.*}} : i32 +// CHECK32-NEXT: ^bb4: // pred: ^bb2 +// CHECK32-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i32 +// CHECK32-NEXT: {{.*}} = llvm.call @other({{.*}}, {{.*}}) : (i32, i32) -> i32 +// CHECK32-NEXT: llvm.return {{.*}} : i32 +^bb4: // pred: ^bb2 + %c0_0 = constant 0 : index + %7 = call @other(%c0_0, %c0_i32) : (index, i32) -> i32 + return %7 : i32 +} + +// CHECK-LABEL: llvm.func @pre(i64) +// CHECK32-LABEL: llvm.func @pre(i32) +func private @pre(index) + +// CHECK-LABEL: llvm.func @body2(i64, i64) +// CHECK32-LABEL: llvm.func @body2(i32, i32) +func private @body2(index, index) + +// CHECK-LABEL: llvm.func @post(i64) +// CHECK32-LABEL: llvm.func @post(i32) +func private @post(index) + +// CHECK-LABEL: func @imperfectly_nested_loops() { +// CHECK-NEXT: llvm.br ^bb1 +func @imperfectly_nested_loops() { +^bb0: + br ^bb1 + +// CHECK-NEXT: ^bb1: // pred: ^bb0 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +^bb1: // pred: ^bb0 + %c0 = constant 0 : index + %c42 = constant 42 : index + br ^bb2(%c0 : index) + +// CHECK-NEXT: ^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb7 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb8 +^bb2(%0: index): // 2 preds: ^bb1, ^bb7 + %1 = cmpi slt, %0, %c42 : index + cond_br %1, ^bb3, ^bb8 + +// CHECK-NEXT: ^bb3: +// CHECK-NEXT: llvm.call @pre({{.*}}) : (i64) -> () +// CHECK-NEXT: llvm.br ^bb4 +^bb3: // pred: ^bb2 + call @pre(%0) : (index) -> () + br ^bb4 + +// CHECK-NEXT: ^bb4: // pred: ^bb3 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(7 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(56 : index) : i64 +// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) +^bb4: // pred: ^bb3 + %c7 = constant 7 : index + %c56 = constant 56 : index + br ^bb5(%c7 : index) + +// CHECK-NEXT: ^bb5({{.*}}: i64): // 2 preds: ^bb4, ^bb6 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb6, ^bb7 +^bb5(%2: index): // 2 preds: ^bb4, ^bb6 + %3 = cmpi slt, %2, %c56 : index + cond_br %3, ^bb6, ^bb7 + +// CHECK-NEXT: ^bb6: // pred: ^bb5 +// CHECK-NEXT: llvm.call @body2({{.*}}, {{.*}}) : (i64, i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(2 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) +^bb6: // pred: ^bb5 + call @body2(%0, %2) : (index, index) -> () + %c2 = constant 2 : index + %4 = addi %2, %c2 : index + br ^bb5(%4 : index) + +// CHECK-NEXT: ^bb7: // pred: ^bb5 +// CHECK-NEXT: llvm.call @post({{.*}}) : (i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +^bb7: // pred: ^bb5 + call @post(%0) : (index) -> () + %c1 = constant 1 : index + %5 = addi %0, %c1 : index + br ^bb2(%5 : index) + +// CHECK-NEXT: ^bb8: // pred: ^bb2 +// CHECK-NEXT: llvm.return +^bb8: // pred: ^bb2 + return +} + +// CHECK-LABEL: llvm.func @mid(i64) +func private @mid(index) + +// CHECK-LABEL: llvm.func @body3(i64, i64) +func private @body3(index, index) + +// A complete function transformation check. +// CHECK-LABEL: func @more_imperfectly_nested_loops() { +// CHECK-NEXT: llvm.br ^bb1 +// CHECK-NEXT:^bb1: // pred: ^bb0 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(0 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(42 : index) : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK-NEXT:^bb2({{.*}}: i64): // 2 preds: ^bb1, ^bb11 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb3, ^bb12 +// CHECK-NEXT:^bb3: // pred: ^bb2 +// CHECK-NEXT: llvm.call @pre({{.*}}) : (i64) -> () +// CHECK-NEXT: llvm.br ^bb4 +// CHECK-NEXT:^bb4: // pred: ^bb3 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(7 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(56 : index) : i64 +// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) +// CHECK-NEXT:^bb5({{.*}}: i64): // 2 preds: ^bb4, ^bb6 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb6, ^bb7 +// CHECK-NEXT:^bb6: // pred: ^bb5 +// CHECK-NEXT: llvm.call @body2({{.*}}, {{.*}}) : (i64, i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(2 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb5({{.*}} : i64) +// CHECK-NEXT:^bb7: // pred: ^bb5 +// CHECK-NEXT: llvm.call @mid({{.*}}) : (i64) -> () +// CHECK-NEXT: llvm.br ^bb8 +// CHECK-NEXT:^bb8: // pred: ^bb7 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(18 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(37 : index) : i64 +// CHECK-NEXT: llvm.br ^bb9({{.*}} : i64) +// CHECK-NEXT:^bb9({{.*}}: i64): // 2 preds: ^bb8, ^bb10 +// CHECK-NEXT: {{.*}} = llvm.icmp "slt" {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.cond_br {{.*}}, ^bb10, ^bb11 +// CHECK-NEXT:^bb10: // pred: ^bb9 +// CHECK-NEXT: llvm.call @body3({{.*}}, {{.*}}) : (i64, i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(3 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb9({{.*}} : i64) +// CHECK-NEXT:^bb11: // pred: ^bb9 +// CHECK-NEXT: llvm.call @post({{.*}}) : (i64) -> () +// CHECK-NEXT: {{.*}} = llvm.mlir.constant(1 : index) : i64 +// CHECK-NEXT: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 +// CHECK-NEXT: llvm.br ^bb2({{.*}} : i64) +// CHECK-NEXT:^bb12: // pred: ^bb2 +// CHECK-NEXT: llvm.return +// CHECK-NEXT: } +func @more_imperfectly_nested_loops() { +^bb0: + br ^bb1 +^bb1: // pred: ^bb0 + %c0 = constant 0 : index + %c42 = constant 42 : index + br ^bb2(%c0 : index) +^bb2(%0: index): // 2 preds: ^bb1, ^bb11 + %1 = cmpi slt, %0, %c42 : index + cond_br %1, ^bb3, ^bb12 +^bb3: // pred: ^bb2 + call @pre(%0) : (index) -> () + br ^bb4 +^bb4: // pred: ^bb3 + %c7 = constant 7 : index + %c56 = constant 56 : index + br ^bb5(%c7 : index) +^bb5(%2: index): // 2 preds: ^bb4, ^bb6 + %3 = cmpi slt, %2, %c56 : index + cond_br %3, ^bb6, ^bb7 +^bb6: // pred: ^bb5 + call @body2(%0, %2) : (index, index) -> () + %c2 = constant 2 : index + %4 = addi %2, %c2 : index + br ^bb5(%4 : index) +^bb7: // pred: ^bb5 + call @mid(%0) : (index) -> () + br ^bb8 +^bb8: // pred: ^bb7 + %c18 = constant 18 : index + %c37 = constant 37 : index + br ^bb9(%c18 : index) +^bb9(%5: index): // 2 preds: ^bb8, ^bb10 + %6 = cmpi slt, %5, %c37 : index + cond_br %6, ^bb10, ^bb11 +^bb10: // pred: ^bb9 + call @body3(%0, %5) : (index, index) -> () + %c3 = constant 3 : index + %7 = addi %5, %c3 : index + br ^bb9(%7 : index) +^bb11: // pred: ^bb9 + call @post(%0) : (index) -> () + %c1 = constant 1 : index + %8 = addi %0, %c1 : index + br ^bb2(%8 : index) +^bb12: // pred: ^bb2 + return +} + +// CHECK-LABEL: llvm.func @get_i64() -> i64 +func private @get_i64() -> (i64) +// CHECK-LABEL: llvm.func @get_f32() -> f32 +func private @get_f32() -> (f32) +// CHECK-LABEL: llvm.func @get_c16() -> !llvm.struct<(f16, f16)> +func private @get_c16() -> (complex) +// CHECK-LABEL: llvm.func @get_c32() -> !llvm.struct<(f32, f32)> +func private @get_c32() -> (complex) +// CHECK-LABEL: llvm.func @get_c64() -> !llvm.struct<(f64, f64)> +func private @get_c64() -> (complex) +// CHECK-LABEL: llvm.func @get_memref() -> !llvm.struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)> +// CHECK32-LABEL: llvm.func @get_memref() -> !llvm.struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)> +func private @get_memref() -> (memref<42x?x10x?xf32>) + +// CHECK-LABEL: llvm.func @multireturn() -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> { +// CHECK32-LABEL: llvm.func @multireturn() -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> { +func @multireturn() -> (i64, f32, memref<42x?x10x?xf32>) { +^bb0: +// CHECK-NEXT: {{.*}} = llvm.call @get_i64() : () -> i64 +// CHECK-NEXT: {{.*}} = llvm.call @get_f32() : () -> f32 +// CHECK-NEXT: {{.*}} = llvm.call @get_memref() : () -> !llvm.struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)> +// CHECK32-NEXT: {{.*}} = llvm.call @get_i64() : () -> i64 +// CHECK32-NEXT: {{.*}} = llvm.call @get_f32() : () -> f32 +// CHECK32-NEXT: {{.