[Mlir-commits] [mlir] e231070 - [MLIR] Create memref dialect and move dialect-specific ops from std.
Julian Gross
llvmlistbot at llvm.org
Mon Mar 15 03:17:17 PDT 2021
Author: Julian Gross
Date: 2021-03-15T11:14:09+01:00
New Revision: e2310704d890ad252aeb1ca28b4b84d29514b1d1
URL: https://github.com/llvm/llvm-project/commit/e2310704d890ad252aeb1ca28b4b84d29514b1d1
DIFF: https://github.com/llvm/llvm-project/commit/e2310704d890ad252aeb1ca28b4b84d29514b1d1.diff
LOG: [MLIR] Create memref dialect and move dialect-specific ops from std.
Create the memref dialect and move dialect-specific ops
from std dialect to this dialect.
Moved ops:
AllocOp -> MemRef_AllocOp
AllocaOp -> MemRef_AllocaOp
AssumeAlignmentOp -> MemRef_AssumeAlignmentOp
DeallocOp -> MemRef_DeallocOp
DimOp -> MemRef_DimOp
MemRefCastOp -> MemRef_CastOp
MemRefReinterpretCastOp -> MemRef_ReinterpretCastOp
GetGlobalMemRefOp -> MemRef_GetGlobalOp
GlobalMemRefOp -> MemRef_GlobalOp
LoadOp -> MemRef_LoadOp
PrefetchOp -> MemRef_PrefetchOp
ReshapeOp -> MemRef_ReshapeOp
StoreOp -> MemRef_StoreOp
SubViewOp -> MemRef_SubViewOp
TransposeOp -> MemRef_TransposeOp
TensorLoadOp -> MemRef_TensorLoadOp
TensorStoreOp -> MemRef_TensorStoreOp
TensorToMemRefOp -> MemRef_BufferCastOp
ViewOp -> MemRef_ViewOp
The roadmap to split the memref dialect from std is discussed here:
https://llvm.discourse.group/t/rfc-split-the-memref-dialect-from-std/2667
Differential Revision: https://reviews.llvm.org/D98041
Added:
mlir/docs/Dialects/MemRef.md
mlir/include/mlir/Dialect/MemRef/CMakeLists.txt
mlir/include/mlir/Dialect/MemRef/EDSC/Intrinsics.h
mlir/include/mlir/Dialect/MemRef/IR/CMakeLists.txt
mlir/include/mlir/Dialect/MemRef/IR/MemRef.h
mlir/include/mlir/Dialect/MemRef/IR/MemRefBase.td
mlir/include/mlir/Dialect/MemRef/IR/MemRefOps.td
mlir/lib/Dialect/MemRef/CMakeLists.txt
mlir/lib/Dialect/MemRef/IR/CMakeLists.txt
mlir/lib/Dialect/MemRef/IR/MemRefDialect.cpp
mlir/lib/Dialect/MemRef/IR/MemRefOps.cpp
Modified:
mlir/docs/BufferDeallocationInternals.md
mlir/docs/Bufferization.md
mlir/docs/Dialects/Linalg.md
mlir/docs/Dialects/Standard.md
mlir/docs/Rationale/UsageOfConst.md
mlir/docs/Traits.md
mlir/docs/Tutorials/Toy/Ch-5.md
mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp
mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp
mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp
mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp
mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp
mlir/include/mlir/Conversion/Passes.td
mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h
mlir/include/mlir/Dialect/Affine/Passes.td
mlir/include/mlir/Dialect/CMakeLists.txt
mlir/include/mlir/Dialect/GPU/GPUOps.td
mlir/include/mlir/Dialect/Linalg/EDSC/FoldedIntrinsics.h
mlir/include/mlir/Dialect/Linalg/IR/LinalgSparseOps.td
mlir/include/mlir/Dialect/Linalg/Passes.h
mlir/include/mlir/Dialect/Linalg/Passes.td
mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h
mlir/include/mlir/Dialect/Linalg/Utils/Utils.h
mlir/include/mlir/Dialect/SCF/Passes.td
mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
mlir/include/mlir/Dialect/StandardOps/EDSC/Intrinsics.h
mlir/include/mlir/Dialect/StandardOps/IR/Ops.h
mlir/include/mlir/Dialect/StandardOps/IR/Ops.td
mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.h
mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.td
mlir/include/mlir/Dialect/StandardOps/Utils/Utils.h
mlir/include/mlir/Dialect/Vector/VectorTransforms.h
mlir/include/mlir/IR/OpDefinition.h
mlir/include/mlir/InitAllDialects.h
mlir/include/mlir/Transforms/Bufferize.h
mlir/include/mlir/Transforms/Passes.h
mlir/include/mlir/Transforms/Passes.td
mlir/include/mlir/Transforms/Utils.h
mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp
mlir/lib/Conversion/AffineToStandard/CMakeLists.txt
mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt
mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
mlir/lib/Conversion/LinalgToStandard/CMakeLists.txt
mlir/lib/Conversion/LinalgToStandard/LinalgToStandard.cpp
mlir/lib/Conversion/PassDetail.h
mlir/lib/Conversion/SCFToGPU/CMakeLists.txt
mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp
mlir/lib/Conversion/ShapeToStandard/CMakeLists.txt
mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp
mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt
mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp
mlir/lib/Conversion/StandardToSPIRV/CMakeLists.txt
mlir/lib/Conversion/StandardToSPIRV/LegalizeStandardForSPIRV.cpp
mlir/lib/Conversion/StandardToSPIRV/StandardToSPIRV.cpp
mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt
mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp
mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp
mlir/lib/Conversion/VectorToSCF/CMakeLists.txt
mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
mlir/lib/Dialect/Affine/IR/AffineOps.cpp
mlir/lib/Dialect/Affine/IR/CMakeLists.txt
mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp
mlir/lib/Dialect/Affine/Transforms/CMakeLists.txt
mlir/lib/Dialect/Affine/Transforms/PassDetail.h
mlir/lib/Dialect/CMakeLists.txt
mlir/lib/Dialect/GPU/CMakeLists.txt
mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp
mlir/lib/Dialect/GPU/Transforms/KernelOutlining.cpp
mlir/lib/Dialect/GPU/Transforms/MemoryPromotion.cpp
mlir/lib/Dialect/Linalg/Analysis/CMakeLists.txt
mlir/lib/Dialect/Linalg/Analysis/DependenceAnalysis.cpp
mlir/lib/Dialect/Linalg/EDSC/CMakeLists.txt
mlir/lib/Dialect/Linalg/IR/CMakeLists.txt
mlir/lib/Dialect/Linalg/IR/LinalgInterfaces.cpp
mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
mlir/lib/Dialect/Linalg/Transforms/Fusion.cpp
mlir/lib/Dialect/Linalg/Transforms/Loops.cpp
mlir/lib/Dialect/Linalg/Transforms/PassDetail.h
mlir/lib/Dialect/Linalg/Transforms/Promotion.cpp
mlir/lib/Dialect/Linalg/Transforms/SparseLowering.cpp
mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp
mlir/lib/Dialect/Linalg/Transforms/Tiling.cpp
mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp
mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
mlir/lib/Dialect/SCF/CMakeLists.txt
mlir/lib/Dialect/SCF/SCF.cpp
mlir/lib/Dialect/SCF/Transforms/Bufferize.cpp
mlir/lib/Dialect/SCF/Transforms/CMakeLists.txt
mlir/lib/Dialect/SCF/Transforms/ParallelLoopFusion.cpp
mlir/lib/Dialect/SCF/Transforms/PassDetail.h
mlir/lib/Dialect/Shape/IR/Shape.cpp
mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
mlir/lib/Dialect/Shape/Transforms/PassDetail.h
mlir/lib/Dialect/StandardOps/CMakeLists.txt
mlir/lib/Dialect/StandardOps/EDSC/Builders.cpp
mlir/lib/Dialect/StandardOps/IR/Ops.cpp
mlir/lib/Dialect/StandardOps/Transforms/Bufferize.cpp
mlir/lib/Dialect/StandardOps/Transforms/CMakeLists.txt
mlir/lib/Dialect/StandardOps/Transforms/ExpandOps.cpp
mlir/lib/Dialect/StandardOps/Transforms/FuncBufferize.cpp
mlir/lib/Dialect/StandardOps/Transforms/PassDetail.h
mlir/lib/Dialect/StandardOps/Transforms/TensorConstantBufferize.cpp
mlir/lib/Dialect/StandardOps/Utils/Utils.cpp
mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
mlir/lib/Dialect/Tensor/Transforms/CMakeLists.txt
mlir/lib/Dialect/Vector/CMakeLists.txt
mlir/lib/Dialect/Vector/VectorOps.cpp
mlir/lib/Dialect/Vector/VectorTransforms.cpp
mlir/lib/Transforms/BufferDeallocation.cpp
mlir/lib/Transforms/BufferOptimizations.cpp
mlir/lib/Transforms/BufferResultsToOutParams.cpp
mlir/lib/Transforms/Bufferize.cpp
mlir/lib/Transforms/CMakeLists.txt
mlir/lib/Transforms/Canonicalizer.cpp
mlir/lib/Transforms/LoopFusion.cpp
mlir/lib/Transforms/MemRefDataFlowOpt.cpp