*}} = llvm.call @get_memref() : () -> !llvm.struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)> + %0 = call @get_i64() : () -> (i64) + %1 = call @get_f32() : () -> (f32) + %2 = call @get_memref() : () -> (memref<42x?x10x?xf32>) +// CHECK-NEXT: {{.*}} = llvm.mlir.undef : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: llvm.return {{.*}} : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK32-NEXT: {{.*}} = llvm.mlir.undef : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.insertvalue {{.*}}, {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: llvm.return {{.*}} : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> + return %0, %1, %2 : i64, f32, memref<42x?x10x?xf32> +} + + +// CHECK-LABEL: llvm.func @multireturn_caller() { +// CHECK32-LABEL: llvm.func @multireturn_caller() { +func @multireturn_caller() { +^bb0: +// CHECK-NEXT: {{.*}} = llvm.call @multireturn() : () -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK-NEXT: {{.*}} = llvm.extractvalue {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i64, array<4 x i64>, array<4 x i64>)>)> +// CHECK32-NEXT: {{.*}} = llvm.call @multireturn() : () -> !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[0] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[1] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> +// CHECK32-NEXT: {{.*}} = llvm.extractvalue {{.*}}[2] : !llvm.struct<(i64, f32, struct<(ptr, ptr, i32, array<4 x i32>, array<4 x i32>)>)> + %0:3 = call @multireturn() : () -> (i64, f32, memref<42x?x10x?xf32>) + %1 = constant 42 : i64 +// CHECK: {{.*}} = llvm.add {{.*}}, {{.*}} : i64 + %2 = addi %0#0, %1 : i64 + %3 = constant 42.0 : f32 +// CHECK: {{.*}} = llvm.fadd {{.*}}, {{.*}} : f32 + %4 = addf %0#1, %3 : f32 + %5 = constant 0 : index + return +} + +// CHECK-LABEL: llvm.func @vector_ops(%arg0: vector<4xf32>, %arg1: vector<4xi1>, %arg2: vector<4xi64>, %arg3: vector<4xi64>) -> vector<4xf32> { +func @vector_ops(%arg0: vector<4xf32>, %arg1: vector<4xi1>, %arg2: vector<4xi64>, %arg3: vector<4xi64>) -> vector<4xf32> { +// CHECK-NEXT: %0 = llvm.mlir.constant(dense<4.200000e+01> : vector<4xf32>) : vector<4xf32> + %0 = constant dense<42.> : vector<4xf32> +// CHECK-NEXT: %1 = llvm.fadd %arg0, %0 : vector<4xf32> + %1 = addf %arg0, %0 : vector<4xf32> +// CHECK-NEXT: %2 = llvm.sdiv %arg2, %arg2 : vector<4xi64> + %3 = divi_signed %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %3 = llvm.udiv %arg2, %arg2 : vector<4xi64> + %4 = divi_unsigned %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %4 = llvm.srem %arg2, %arg2 : vector<4xi64> + %5 = remi_signed %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %5 = llvm.urem %arg2, %arg2 : vector<4xi64> + %6 = remi_unsigned %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %6 = llvm.fdiv %arg0, %0 : vector<4xf32> + %7 = divf %arg0, %0 : vector<4xf32> +// CHECK-NEXT: %7 = llvm.frem %arg0, %0 : vector<4xf32> + %8 = remf %arg0, %0 : vector<4xf32> +// CHECK-NEXT: %8 = llvm.and %arg2, %arg3 : vector<4xi64> + %9 = and %arg2, %arg3 : vector<4xi64> +// CHECK-NEXT: %9 = llvm.or %arg2, %arg3 : vector<4xi64> + %10 = or %arg2, %arg3 : vector<4xi64> +// CHECK-NEXT: %10 = llvm.xor %arg2, %arg3 : vector<4xi64> + %11 = xor %arg2, %arg3 : vector<4xi64> +// CHECK-NEXT: %11 = llvm.shl %arg2, %arg2 : vector<4xi64> + %12 = shift_left %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %12 = llvm.ashr %arg2, %arg2 : vector<4xi64> + %13 = shift_right_signed %arg2, %arg2 : vector<4xi64> +// CHECK-NEXT: %13 = llvm.lshr %arg2, %arg2 : vector<4xi64> + %14 = shift_right_unsigned %arg2, %arg2 : vector<4xi64> + return %1 : vector<4xf32> +} + +// CHECK-LABEL: @ops +func @ops(f32, f32, i32, i32, f64) -> (f32, i32) { +^bb0(%arg0: f32, %arg1: f32, %arg2: i32, %arg3: i32, %arg4: f64): +// CHECK-NEXT: %0 = llvm.fsub %arg0, %arg1 : f32 + %0 = subf %arg0, %arg1: f32 +// CHECK-NEXT: %1 = llvm.sub %arg2, %arg3 : i32 + %1 = subi %arg2, %arg3: i32 +// CHECK-NEXT: %2 = llvm.icmp "slt" %arg2, %1 : i32 + %2 = cmpi slt, %arg2, %1 : i32 +// CHECK-NEXT: %3 = llvm.sdiv %arg2, %arg3 : i32 + %3 = divi_signed %arg2, %arg3 : i32 +// CHECK-NEXT: %4 = llvm.udiv %arg2, %arg3 : i32 + %4 = divi_unsigned %arg2, %arg3 : i32 +// CHECK-NEXT: %5 = llvm.srem %arg2, %arg3 : i32 + %5 = remi_signed %arg2, %arg3 : i32 +// CHECK-NEXT: %6 = llvm.urem %arg2, %arg3 : i32 + %6 = remi_unsigned %arg2, %arg3 : i32 +// CHECK-NEXT: %7 = llvm.select %2, %arg2, %arg3 : i1, i32 + %7 = select %2, %arg2, %arg3 : i32 +// CHECK-NEXT: %8 = llvm.fdiv %arg0, %arg1 : f32 + %8 = divf %arg0, %arg1 : f32 +// CHECK-NEXT: %9 = llvm.frem %arg0, %arg1 : f32 + %9 = remf %arg0, %arg1 : f32 +// CHECK-NEXT: %10 = llvm.and %arg2, %arg3 : i32 + %10 = and %arg2, %arg3 : i32 +// CHECK-NEXT: %11 = llvm.or %arg2, %arg3 : i32 + %11 = or %arg2, %arg3 : i32 +// CHECK-NEXT: %12 = llvm.xor %arg2, %arg3 : i32 + %12 = xor %arg2, %arg3 : i32 +// CHECK-NEXT: %13 = "llvm.intr.exp"(%arg0) : (f32) -> f32 + %13 = math.exp %arg0 : f32 +// CHECK-NEXT: %14 = "llvm.intr.exp2"(%arg0) : (f32) -> f32 + %14 = math.exp2 %arg0 : f32 +// CHECK-NEXT: %15 = llvm.mlir.constant(7.900000e-01 : f64) : f64 + %15 = constant 7.9e-01 : f64 +// CHECK-NEXT: %16 = llvm.shl %arg2, %arg3 : i32 + %16 = shift_left %arg2, %arg3 : i32 +// CHECK-NEXT: %17 = llvm.ashr %arg2, %arg3 : i32 + %17 = shift_right_signed %arg2, %arg3 : i32 +// CHECK-NEXT: %18 = llvm.lshr %arg2, %arg3 : i32 + %18 = shift_right_unsigned %arg2, %arg3 : i32 +// CHECK-NEXT: %{{[0-9]+}} = "llvm.intr.sqrt"(%arg0) : (f32) -> f32 + %19 = math.sqrt %arg0 : f32 +// CHECK-NEXT: %{{[0-9]+}} = "llvm.intr.sqrt"(%arg4) : (f64) -> f64 + %20 = math.sqrt %arg4 : f64 + return %0, %4 : f32, i32 +} + +// Checking conversion of index types to integers using i1, assuming no target +// system would have a 1-bit address space. Otherwise, we would have had to +// make this test dependent on the pointer size on the target system. +// CHECK-LABEL: @index_cast +func @index_cast(%arg0: index, %arg1: i1) { +// CHECK-NEXT: = llvm.trunc %arg0 : i{{.*}} to i1 + %0 = index_cast %arg0: index to i1 +// CHECK-NEXT: = llvm.sext %arg1 : i1 to i{{.*}} + %1 = index_cast %arg1: i1 to index + return +} + +// CHECK-LABEL: @vector_index_cast +func @vector_index_cast(%arg0: vector<2xindex>, %arg1: vector<2xi1>) { +// CHECK-NEXT: = llvm.trunc %{{.*}} : vector<2xi{{.*}}> to vector<2xi1> + %0 = index_cast %arg0: vector<2xindex> to vector<2xi1> +// CHECK-NEXT: = llvm.sext %{{.*}} : vector<2xi1> to vector<2xi{{.*}}> + %1 = index_cast %arg1: vector<2xi1> to vector<2xindex> + return +} + +// Checking conversion of signed integer types to floating point. +// CHECK-LABEL: @sitofp +func @sitofp(%arg0 : i32, %arg1 : i64) { +// CHECK-NEXT: = llvm.sitofp {{.*}} : i32 to f32 + %0 = sitofp %arg0: i32 to f32 +// CHECK-NEXT: = llvm.sitofp {{.*}} : i32 to f64 + %1 = sitofp %arg0: i32 to f64 +// CHECK-NEXT: = llvm.sitofp {{.*}} : i64 to f32 + %2 = sitofp %arg1: i64 to f32 +// CHECK-NEXT: = llvm.sitofp {{.*}} : i64 to f64 + %3 = sitofp %arg1: i64 to f64 + return +} + +// Checking conversion of integer vectors to floating point vector types. +// CHECK-LABEL: @sitofp_vector +func @sitofp_vector(%arg0 : vector<2xi16>, %arg1 : vector<2xi32>, %arg2 : vector<2xi64>) { +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi16> to vector<2xf32> + %0 = sitofp %arg0: vector<2xi16> to vector<2xf32> +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi16> to vector<2xf64> + %1 = sitofp %arg0: vector<2xi16> to vector<2xf64> +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi32> to vector<2xf32> + %2 = sitofp %arg1: vector<2xi32> to vector<2xf32> +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi32> to vector<2xf64> + %3 = sitofp %arg1: vector<2xi32> to vector<2xf64> +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi64> to vector<2xf32> + %4 = sitofp %arg2: vector<2xi64> to vector<2xf32> +// CHECK-NEXT: = llvm.sitofp {{.*}} : vector<2xi64> to vector<2xf64> + %5 = sitofp %arg2: vector<2xi64> to vector<2xf64> + return +} + +// Checking conversion of unsigned integer types to floating point. +// CHECK-LABEL: @uitofp +func @uitofp(%arg0 : i32, %arg1 : i64) { +// CHECK-NEXT: = llvm.uitofp {{.*}} : i32 to f32 + %0 = uitofp %arg0: i32 to f32 +// CHECK-NEXT: = llvm.uitofp {{.*}} : i32 to f64 + %1 = uitofp %arg0: i32 to f64 +// CHECK-NEXT: = llvm.uitofp {{.*}} : i64 to f32 + %2 = uitofp %arg1: i64 to f32 +// CHECK-NEXT: = llvm.uitofp {{.*}} : i64 to f64 + %3 = uitofp %arg1: i64 to f64 + return } +// Checking conversion of integer types to floating point. +// CHECK-LABEL: @fpext +func @fpext(%arg0 : f16, %arg1 : f32) { +// CHECK-NEXT: = llvm.fpext {{.*}} : f16 to f32 + %0 = fpext %arg0: f16 to f32 +// CHECK-NEXT: = llvm.fpext {{.*}} : f16 to f64 + %1 = fpext %arg0: f16 to f64 +// CHECK-NEXT: = llvm.fpext {{.*}} : f32 to f64 + %2 = fpext %arg1: f32 to f64 + return +} + +// Checking conversion of integer types to floating point. +// CHECK-LABEL: @fpext +func @fpext_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>) { +// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf16> to vector<2xf32> + %0 = fpext %arg0: vector<2xf16> to vector<2xf32> +// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf16> to vector<2xf64> + %1 = fpext %arg0: vector<2xf16> to vector<2xf64> +// CHECK-NEXT: = llvm.fpext {{.*}} : vector<2xf32> to vector<2xf64> + %2 = fpext %arg1: vector<2xf32> to vector<2xf64> + return +} + +// Checking conversion of floating point to integer types. +// CHECK-LABEL: @fptosi +func @fptosi(%arg0 : f32, %arg1 : f64) { +// CHECK-NEXT: = llvm.fptosi {{.*}} : f32 to i32 + %0 = fptosi %arg0: f32 to i32 +// CHECK-NEXT: = llvm.fptosi {{.*}} : f32 to i64 + %1 = fptosi %arg0: f32 to i64 +// CHECK-NEXT: = llvm.fptosi {{.*}} : f64 to i32 + %2 = fptosi %arg1: f64 to i32 +// CHECK-NEXT: = llvm.fptosi {{.*}} : f64 to i64 + %3 = fptosi %arg1: f64 to i64 + return +} + +// Checking conversion of floating point vectors to integer vector types. +// CHECK-LABEL: @fptosi_vector +func @fptosi_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>, %arg2 : vector<2xf64>) { +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf16> to vector<2xi32> + %0 = fptosi %arg0: vector<2xf16> to vector<2xi32> +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf16> to vector<2xi64> + %1 = fptosi %arg0: vector<2xf16> to vector<2xi64> +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf32> to vector<2xi32> + %2 = fptosi %arg1: vector<2xf32> to vector<2xi32> +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf32> to vector<2xi64> + %3 = fptosi %arg1: vector<2xf32> to vector<2xi64> +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf64> to vector<2xi32> + %4 = fptosi %arg2: vector<2xf64> to vector<2xi32> +// CHECK-NEXT: = llvm.fptosi {{.*}} : vector<2xf64> to vector<2xi64> + %5 = fptosi %arg2: vector<2xf64> to vector<2xi64> + return +} + +// Checking conversion of floating point to integer types. +// CHECK-LABEL: @fptoui +func @fptoui(%arg0 : f32, %arg1 : f64) { +// CHECK-NEXT: = llvm.fptoui {{.*}} : f32 to i32 + %0 = fptoui %arg0: f32 to i32 +// CHECK-NEXT: = llvm.fptoui {{.*}} : f32 to i64 + %1 = fptoui %arg0: f32 to i64 +// CHECK-NEXT: = llvm.fptoui {{.*}} : f64 to i32 + %2 = fptoui %arg1: f64 to i32 +// CHECK-NEXT: = llvm.fptoui {{.*}} : f64 to i64 + %3 = fptoui %arg1: f64 to i64 + return +} + +// Checking conversion of floating point vectors to integer vector types. +// CHECK-LABEL: @fptoui_vector +func @fptoui_vector(%arg0 : vector<2xf16>, %arg1 : vector<2xf32>, %arg2 : vector<2xf64>) { +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf16> to vector<2xi32> + %0 = fptoui %arg0: vector<2xf16> to vector<2xi32> +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf16> to vector<2xi64> + %1 = fptoui %arg0: vector<2xf16> to vector<2xi64> +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf32> to vector<2xi32> + %2 = fptoui %arg1: vector<2xf32> to vector<2xi32> +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf32> to vector<2xi64> + %3 = fptoui %arg1: vector<2xf32> to vector<2xi64> +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf64> to vector<2xi32> + %4 = fptoui %arg2: vector<2xf64> to vector<2xi32> +// CHECK-NEXT: = llvm.fptoui {{.*}} : vector<2xf64> to vector<2xi64> + %5 = fptoui %arg2: vector<2xf64> to vector<2xi64> + return +} + +// Checking conversion of integer vectors to floating point vector types. +// CHECK-LABEL: @uitofp_vector +func @uitofp_vector(%arg0 : vector<2xi16>, %arg1 : vector<2xi32>, %arg2 : vector<2xi64>) { +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi16> to vector<2xf32> + %0 = uitofp %arg0: vector<2xi16> to vector<2xf32> +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi16> to vector<2xf64> + %1 = uitofp %arg0: vector<2xi16> to vector<2xf64> +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi32> to vector<2xf32> + %2 = uitofp %arg1: vector<2xi32> to vector<2xf32> +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi32> to vector<2xf64> + %3 = uitofp %arg1: vector<2xi32> to vector<2xf64> +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi64> to vector<2xf32> + %4 = uitofp %arg2: vector<2xi64> to vector<2xf32> +// CHECK-NEXT: = llvm.uitofp {{.*}} : vector<2xi64> to vector<2xf64> + %5 = uitofp %arg2: vector<2xi64> to vector<2xf64> + return +} + +// Checking conversion of integer types to floating point. +// CHECK-LABEL: @fptrunc +func @fptrunc(%arg0 : f32, %arg1 : f64) { +// CHECK-NEXT: = llvm.fptrunc {{.*}} : f32 to f16 + %0 = fptrunc %arg0: f32 to f16 +// CHECK-NEXT: = llvm.fptrunc {{.*}} : f64 to f16 + %1 = fptrunc %arg1: f64 to f16 +// CHECK-NEXT: = llvm.fptrunc {{.*}} : f64 to f32 + %2 = fptrunc %arg1: f64 to f32 + return +} + +// Checking conversion of integer types to floating point. +// CHECK-LABEL: @fptrunc +func @fptrunc_vector(%arg0 : vector<2xf32>, %arg1 : vector<2xf64>) { +// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf32> to vector<2xf16> + %0 = fptrunc %arg0: vector<2xf32> to vector<2xf16> +// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf64> to vector<2xf16> + %1 = fptrunc %arg1: vector<2xf64> to vector<2xf16> +// CHECK-NEXT: = llvm.fptrunc {{.*}} : vector<2xf64> to vector<2xf32> + %2 = fptrunc %arg1: vector<2xf64> to vector<2xf32> + return +} + +// Check sign and zero extension and truncation of integers. +// CHECK-LABEL: @integer_extension_and_truncation +func @integer_extension_and_truncation(%arg0 : i3) { +// CHECK-NEXT: = llvm.sext %arg0 : i3 to i6 + %0 = sexti %arg0 : i3 to i6 +// CHECK-NEXT: = llvm.zext %arg0 : i3 to i6 + %1 = zexti %arg0 : i3 to i6 +// CHECK-NEXT: = llvm.trunc %arg0 : i3 to i2 + %2 = trunci %arg0 : i3 to i2 + return +} + +// CHECK-LABEL: @dfs_block_order +func @dfs_block_order(%arg0: i32) -> (i32) { +// CHECK-NEXT: %[[CST:.*]] = llvm.mlir.constant(42 : i32) : i32 + %0 = constant 42 : i32 +// CHECK-NEXT: llvm.br ^bb2 + br ^bb2 + +// CHECK-NEXT: ^bb1: +// CHECK-NEXT: %[[ADD:.