mlir/lib/Transforms/NormalizeMemRefs.cpp
mlir/lib/Transforms/PassDetail.h
mlir/lib/Transforms/PipelineDataTransfer.cpp
mlir/lib/Transforms/Utils/CMakeLists.txt
mlir/lib/Transforms/Utils/LoopUtils.cpp
mlir/lib/Transforms/Utils/Utils.cpp
mlir/test/Analysis/test-alias-analysis.mlir
mlir/test/Analysis/test-liveness.mlir
mlir/test/CAPI/ir.c
mlir/test/Conversion/AffineToStandard/lower-affine-gpu.mlir
mlir/test/Conversion/AffineToStandard/lower-affine-to-vector.mlir
mlir/test/Conversion/AffineToStandard/lower-affine.mlir
mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir
mlir/test/Conversion/GPUCommon/memory-attrbution.mlir
mlir/test/Conversion/GPUToSPIRV/load-store.mlir
mlir/test/Conversion/GPUToVulkan/lower-gpu-launch-vulkan-launch.mlir
mlir/test/Conversion/LinalgToVector/linalg-to-vector.mlir
mlir/test/Conversion/SCFToGPU/no_blocks_no_threads.mlir
mlir/test/Conversion/SCFToGPU/parallel_loop.mlir
mlir/test/Conversion/SCFToGPU/step_one.mlir
mlir/test/Conversion/SCFToGPU/step_positive.mlir
mlir/test/Conversion/SCFToSPIRV/for.mlir
mlir/test/Conversion/SCFToSPIRV/if.mlir
mlir/test/Conversion/SPIRVToLLVM/lower-host-to-llvm-calls.mlir
mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir
mlir/test/Conversion/StandardToLLVM/calling-convention.mlir
mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir
mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir
mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir
mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir
mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir
mlir/test/Conversion/StandardToSPIRV/alloc.mlir
mlir/test/Conversion/StandardToSPIRV/legalization.mlir
mlir/test/Conversion/StandardToSPIRV/std-ops-to-spirv.mlir
mlir/test/Conversion/StandardToSPIRV/subview-to-spirv.mlir
mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir
mlir/test/Conversion/VectorToSCF/vector-to-loops.mlir
mlir/test/Dialect/Affine/SuperVectorize/uniform_divergent.mlir
mlir/test/Dialect/Affine/SuperVectorize/vector_utils.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_1d.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_2d.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_3d.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_2d.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_transpose_2d.mlir
mlir/test/Dialect/Affine/SuperVectorize/vectorize_transpose_2d.mlir
mlir/test/Dialect/Affine/affine-data-copy.mlir
mlir/test/Dialect/Affine/affine-loop-invariant-code-motion.mlir
mlir/test/Dialect/Affine/affine-loop-normalize.mlir
mlir/test/Dialect/Affine/canonicalize.mlir
mlir/test/Dialect/Affine/dma-generate.mlir
mlir/test/Dialect/Affine/dma.mlir
mlir/test/Dialect/Affine/invalid.mlir
mlir/test/Dialect/Affine/load-store-invalid.mlir
mlir/test/Dialect/Affine/load-store.mlir
mlir/test/Dialect/Affine/loop-tiling-parametric.mlir
mlir/test/Dialect/Affine/loop-tiling-validity.mlir
mlir/test/Dialect/Affine/loop-tiling.mlir
mlir/test/Dialect/Affine/memref-stride-calculation.mlir
mlir/test/Dialect/Affine/ops.mlir
mlir/test/Dialect/Affine/parallelize.mlir
mlir/test/Dialect/Affine/slicing-utils.mlir
mlir/test/Dialect/Affine/unroll.mlir
mlir/test/Dialect/Async/async-parallel-for.mlir
mlir/test/Dialect/Async/async-to-async-runtime.mlir
mlir/test/Dialect/GPU/all-reduce-max.mlir
mlir/test/Dialect/GPU/all-reduce.mlir
mlir/test/Dialect/GPU/multiple-all-reduce.mlir
mlir/test/Dialect/GPU/ops.mlir
mlir/test/Dialect/GPU/outlining.mlir
mlir/test/Dialect/GPU/promotion.mlir
mlir/test/Dialect/Linalg/affine.mlir
mlir/test/Dialect/Linalg/bufferize.mlir
mlir/test/Dialect/Linalg/canonicalize.mlir
mlir/test/Dialect/Linalg/convert-elementwise-to-linalg.mlir
mlir/test/Dialect/Linalg/fold-affine-min-scf.mlir
mlir/test/Dialect/Linalg/forward-vector-transfers.mlir
mlir/test/Dialect/Linalg/fusion-2-level.mlir
mlir/test/Dialect/Linalg/fusion-indexed-generic.mlir
mlir/test/Dialect/Linalg/fusion-pattern.mlir
mlir/test/Dialect/Linalg/fusion-sequence.mlir
mlir/test/Dialect/Linalg/fusion-tensor-pattern.mlir
mlir/test/Dialect/Linalg/fusion-tensor.mlir
mlir/test/Dialect/Linalg/fusion.mlir
mlir/test/Dialect/Linalg/hoist-padding.mlir
mlir/test/Dialect/Linalg/invalid.mlir
mlir/test/Dialect/Linalg/loops.mlir
mlir/test/Dialect/Linalg/parallel-loops.mlir
mlir/test/Dialect/Linalg/promote.mlir
mlir/test/Dialect/Linalg/promotion_options.mlir
mlir/test/Dialect/Linalg/reshape_fusion.mlir
mlir/test/Dialect/Linalg/roundtrip.mlir
mlir/test/Dialect/Linalg/sparse_1d.mlir
mlir/test/Dialect/Linalg/sparse_2d.mlir
mlir/test/Dialect/Linalg/sparse_3d.mlir
mlir/test/Dialect/Linalg/sparse_lower.mlir
mlir/test/Dialect/Linalg/sparse_nd.mlir
mlir/test/Dialect/Linalg/sparse_storage.mlir
mlir/test/Dialect/Linalg/sparse_vector.mlir
mlir/test/Dialect/Linalg/standard.mlir
mlir/test/Dialect/Linalg/tile-and-distribute.mlir
mlir/test/Dialect/Linalg/tile-and-fuse-tensors.mlir
mlir/test/Dialect/Linalg/tile-conv-padding.mlir
mlir/test/Dialect/Linalg/tile-conv.mlir
mlir/test/Dialect/Linalg/tile-parallel-reduce.mlir
mlir/test/Dialect/Linalg/tile-parallel.mlir
mlir/test/Dialect/Linalg/tile-simple-conv.mlir
mlir/test/Dialect/Linalg/tile-tensors.mlir
mlir/test/Dialect/Linalg/tile.mlir
mlir/test/Dialect/Linalg/transform-patterns.mlir
mlir/test/Dialect/Linalg/vectorization.mlir
mlir/test/Dialect/OpenACC/invalid.mlir
mlir/test/Dialect/OpenACC/ops.mlir
mlir/test/Dialect/SCF/bufferize.mlir
mlir/test/Dialect/SCF/canonicalize.mlir
mlir/test/Dialect/SCF/for-loop-specialization.mlir
mlir/test/Dialect/SCF/loop-unroll.mlir
mlir/test/Dialect/SCF/ops.mlir
mlir/test/Dialect/SCF/parallel-loop-fusion.mlir
mlir/test/Dialect/SCF/parallel-loop-specialization.mlir
mlir/test/Dialect/SCF/parallel-loop-tiling.mlir
mlir/test/Dialect/Shape/bufferize.mlir
mlir/test/Dialect/Standard/bufferize.mlir
mlir/test/Dialect/Standard/canonicalize.mlir
mlir/test/Dialect/Standard/expand-ops.mlir
mlir/test/Dialect/Standard/func-bufferize.mlir
mlir/test/Dialect/Standard/invalid.mlir
mlir/test/Dialect/Standard/ops.mlir
mlir/test/Dialect/Standard/tensor-constant-bufferize.mlir
mlir/test/Dialect/Tensor/bufferize.mlir
mlir/test/Dialect/Tensor/canonicalize.mlir
mlir/test/Dialect/Vector/canonicalize.mlir
mlir/test/Dialect/Vector/vector-contract-matvec-transforms.mlir
mlir/test/Dialect/Vector/vector-transfer-full-partial-split.mlir
mlir/test/Dialect/Vector/vector-transfer-lowering.mlir
mlir/test/Dialect/Vector/vector-transforms.mlir
mlir/test/EDSC/CMakeLists.txt
mlir/test/EDSC/builder-api-test.cpp
mlir/test/Examples/Toy/Ch5/affine-lowering.mlir
mlir/test/Examples/Toy/Ch6/affine-lowering.mlir
mlir/test/Examples/Toy/Ch7/affine-lowering.mlir
mlir/test/IR/core-ops.mlir
mlir/test/IR/invalid-ops.mlir
mlir/test/IR/invalid.mlir
mlir/test/IR/memory-ops.mlir
mlir/test/IR/parser.mlir
mlir/test/IR/slice.mlir
mlir/test/Integration/Dialect/Async/CPU/microbench-linalg-async-parallel-for.mlir
mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-1d.mlir
mlir/test/Integration/Dialect/Async/CPU/test-async-parallel-for-2d.mlir
mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir
mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major.mlir
mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major_as_row_major.mlir
mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir
mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir
mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-ncw-filter-wcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-input-nwc-filter-wcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-ncw-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-nwc-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nchw-filter-hwcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-input-nhwc-filter-hwcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nchw-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-nhwc-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ncdhw-filter-dhwcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-input-ndhwc-filter-dhwcf-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ncdhw-call.mlir
mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-ndhwc-call.mlir
mlir/test/Integration/Dialect/Standard/CPU/test-ceil-floor-pos-neg.mlir
mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-sparse-saxpy-jagged-matvec.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir
mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir
mlir/test/Integration/Sparse/CPU/frostt-example.mlir
mlir/test/Integration/Sparse/CPU/matrix-market-example.mlir
mlir/test/Integration/Sparse/CPU/sparse_sampled_matmul.mlir
mlir/test/Integration/Sparse/CPU/sparse_sum.mlir
mlir/test/Transforms/buffer-deallocation.mlir
mlir/test/Transforms/buffer-hoisting.mlir
mlir/test/Transforms/buffer-loop-hoisting.mlir
mlir/test/Transforms/buffer-results-to-out-params.mlir
mlir/test/Transforms/canonicalize-block-merge.mlir
mlir/test/Transforms/canonicalize.mlir
mlir/test/Transforms/constant-fold.mlir
mlir/test/Transforms/copy-removal.mlir
mlir/test/Transforms/cse.mlir
mlir/test/Transforms/finalizing-bufferize.mlir
mlir/test/Transforms/loop-fusion-dependence-check.mlir
mlir/test/Transforms/loop-fusion-slice-computation.mlir
mlir/test/Transforms/loop-fusion-transformation.mlir
mlir/test/Transforms/loop-fusion.mlir
mlir/test/Transforms/loop-invariant-code-motion.mlir
mlir/test/Transforms/memref-bound-check.mlir
mlir/test/Transforms/memref-dataflow-opt.mlir
mlir/test/Transforms/memref-dependence-check.mlir
mlir/test/Transforms/normalize-memrefs-ops.mlir
mlir/test/Transforms/normalize-memrefs.mlir
mlir/test/Transforms/parametric-tiling.mlir
mlir/test/Transforms/pipeline-data-transfer.mlir
mlir/test/Transforms/promote-buffers-to-stack.mlir
mlir/test/lib/Dialect/Affine/TestAffineDataCopy.cpp
mlir/test/lib/Dialect/Test/TestDialect.cpp
mlir/test/lib/Dialect/Test/TestPatterns.cpp
mlir/test/lib/Transforms/TestConvVectorization.cpp
mlir/test/lib/Transforms/TestGpuMemoryPromotion.cpp
mlir/test/lib/Transforms/TestGpuRewrite.cpp
mlir/test/lib/Transforms/TestLinalgCodegenStrategy.cpp
mlir/test/lib/Transforms/TestLinalgFusionTransforms.cpp
mlir/test/lib/Transforms/TestLinalgTransforms.cpp
mlir/test/lib/Transforms/TestMemRefStrideCalculation.cpp
mlir/test/lib/Transforms/TestSparsification.cpp
mlir/test/lib/Transforms/TestVectorTransforms.cpp
mlir/test/mlir-cpu-runner/async-value.mlir
mlir/test/mlir-cpu-runner/async.mlir
mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir
mlir/test/mlir-cpu-runner/global_memref.mlir
mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir
mlir/test/mlir-cpu-runner/memref_reshape.mlir
mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir
mlir/test/mlir-cpu-runner/unranked_memref.mlir
mlir/test/mlir-cpu-runner/utils.mlir
mlir/test/mlir-opt/commandline.mlir
mlir/test/mlir-reduce/multiple-function.mlir
mlir/test/mlir-reduce/simple-test.mlir
mlir/unittests/ExecutionEngine/Invoke.cpp
Removed:
################################################################################
diff --git a/mlir/docs/BufferDeallocationInternals.md b/mlir/docs/BufferDeallocationInternals.md
index acc775fd9f30..9626acad8825 100644
--- a/mlir/docs/BufferDeallocationInternals.md
+++ b/mlir/docs/BufferDeallocationInternals.md
@@ -779,8 +779,8 @@ the deallocation of the source value.