*]] = llvm.add %arg0, %[[CST]] : i32 +// CHECK-NEXT: llvm.return %[[ADD]] : i32 +^bb1: + %2 = addi %arg0, %0 : i32 + return %2 : i32 + +// CHECK-NEXT: ^bb2: +^bb2: +// CHECK-NEXT: llvm.br ^bb1 + br ^bb1 +} + +// CHECK-LABEL: func @fcmp(%arg0: f32, %arg1: f32) { +func @fcmp(f32, f32) -> () { +^bb0(%arg0: f32, %arg1: f32): + // CHECK: llvm.fcmp "oeq" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ogt" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "oge" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "olt" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ole" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "one" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ord" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ueq" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ugt" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "uge" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ult" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "ule" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "une" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.fcmp "uno" %arg0, %arg1 : f32 + // CHECK-NEXT: llvm.return + %1 = cmpf oeq, %arg0, %arg1 : f32 + %2 = cmpf ogt, %arg0, %arg1 : f32 + %3 = cmpf oge, %arg0, %arg1 : f32 + %4 = cmpf olt, %arg0, %arg1 : f32 + %5 = cmpf ole, %arg0, %arg1 : f32 + %6 = cmpf one, %arg0, %arg1 : f32 + %7 = cmpf ord, %arg0, %arg1 : f32 + %8 = cmpf ueq, %arg0, %arg1 : f32 + %9 = cmpf ugt, %arg0, %arg1 : f32 + %10 = cmpf uge, %arg0, %arg1 : f32 + %11 = cmpf ult, %arg0, %arg1 : f32 + %12 = cmpf ule, %arg0, %arg1 : f32 + %13 = cmpf une, %arg0, %arg1 : f32 + %14 = cmpf uno, %arg0, %arg1 : f32 + + return +} + +// CHECK-LABEL: @splat +// CHECK-SAME: %[[A:arg[0-9]+]]: vector<4xf32> +// CHECK-SAME: %[[ELT:arg[0-9]+]]: f32 +func @splat(%a: vector<4xf32>, %b: f32) -> vector<4xf32> { + %vb = splat %b : vector<4xf32> + %r = mulf %a, %vb : vector<4xf32> + return %r : vector<4xf32> +} +// CHECK-NEXT: %[[UNDEF:[0-9]+]] = llvm.mlir.undef : vector<4xf32> +// CHECK-NEXT: %[[ZERO:[0-9]+]] = llvm.mlir.constant(0 : i32) : i32 +// CHECK-NEXT: %[[V:[0-9]+]] = llvm.insertelement %[[ELT]], %[[UNDEF]][%[[ZERO]] : i32] : vector<4xf32> +// CHECK-NEXT: %[[SPLAT:[0-9]+]] = llvm.shufflevector %[[V]], %[[UNDEF]] [0 : i32, 0 : i32, 0 : i32, 0 : i32] +// CHECK-NEXT: %[[SCALE:[0-9]+]] = llvm.fmul %[[A]], %[[SPLAT]] : vector<4xf32> +// CHECK-NEXT: llvm.return %[[SCALE]] : vector<4xf32> + +// ----- + +// CHECK-LABEL: func @atomic_rmw +func @atomic_rmw(%I : memref<10xi32>, %ival : i32, %F : memref<10xf32>, %fval : f32, %i : index) { + atomic_rmw assign %fval, %F[%i] : (f32, memref<10xf32>) -> f32 + // CHECK: llvm.atomicrmw xchg %{{.*}}, %{{.*}} acq_rel + atomic_rmw addi %ival, %I[%i] : (i32, memref<10xi32>) -> i32 + // CHECK: llvm.atomicrmw add %{{.*}}, %{{.*}} acq_rel + atomic_rmw maxs %ival, %I[%i] : (i32, memref<10xi32>) -> i32 + // CHECK: llvm.atomicrmw max %{{.*}}, %{{.*}} acq_rel + atomic_rmw mins %ival, %I[%i] : (i32, memref<10xi32>) -> i32 + // CHECK: llvm.atomicrmw min %{{.*}}, %{{.*}} acq_rel + atomic_rmw maxu %ival, %I[%i] : (i32, memref<10xi32>) -> i32 + // CHECK: llvm.atomicrmw umax %{{.*}}, %{{.*}} acq_rel + atomic_rmw minu %ival, %I[%i] : (i32, memref<10xi32>) -> i32 + // CHECK: llvm.atomicrmw umin %{{.*}}, %{{.*}} acq_rel + atomic_rmw addf %fval, %F[%i] : (f32, memref<10xf32>) -> f32 + // CHECK: llvm.atomicrmw fadd %{{.*}}, %{{.*}} acq_rel + return +} + +// ----- + +// CHECK-LABEL: func @generic_atomic_rmw +func @generic_atomic_rmw(%I : memref<10xi32>, %i : index) -> i32 { + %x = generic_atomic_rmw %I[%i] : memref<10xi32> { + ^bb0(%old_value : i32): + %c1 = constant 1 : i32 + atomic_yield %c1 : i32 + } + // CHECK: [[init:%.*]] = llvm.load %{{.*}} : !llvm.ptr + // CHECK-NEXT: llvm.br ^bb1([[init]] : i32) + // CHECK-NEXT: ^bb1([[loaded:%.*]]: i32): + // CHECK-NEXT: [[c1:%.*]] = llvm.mlir.constant(1 : i32) + // CHECK-NEXT: [[pair:%.*]] = llvm.cmpxchg %{{.*}}, [[loaded]], [[c1]] + // CHECK-SAME: acq_rel monotonic : i32 + // CHECK-NEXT: [[new:%.*]] = llvm.extractvalue [[pair]][0] + // CHECK-NEXT: [[ok:%.*]] = llvm.extractvalue [[pair]][1] + // CHECK-NEXT: llvm.cond_br [[ok]], ^bb2, ^bb1([[new]] : i32) + // CHECK-NEXT: ^bb2: + %c2 = constant 2 : i32 + %add = addi %c2, %x : i32 + return %add : i32 + // CHECK-NEXT: [[c2:%.*]] = llvm.mlir.constant(2 : i32) + // CHECK-NEXT: [[add:%.*]] = llvm.add [[c2]], [[new]] : i32 + // CHECK-NEXT: llvm.return [[add]] +} + +// ----- + +// CHECK-LABEL: func @rank_of_unranked +// CHECK32-LABEL: func @rank_of_unranked +func @rank_of_unranked(%unranked: memref<*xi32>) { + %rank = rank %unranked : memref<*xi32> + return +} +// CHECK-NEXT: llvm.mlir.undef +// CHECK-NEXT: llvm.insertvalue +// CHECK-NEXT: llvm.insertvalue +// CHECK-NEXT: llvm.extractvalue %{{.*}}[0] : !llvm.struct<(i64, ptr)> +// CHECK32: llvm.extractvalue %{{.*}}[0] : !llvm.struct<(i32, ptr)> + +// CHECK-LABEL: func @rank_of_ranked +// CHECK32-LABEL: func @rank_of_ranked +func @rank_of_ranked(%ranked: memref) { + %rank = rank %ranked : memref + return +} +// CHECK: llvm.mlir.constant(1 : index) : i64 +// CHECK32: llvm.mlir.constant(1 : index) : i32 + + // ----- // CHECK-LABEL: func @log1p( @@ -147,99 +997,6 @@ // ----- -// CHECK-LABEL: func @transpose -// CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.extractvalue {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[3, 2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.extractvalue {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.extractvalue {{.*}}[3, 2] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -// CHECK: llvm.insertvalue {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<3 x i64>, array<3 x i64>)> -func @transpose(%arg0: memref) { - %0 = memref.transpose %arg0 (i, j, k) -> (k, i, j) : memref to memref (d2 * s1 + s0 + d0 * s2 + d1)>> - return -} - -// ----- - -// CHECK: llvm.mlir.global external @gv0() : !llvm.array<2 x f32> -memref.global @gv0 : memref<2xf32> = uninitialized - -// CHECK: llvm.mlir.global private @gv1() : !llvm.array<2 x f32> -memref.global "private" @gv1 : memref<2xf32> - -// CHECK: llvm.mlir.global external @gv2(dense<{{\[\[}}0.000000e+00, 1.000000e+00, 2.000000e+00], [3.000000e+00, 4.000000e+00, 5.000000e+00]]> : tensor<2x3xf32>) : !llvm.array<2 x array<3 x f32>> -memref.global @gv2 : memref<2x3xf32> = dense<[[0.0, 1.0, 2.0], [3.0, 4.0, 5.0]]> - -// Test 1D memref. -// CHECK-LABEL: func @get_gv0_memref -func @get_gv0_memref() { - %0 = memref.get_global @gv0 : memref<2xf32> - // CHECK: %[[DIM:.*]] = llvm.mlir.constant(2 : index) : i64 - // CHECK: %[[STRIDE:.*]] = llvm.mlir.constant(1 : index) : i64 - // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv0 : !llvm.ptr> - // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]], %[[ZERO]]] : (!llvm.ptr>, i64, i64) -> !llvm.ptr - // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 - // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: llvm.insertvalue %[[DIM]], {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - // CHECK: llvm.insertvalue %[[STRIDE]], {{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<1 x i64>, array<1 x i64>)> - return -} - -// Test 2D memref. -// CHECK-LABEL: func @get_gv2_memref -func @get_gv2_memref() { - // CHECK: %[[DIM0:.*]] = llvm.mlir.constant(2 : index) : i64 - // CHECK: %[[DIM1:.*]] = llvm.mlir.constant(3 : index) : i64 - // CHECK: %[[STRIDE1:.*]] = llvm.mlir.constant(1 : index) : i64 - // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv2 : !llvm.ptr>> - // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]], %[[ZERO]], %[[ZERO]]] : (!llvm.ptr>>, i64, i64, i64) -> !llvm.ptr - // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 - // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DIM0]], {{.*}}[3, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DIM1]], {{.*}}[3, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[DIM1]], {{.*}}[4, 0] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - // CHECK: llvm.insertvalue %[[STRIDE1]], {{.*}}[4, 1] : !llvm.struct<(ptr, ptr, i64, array<2 x i64>, array<2 x i64>)> - - %0 = memref.get_global @gv2 : memref<2x3xf32> - return -} - -// Test scalar memref. -// CHECK: llvm.mlir.global external @gv3(1.000000e+00 : f32) : f32 -memref.global @gv3 : memref = dense<1.0> - -// CHECK-LABEL: func @get_gv3_memref -func @get_gv3_memref() { - // CHECK: %[[ADDR:.