## Known Limitations
BufferDeallocation introduces additional copies using allocations from the
-“std” dialect (“std.alloc”). Analogous, all deallocations use the “std”
-dialect-free operation “std.dealloc”. The actual copy process is realized using
-“linalg.copy”. Furthermore, buffers are essentially immutable after their
-creation in a block. Another limitations are known in the case using
-unstructered control flow.
+“memref” dialect (“memref.alloc”). Analogous, all deallocations use the
+“memref” dialect-free operation “memref.dealloc”. The actual copy process is
+realized using “linalg.copy”. Furthermore, buffers are essentially immutable
+after their creation in a block. Another limitations are known in the case
+using unstructered control flow.
diff --git a/mlir/docs/Bufferization.md b/mlir/docs/Bufferization.md
index a131b2b7ba87..de0648deea7a 100644
--- a/mlir/docs/Bufferization.md
+++ b/mlir/docs/Bufferization.md
@@ -190,8 +190,8 @@ One convenient utility provided by the MLIR bufferization infrastructure is the
`BufferizeTypeConverter`, which comes pre-loaded with the necessary conversions
and materializations between `tensor` and `memref`.
-In this case, the `StandardOpsDialect` is marked as legal, so the `tensor_load`
-and `tensor_to_memref` ops, which are inserted automatically by the dialect
+In this case, the `MemRefOpsDialect` is marked as legal, so the `tensor_load`
+and `buffer_cast` ops, which are inserted automatically by the dialect
conversion framework as materializations, are legal. There is a helper
`populateBufferizeMaterializationLegality`
([code](https://github.com/llvm/llvm-project/blob/a0b65a7bcd6065688189b3d678c42ed6af9603db/mlir/include/mlir/Transforms/Bufferize.h#L53))
@@ -247,7 +247,7 @@ from the program.
The easiest way to write a finalizing bufferize pass is to not write one at all!
MLIR provides a pass `finalizing-bufferize` which eliminates the `tensor_load` /
-`tensor_to_memref` materialization ops inserted by partial bufferization passes
+`buffer_cast` materialization ops inserted by partial bufferization passes
and emits an error if that is not sufficient to remove all tensors from the
program.
@@ -268,7 +268,7 @@ recommended in new code. A helper,
`populateEliminateBufferizeMaterializationsPatterns`
([code](https://github.com/llvm/llvm-project/blob/a0b65a7bcd6065688189b3d678c42ed6af9603db/mlir/include/mlir/Transforms/Bufferize.h#L58))
is available for such passes to provide patterns that eliminate `tensor_load`
-and `tensor_to_memref`.
+and `buffer_cast`.