*]] = llvm.mlir.addressof @gv3 : !llvm.ptr - // CHECK: %[[ZERO:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: %[[GEP:.*]] = llvm.getelementptr %[[ADDR]][%[[ZERO]]] : (!llvm.ptr, i64) -> !llvm.ptr - // CHECK: %[[DEADBEEF:.*]] = llvm.mlir.constant(3735928559 : index) : i64 - // CHECK: %[[DEADBEEFPTR:.*]] = llvm.inttoptr %[[DEADBEEF]] : i64 to !llvm.ptr - // CHECK: llvm.mlir.undef : !llvm.struct<(ptr, ptr, i64)> - // CHECK: llvm.insertvalue %[[DEADBEEFPTR]], {{.*}}[0] : !llvm.struct<(ptr, ptr, i64)> - // CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr, ptr, i64)> - // CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64 - // CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr, ptr, i64)> - %0 = memref.get_global @gv3 : memref - return -} - // This should not trigger an assertion by creating an LLVM::CallOp with a // nullptr result type. diff --git a/mlir/test/Integration/Dialect/Async/CPU/microbench-linalg-async-parallel-for.mlir b/mlir/test/Integration/Dialect/Async/CPU/microbench-linalg-async-parallel-for.mlir --- a/mlir/test/Integration/Dialect/Async/CPU/microbench-linalg-async-parallel-for.mlir +++ b/mlir/test/Integration/Dialect/Async/CPU/microbench-linalg-async-parallel-for.mlir @@ -8,6 +8,7 @@ // RUN: -convert-scf-to-std \ // RUN: -std-expand \ // RUN: -convert-vector-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O3 \ @@ -20,6 +21,7 @@ // RUN: -convert-linalg-to-loops \ // RUN: -convert-scf-to-std \ // RUN: -convert-vector-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O3 \ diff --git a/mlir/test/Integration/Dialect/Async/CPU/microbench-scf-async-parallel-for.mlir b/mlir/test/Integration/Dialect/Async/CPU/microbench-scf-async-parallel-for.mlir --- a/mlir/test/Integration/Dialect/Async/CPU/microbench-scf-async-parallel-for.mlir +++ b/mlir/test/Integration/Dialect/Async/CPU/microbench-scf-async-parallel-for.mlir @@ -8,7 +8,8 @@ // RUN: -convert-scf-to-std \ // RUN: -std-expand \ // RUN: -convert-vector-to-llvm \ -// RUN: -convert-std-to-llvm -print-ir-after-all \ +// RUN: -convert-memref-to-llvm \ +// RUN: -convert-std-to-llvm -print-ir-after-all \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O3 \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ @@ -26,6 +27,7 @@ // RUN: -convert-scf-to-std \ // RUN: -std-expand \ // RUN: -convert-vector-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O3 \ @@ -38,6 +40,7 @@ // RUN: -convert-linalg-to-loops \ // RUN: -convert-scf-to-std \ // RUN: -convert-vector-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O3 \ diff --git a/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-1d.mlir b/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-1d.mlir --- a/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-1d.mlir +++ b/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-1d.mlir @@ -4,6 +4,7 @@ // RUN: -async-runtime-ref-counting-opt \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ @@ -16,6 +17,7 @@ // RUN: -async-runtime-policy-based-ref-counting \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ @@ -31,6 +33,7 @@ // RUN: -async-runtime-ref-counting-opt \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ diff --git a/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-2d.mlir b/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-2d.mlir --- a/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-2d.mlir +++ b/mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-2d.mlir @@ -4,6 +4,7 @@ // RUN: -async-runtime-ref-counting-opt \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ @@ -16,6 +17,7 @@ // RUN: -async-runtime-policy-based-ref-counting \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ @@ -31,6 +33,7 @@ // RUN: -async-runtime-ref-counting-opt \ // RUN: -convert-async-to-llvm \ // RUN: -convert-scf-to-std \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void -O0 \ diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir @@ -5,7 +5,7 @@ // RUN: mlir-opt -test-linalg-codegen-strategy="anchor-op=linalg.copy register-tile-sizes=4,32 vectorize" | \ // RUN: mlir-opt -canonicalize -convert-vector-to-scf -lower-affine -convert-linalg-to-loops | \ -// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // Activate to dump assembly // R_UN: -dump-object-file -object-filename=/tmp/a.o \ diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major_as_row_major.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major_as_row_major.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major_as_row_major.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major_as_row_major.mlir @@ -8,7 +8,7 @@ // R_UN: mlir-opt -test-linalg-codegen-strategy="anchor-op=linalg.copy register-tile-sizes=4,16 vectorize" | \ // RUN: mlir-opt -canonicalize -convert-vector-to-scf -lower-affine -convert-linalg-to-loops | \ -// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // Activate to dump assembly // R_UN: -dump-object-file -object-filename=/tmp/a.o \ diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir @@ -5,7 +5,7 @@ // RUN: mlir-opt -test-linalg-codegen-strategy="anchor-func=matmul anchor-op=linalg.copy register-tile-sizes=4,32 vectorize" | \ // RUN: mlir-opt -canonicalize -convert-vector-to-scf -lower-affine -convert-linalg-to-loops | \ -// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm -mlir-disable-threading | \ +// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm -mlir-disable-threading | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // Activate to dump assembly // R_UN: -dump-object-file -object-filename=/tmp/a.o \ diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-comprehensive-bufferize.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-comprehensive-bufferize.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-comprehensive-bufferize.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-comprehensive-bufferize.mlir @@ -1,6 +1,6 @@ // RUN: mlir-opt %s -canonicalize -cse -linalg-comprehensive-module-bufferize |\ // RUN: mlir-opt -convert-vector-to-scf -lower-affine -convert-linalg-to-loops |\ -// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt -canonicalize -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -O3 -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext |\ diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-ncw-filter-wcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-ncw-filter-wcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-ncw-filter-wcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-ncw-filter-wcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-nwc-filter-wcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-nwc-filter-wcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-nwc-filter-wcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-nwc-filter-wcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-ncw-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-ncw-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-ncw-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-ncw-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-nwc-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-nwc-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-nwc-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-nwc-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,2" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,2" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nchw-filter-hwcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nchw-filter-hwcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nchw-filter-hwcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nchw-filter-hwcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,0,4,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,0,4,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nhwc-filter-hwcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nhwc-filter-hwcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nhwc-filter-hwcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nhwc-filter-hwcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,3,3,2" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,3,3,2" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nchw-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nchw-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nchw-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nchw-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,0,4,4" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,0,4,4" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nhwc-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nhwc-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nhwc-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nhwc-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,3,3,2" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,3,3,2" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,2,2" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=2,2,2" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ncdhw-filter-dhwcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ncdhw-filter-dhwcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ncdhw-filter-dhwcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ncdhw-filter-dhwcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,5,5,5" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,5,5,5" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ndhwc-filter-dhwcf-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ndhwc-filter-dhwcf-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ndhwc-filter-dhwcf-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ndhwc-filter-dhwcf-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,5,5,5" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,5,5,5" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ncdhw-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ncdhw-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ncdhw-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ncdhw-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,5,5,5" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,0,5,5,5" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ndhwc-call.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ndhwc-call.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ndhwc-call.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ndhwc-call.mlir @@ -1,21 +1,21 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,5,5,5" -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s -// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s // RUN: mlir-opt %s -linalg-tile="linalg-tile-sizes=0,5,5,5" \ -// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -test-conv-vectorization="tile-sizes=1,1,1,1,1,3,3,3,3" -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-vector-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-elementwise.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-elementwise.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-elementwise.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-elementwise.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-elementwise-to-linalg -std-bufferize -tensor-constant-bufferize -linalg-bufferize -tensor-bufferize -func-bufferize -convert-linalg-to-loops -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-elementwise-to-linalg -std-bufferize -tensor-constant-bufferize -linalg-bufferize -tensor-bufferize -func-bufferize -convert-linalg-to-loops -convert-linalg-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-padtensor.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-padtensor.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-padtensor.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-padtensor.mlir @@ -1,7 +1,7 @@ // RUN: mlir-opt %s -linalg-bufferize -std-bufferize \ // RUN: -tensor-constant-bufferize -tensor-bufferize -func-bufferize \ // RUN: -finalizing-bufferize \ -// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert-multiple-uses.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert-multiple-uses.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert-multiple-uses.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert-multiple-uses.mlir @@ -1,7 +1,7 @@ // RUN: mlir-opt %s -linalg-bufferize -std-bufferize \ // RUN: -tensor-constant-bufferize -tensor-bufferize -func-bufferize \ // RUN: -finalizing-bufferize \ -// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-subtensor-insert.mlir @@ -1,7 +1,7 @@ // RUN: mlir-opt %s -linalg-bufferize -std-bufferize \ // RUN: -tensor-constant-bufferize -tensor-bufferize -func-bufferize \ // RUN: -finalizing-bufferize \ -// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-e2e.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-e2e.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-e2e.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-e2e.mlir @@ -1,6 +1,6 @@ // RUN: mlir-opt %s -tensor-constant-bufferize -std-bufferize -linalg-bufferize \ // RUN: -tensor-bufferize -func-bufferize -finalizing-bufferize -convert-linalg-to-loops \ -// RUN: -convert-linalg-to-llvm -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-matmul.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-matmul.mlir --- a/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-matmul.mlir +++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-tensor-matmul.mlir @@ -1,6 +1,6 @@ // RUN: mlir-opt %s -linalg-bufferize -std-bufferize -tensor-constant-bufferize \ // RUN: -tensor-bufferize -func-bufferize -finalizing-bufferize -convert-linalg-to-loops -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s @@ -9,7 +9,7 @@ // RUN: -scf-bufferize -std-bufferize -tensor-constant-bufferize -tensor-bufferize \ // RUN: -func-bufferize \ // RUN: -finalizing-bufferize -convert-linalg-to-loops -convert-scf-to-std -convert-scf-to-std \ -// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std -convert-std-to-llvm | \ +// RUN: -convert-linalg-to-llvm -lower-affine -convert-scf-to-std --convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/dense_output.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/dense_output.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/dense_output.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/dense_output.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/test.mtx" \ // RUN: TENSOR1="%mlir_integration_test_dir/data/zero.mtx" \ // RUN: mlir-cpu-runner \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_flatten.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_flatten.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_flatten.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_flatten.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/test.tns" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_matvec.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_matvec.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_matvec.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_matvec.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/wide.mtx" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_mttkrp.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_mttkrp.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_mttkrp.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_mttkrp.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/mttkrp_b.tns" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_out_simple.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_out_simple.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_out_simple.