## Changes since [the talk](#the-talk)
diff --git a/mlir/docs/Dialects/Linalg.md b/mlir/docs/Dialects/Linalg.md
index e606e67c1e7c..4be06d2d4cf6 100644
--- a/mlir/docs/Dialects/Linalg.md
+++ b/mlir/docs/Dialects/Linalg.md
@@ -406,9 +406,9 @@ into a form that will resemble:
#map0 = affine_map<(d0, d1)[s0, s1, s2] -> (d0 * s1 + s0 + d1 * s2)>
func @example(%arg0: memref<?x?xf32>, %arg1: memref<?x?xf32>, %arg2: memref<?x?xf32>) {
- %0 = memref_cast %arg0 : memref<?x?xf32> to memref<?x?xf32, #map0>
- %1 = memref_cast %arg1 : memref<?x?xf32> to memref<?x?xf32, #map0>
- %2 = memref_cast %arg2 : memref<?x?xf32> to memref<?x?xf32, #map0>
+ %0 = memref.cast %arg0 : memref<?x?xf32> to memref<?x?xf32, #map0>
+ %1 = memref.cast %arg1 : memref<?x?xf32> to memref<?x?xf32, #map0>
+ %2 = memref.cast %arg2 : memref<?x?xf32> to memref<?x?xf32, #map0>
call @pointwise_add(%0, %1, %2) : (memref<?x?xf32, #map0>, memref<?x?xf32, #map0>, memref<?x?xf32, #map0>) -> ()
return
}
@@ -518,9 +518,9 @@ A set of ops that manipulate metadata but do not move memory. These ops take
generally alias the operand `view`. At the moment the existing ops are:
```
-* `std.view`,
+* `memref.view`,
* `std.subview`,
-* `std.transpose`.
+* `memref.transpose`.
* `linalg.range`,
* `linalg.slice`,
* `linalg.reshape`,
diff --git a/mlir/docs/Dialects/MemRef.md b/mlir/docs/Dialects/MemRef.md
new file mode 100644
index 000000000000..d7b62d60817c
--- /dev/null
+++ b/mlir/docs/Dialects/MemRef.md
@@ -0,0 +1,76 @@
+# 'memref' Dialect
+
+This dialect provides documentation for operations within the MemRef dialect.
+
+**Please post an RFC on the [forum](https://llvm.discourse.group/c/mlir/31)
+before adding or changing any operation in this dialect.**
+
+[TOC]
+
+## Operations
+
+[include "Dialects/MemRefOps.md"]
+
+### 'dma_start' operation
+
+Syntax:
+
+```
+operation ::= `dma_start` ssa-use`[`ssa-use-list`]` `,`
+ ssa-use`[`ssa-use-list`]` `,` ssa-use `,`
+ ssa-use`[`ssa-use-list`]` (`,` ssa-use `,` ssa-use)?
+ `:` memref-type `,` memref-type `,` memref-type
+```
+
+Starts a non-blocking DMA operation that transfers data from a source memref to
+a destination memref. The operands include the source and destination memref's
+each followed by its indices, size of the data transfer in terms of the number
+of elements (of the elemental type of the memref), a tag memref with its
+indices, and optionally two additional arguments corresponding to the stride (in
+terms of number of elements) and the number of elements to transfer per stride.
+The tag location is used by a dma_wait operation to check for completion. The
+indices of the source memref, destination memref, and the tag memref have the
+same restrictions as any load/store operation in an affine context (whenever DMA
+operations appear in an affine context). See
+[restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
+in affine contexts. This allows powerful static analysis and transformations in
+the presence of such DMAs including rescheduling, pipelining / overlap with
+computation, and checking for matching start/end operations. The source and
+destination memref need not be of the same dimensionality, but need to have the
+same elemental type.
+
+For example, a `dma_start` operation that transfers 32 vector elements from a
+memref `%src` at location `[%i, %j]` to memref `%dst` at `[%k, %l]` would be
+specified as shown below.
+
+Example:
+
+```mlir
+%size = constant 32 : index
+%tag = alloc() : memref<1 x i32, affine_map<(d0) -> (d0)>, 4>
+%idx = constant 0 : index
+dma_start %src[%i, %j], %dst[%k, %l], %size, %tag[%idx] :
+ memref<40 x 8 x vector<16xf32>, affine_map<(d0, d1) -> (d0, d1)>, 0>,
+ memref<2 x 4 x vector<16xf32>, affine_map<(d0, d1) -> (d0, d1)>, 2>,
+ memref<1 x i32>, affine_map<(d0) -> (d0)>, 4>
+```
+
+### 'dma_wait' operation
+
+Syntax:
+
+```
+operation ::= `dma_wait` ssa-use`[`ssa-use-list`]` `,` ssa-use `:` memref-type
+```
+
+Blocks until the completion of a DMA operation associated with the tag element
+specified with a tag memref and its indices. The operands include the tag memref
+followed by its indices and the number of elements associated with the DMA being
+waited on. The indices of the tag memref have the same restrictions as
+load/store indices.