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_out_simple.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/test.mtx" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sampled_matmul.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sampled_matmul.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sampled_matmul.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sampled_matmul.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/test.mtx" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_spmm.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_spmm.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_spmm.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_spmm.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/wide.mtx" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sum.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sum.mlir --- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sum.mlir +++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_sum.mlir @@ -3,7 +3,7 @@ // RUN: --convert-linalg-to-loops --convert-vector-to-scf --convert-scf-to-std \ // RUN: --func-bufferize --tensor-constant-bufferize --tensor-bufferize \ // RUN: --std-bufferize --finalizing-bufferize \ -// RUN: --convert-vector-to-llvm --convert-std-to-llvm | \ +// RUN: --convert-vector-to-llvm --convert-memref-to-llvm --convert-std-to-llvm | \ // RUN: TENSOR0="%mlir_integration_test_dir/data/test.mtx" \ // RUN: mlir-cpu-runner \ // RUN: -e entry -entry-point-result=void \ diff --git a/mlir/test/Integration/Dialect/Standard/CPU/test-ceil-floor-pos-neg.mlir b/mlir/test/Integration/Dialect/Standard/CPU/test-ceil-floor-pos-neg.mlir --- a/mlir/test/Integration/Dialect/Standard/CPU/test-ceil-floor-pos-neg.mlir +++ b/mlir/test/Integration/Dialect/Standard/CPU/test-ceil-floor-pos-neg.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -std-expand -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -std-expand -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir b/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir --- a/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir +++ b/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext | FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-saxpy-jagged-matvec.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-saxpy-jagged-matvec.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-saxpy-jagged-matvec.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-saxpy-jagged-matvec.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-1d.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-1d.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-1d.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-1d.mlir @@ -1,19 +1,19 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir @@ -1,19 +1,19 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-3d.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-3d.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-3d.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-3d.mlir @@ -1,19 +1,19 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf='full-unroll=true lower-permutation-maps=true' -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir @@ -1,9 +1,9 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf=full-unroll=true -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf=full-unroll=true -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir @@ -1,9 +1,9 @@ -// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext,%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s -// RUN: mlir-opt %s -convert-vector-to-scf=full-unroll=true -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-vector-to-scf=full-unroll=true -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext,%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e entry -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_c_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir --- a/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir +++ b/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir @@ -1,11 +1,11 @@ // RUN: mlir-opt %s -test-vector-to-forloop -convert-vector-to-scf \ -// RUN: -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | \ +// RUN: -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext | \ // RUN: FileCheck %s // RUN: mlir-opt %s -convert-vector-to-scf -lower-affine \ -// RUN: -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e main \ +// RUN: -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e main \ // RUN: -entry-point-result=void \ // RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext | \ // RUN: FileCheck %s diff --git a/mlir/test/lib/Conversion/StandardToLLVM/CMakeLists.txt b/mlir/test/lib/Conversion/StandardToLLVM/CMakeLists.txt --- a/mlir/test/lib/Conversion/StandardToLLVM/CMakeLists.txt +++ b/mlir/test/lib/Conversion/StandardToLLVM/CMakeLists.txt @@ -5,6 +5,7 @@ EXCLUDE_FROM_LIBMLIR LINK_LIBS PUBLIC + MLIRLLVMCommonConversion MLIRLLVMIR MLIRLLVMIRTransforms MLIRPass diff --git a/mlir/test/lib/Conversion/StandardToLLVM/TestConvertCallOp.cpp b/mlir/test/lib/Conversion/StandardToLLVM/TestConvertCallOp.cpp --- a/mlir/test/lib/Conversion/StandardToLLVM/TestConvertCallOp.cpp +++ b/mlir/test/lib/Conversion/StandardToLLVM/TestConvertCallOp.cpp @@ -8,8 +8,8 @@ #include "TestDialect.h" #include "TestTypes.h" +#include "mlir/Conversion/LLVMCommon/Pattern.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVM.h" -#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Dialect/LLVMIR/LLVMDialect.h" #include "mlir/Dialect/StandardOps/IR/Ops.h" #include "mlir/Pass/Pass.h" diff --git a/mlir/test/mlir-cpu-runner/async-value.mlir b/mlir/test/mlir-cpu-runner/async-value.mlir --- a/mlir/test/mlir-cpu-runner/async-value.mlir +++ b/mlir/test/mlir-cpu-runner/async-value.mlir @@ -3,6 +3,7 @@ // RUN: -async-runtime-ref-counting-opt \ // RUN: -convert-async-to-llvm \ // RUN: -convert-vector-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e main -entry-point-result=void -O0 \ diff --git a/mlir/test/mlir-cpu-runner/async.mlir b/mlir/test/mlir-cpu-runner/async.mlir --- a/mlir/test/mlir-cpu-runner/async.mlir +++ b/mlir/test/mlir-cpu-runner/async.mlir @@ -5,6 +5,7 @@ // RUN: -convert-linalg-to-loops \ // RUN: -convert-scf-to-std \ // RUN: -convert-linalg-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm \ // RUN: | mlir-cpu-runner \ // RUN: -e main -entry-point-result=void -O0 \ diff --git a/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir b/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir --- a/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir +++ b/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-std-to-llvm='use-bare-ptr-memref-call-conv=1' | mlir-cpu-runner -shared-libs=%linalg_test_lib_dir/libmlir_c_runner_utils%shlibext -entry-point-result=void | FileCheck %s +// RUN: mlir-opt %s -convert-scf-to-std -convert-memref-to-llvm -convert-std-to-llvm='use-bare-ptr-memref-call-conv=1' | mlir-cpu-runner -shared-libs=%linalg_test_lib_dir/libmlir_c_runner_utils%shlibext -entry-point-result=void | FileCheck %s // Verify bare pointer memref calling convention. `simple_add1_add2_test` // gets two 2xf32 memrefs, adds 1.0f to the first one and 2.0f to the second diff --git a/mlir/test/mlir-cpu-runner/copy.mlir b/mlir/test/mlir-cpu-runner/copy.mlir --- a/mlir/test/mlir-cpu-runner/copy.mlir +++ b/mlir/test/mlir-cpu-runner/copy.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-std-to-llvm \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-memref-to-llvm -convert-std-to-llvm \ // RUN: | mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/mlir-cpu-runner/global_memref.mlir b/mlir/test/mlir-cpu-runner/global_memref.mlir --- a/mlir/test/mlir-cpu-runner/global_memref.mlir +++ b/mlir/test/mlir-cpu-runner/global_memref.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-std-to-llvm | mlir-cpu-runner -e main -entry-point-result=void -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext | FileCheck %s +// RUN: mlir-opt %s -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e main -entry-point-result=void -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext | FileCheck %s func private @print_memref_f32(memref<*xf32>) attributes { llvm.emit_c_interface } func private @print_memref_i32(memref<*xi32>) attributes { llvm.emit_c_interface } diff --git a/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir b/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir --- a/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir +++ b/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -convert-std-to-llvm \ +// RUN: mlir-opt %s -convert-scf-to-std -convert-memref-to-llvm -convert-std-to-llvm \ // RUN: | mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/mlir-cpu-runner/memref_reshape.mlir b/mlir/test/mlir-cpu-runner/memref_reshape.mlir --- a/mlir/test/mlir-cpu-runner/memref_reshape.mlir +++ b/mlir/test/mlir-cpu-runner/memref_reshape.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt %s -convert-scf-to-std -std-expand -convert-std-to-llvm \ +// RUN: mlir-opt %s -convert-scf-to-std -std-expand -convert-memref-to-llvm -convert-std-to-llvm \ // RUN: | mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext \ // RUN: | FileCheck %s diff --git a/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir b/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir --- a/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir +++ b/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir @@ -1,4 +1,4 @@ -// RUN: mlir-opt -convert-linalg-to-loops -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-std-to-llvm %s | mlir-cpu-runner -O3 -e main -entry-point-result=void -shared-libs=%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext | FileCheck %s +// RUN: mlir-opt -convert-linalg-to-loops -lower-affine -convert-scf-to-std -convert-vector-to-llvm -convert-memref-to-llvm -convert-std-to-llvm %s | mlir-cpu-runner -O3 -e main -entry-point-result=void -shared-libs=%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext | FileCheck %s func @main() { %A = memref.alloc() : memref<16x16xf32> diff --git a/mlir/test/mlir-cpu-runner/unranked_memref.mlir b/mlir/test/mlir-cpu-runner/unranked_memref.mlir --- a/mlir/test/mlir-cpu-runner/unranked_memref.mlir +++ b/mlir/test/mlir-cpu-runner/unranked_memref.mlir @@ -1,6 +1,7 @@ // RUN: mlir-opt %s -convert-linalg-to-loops \ // RUN: -convert-scf-to-std \ // RUN: -convert-linalg-to-llvm \ +// RUN: -convert-memref-to-llvm \ // RUN: -convert-std-to-llvm | \ // RUN: mlir-cpu-runner -e main -entry-point-result=void \ // RUN: -shared-libs=%mlir_runner_utils_dir/libmlir_runner_utils%shlibext,%mlir_runner_utils_dir/libmlir_c_runner_utils%shlibext | FileCheck %s diff --git a/mlir/test/mlir-cpu-runner/utils.mlir b/mlir/test/mlir-cpu-runner/utils.mlir --- a/mlir/test/mlir-cpu-runner/utils.mlir +++ b/mlir/test/mlir-cpu-runner/utils.mlir @@ -1,7 +1,7 @@ -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_0d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-0D -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_1d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-1D -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_3d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-3D -// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e vector_splat_2d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-VECTOR-SPLAT-2D +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_0d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-0D +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_1d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-1D +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e print_3d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-3D +// RUN: mlir-opt %s -convert-linalg-to-loops -convert-scf-to-std -convert-linalg-to-llvm -convert-memref-to-llvm -convert-std-to-llvm | mlir-cpu-runner -e vector_splat_2d -entry-point-result=void -shared-libs=%linalg_test_lib_dir/libmlir_runner_utils%shlibext | FileCheck %s --check-prefix=PRINT-VECTOR-SPLAT-2D func @print_0d() { %f = constant 2.00000e+00 : f32 diff --git a/mlir/test/python/execution_engine.py b/mlir/test/python/execution_engine.py --- a/mlir/test/python/execution_engine.py +++ b/mlir/test/python/execution_engine.py @@ -56,7 +56,7 @@ def lowerToLLVM(module): import mlir.conversions - pm = PassManager.parse("convert-std-to-llvm") + pm = PassManager.parse("convert-memref-to-llvm,convert-std-to-llvm") pm.run(module) return module diff --git a/mlir/test/python/integration/dialects/linalg/opsrun.py b/mlir/test/python/integration/dialects/linalg/opsrun.py --- a/mlir/test/python/integration/dialects/linalg/opsrun.py +++ b/mlir/test/python/integration/dialects/linalg/opsrun.py @@ -127,7 +127,7 @@ boilerplate) pm = PassManager.parse("func(convert-linalg-to-loops, lower-affine, " + "convert-scf-to-std), convert-vector-to-llvm," + - "convert-std-to-llvm") + "convert-memref-to-llvm,convert-std-to-llvm") pm.run(mod) return mod diff --git a/mlir/tools/mlir-vulkan-runner/CMakeLists.txt b/mlir/tools/mlir-vulkan-runner/CMakeLists.txt --- a/mlir/tools/mlir-vulkan-runner/CMakeLists.txt +++ b/mlir/tools/mlir-vulkan-runner/CMakeLists.txt @@ -64,6 +64,7 @@ MLIRLLVMCommonConversion MLIRLLVMToLLVMIRTranslation MLIRMemRef + MLIRMemRefToLLVM MLIRParser MLIRSPIRV MLIRSPIRVTransforms diff --git a/mlir/tools/mlir-vulkan-runner/mlir-vulkan-runner.cpp b/mlir/tools/mlir-vulkan-runner/mlir-vulkan-runner.cpp --- a/mlir/tools/mlir-vulkan-runner/mlir-vulkan-runner.cpp +++ b/mlir/tools/mlir-vulkan-runner/mlir-vulkan-runner.cpp @@ -15,6 +15,7 @@ #include "mlir/Conversion/GPUToSPIRV/GPUToSPIRVPass.h" #include "mlir/Conversion/GPUToVulkan/ConvertGPUToVulkanPass.h" #include "mlir/Conversion/LLVMCommon/LoweringOptions.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Conversion/StandardToSPIRV/StandardToSPIRVPass.h" #include "mlir/Dialect/GPU/GPUDialect.h" @@ -50,6 +51,7 @@ passManager.addPass(createConvertGpuLaunchFuncToVulkanLaunchFuncPass()); LowerToLLVMOptions llvmOptions(module.getContext(), DataLayout(module)); llvmOptions.emitCWrappers = true; + passManager.addPass(createMemRefToLLVMPass()); passManager.addPass(createLowerToLLVMPass(llvmOptions)); passManager.addPass(createConvertVulkanLaunchFuncToVulkanCallsPass()); return passManager.run(module); diff --git a/mlir/unittests/ExecutionEngine/CMakeLists.txt b/mlir/unittests/ExecutionEngine/CMakeLists.txt --- a/mlir/unittests/ExecutionEngine/CMakeLists.txt +++ b/mlir/unittests/ExecutionEngine/CMakeLists.txt @@ -7,6 +7,7 @@ PRIVATE MLIRExecutionEngine MLIRLinalgToLLVM + MLIRMemRefToLLVM ${dialect_libs} ) diff --git a/mlir/unittests/ExecutionEngine/Invoke.cpp b/mlir/unittests/ExecutionEngine/Invoke.cpp --- a/mlir/unittests/ExecutionEngine/Invoke.cpp +++ b/mlir/unittests/ExecutionEngine/Invoke.cpp @@ -7,6 +7,7 @@ //===----------------------------------------------------------------------===// #include "mlir/Conversion/LinalgToLLVM/LinalgToLLVM.h" +#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h" #include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h" #include "mlir/Conversion/VectorToLLVM/ConvertVectorToLLVM.h" #include "mlir/Conversion/VectorToSCF/VectorToSCF.h" @@ -39,6 +40,7 @@ /// dialects lowering to LLVM Dialect. static LogicalResult lowerToLLVMDialect(ModuleOp module) { PassManager pm(module.getContext()); + pm.addPass(mlir::createMemRefToLLVMPass()); pm.addPass(mlir::createLowerToLLVMPass()); return pm.run(module); }