+
+Example:
+
+```mlir
+dma_wait %tag[%idx], %size : memref<1 x i32, affine_map<(d0) -> (d0)>, 4>
+```
diff --git a/mlir/docs/Dialects/Standard.md b/mlir/docs/Dialects/Standard.md
index 8cc92f59228c..942a97676c8f 100644
--- a/mlir/docs/Dialects/Standard.md
+++ b/mlir/docs/Dialects/Standard.md
@@ -13,67 +13,3 @@ before adding or changing any operation in this dialect.**
## Operations
[include "Dialects/StandardOps.md"]
-
-### 'dma_start' operation
-
-Syntax:
-
-```
-operation ::= `dma_start` ssa-use`[`ssa-use-list`]` `,`
- ssa-use`[`ssa-use-list`]` `,` ssa-use `,`
- ssa-use`[`ssa-use-list`]` (`,` ssa-use `,` ssa-use)?
- `:` memref-type `,` memref-type `,` memref-type
-```
-
-Starts a non-blocking DMA operation that transfers data from a source memref to
-a destination memref. The operands include the source and destination memref's
-each followed by its indices, size of the data transfer in terms of the number
-of elements (of the elemental type of the memref), a tag memref with its
-indices, and optionally two additional arguments corresponding to the stride (in
-terms of number of elements) and the number of elements to transfer per stride.
-The tag location is used by a dma_wait operation to check for completion. The
-indices of the source memref, destination memref, and the tag memref have the
-same restrictions as any load/store operation in an affine context (whenever DMA
-operations appear in an affine context). See
-[restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
-in affine contexts. This allows powerful static analysis and transformations in
-the presence of such DMAs including rescheduling, pipelining / overlap with
-computation, and checking for matching start/end operations. The source and
-destination memref need not be of the same dimensionality, but need to have the
-same elemental type.
-
-For example, a `dma_start` operation that transfers 32 vector elements from a
-memref `%src` at location `[%i, %j]` to memref `%dst` at `[%k, %l]` would be
-specified as shown below.
-
-Example:
-
-```mlir
-%size = constant 32 : index
-%tag = alloc() : memref<1 x i32, affine_map<(d0) -> (d0)>, 4>
-%idx = constant 0 : index
-dma_start %src[%i, %j], %dst[%k, %l], %size, %tag[%idx] :
- memref<40 x 8 x vector<16xf32>, affine_map<(d0, d1) -> (d0, d1)>, 0>,
- memref<2 x 4 x vector<16xf32>, affine_map<(d0, d1) -> (d0, d1)>, 2>,
- memref<1 x i32>, affine_map<(d0) -> (d0)>, 4>
-```
-
-### 'dma_wait' operation
-
-Syntax:
-
-```
-operation ::= `dma_wait` ssa-use`[`ssa-use-list`]` `,` ssa-use `:` memref-type
-```
-
-Blocks until the completion of a DMA operation associated with the tag element
-specified with a tag memref and its indices. The operands include the tag memref
-followed by its indices and the number of elements associated with the DMA being
-waited on. The indices of the tag memref have the same restrictions as
-load/store indices.
-
-Example:
-
-```mlir
-dma_wait %tag[%idx], %size : memref<1 x i32, affine_map<(d0) -> (d0)>, 4>
-```
diff --git a/mlir/docs/Rationale/UsageOfConst.md b/mlir/docs/Rationale/UsageOfConst.md
index 3082dda7ad23..4ca9d09d3017 100644
--- a/mlir/docs/Rationale/UsageOfConst.md
+++ b/mlir/docs/Rationale/UsageOfConst.md
@@ -200,7 +200,7 @@ for.
### The `OpPointer` and `ConstOpPointer` Classes
The "typed operation" classes for registered operations (e.g. like `DimOp` for
-the "std.dim" operation in standard ops) contain a pointer to an operation and
+the "memref.dim" operation in memref ops) contain a pointer to an operation and
provide typed APIs for processing it.
However, this is a problem for our current `const` design - `const DimOp` means
diff --git a/mlir/docs/Traits.md b/mlir/docs/Traits.md
index 0b85929aab93..8a2182757812 100644
--- a/mlir/docs/Traits.md
+++ b/mlir/docs/Traits.md
@@ -211,7 +211,7 @@ are nested inside of other operations that themselves have this trait.
This trait is carried by region holding operations that define a new scope for
automatic allocation. Such allocations are automatically freed when control is
transferred back from the regions of such operations. As an example, allocations
-performed by [`std.alloca`](Dialects/Standard.md#stdalloca-allocaop) are
+performed by [`memref.alloca`](Dialects/MemRef.md#memrefalloca-allocaop) are
automatically freed when control leaves the region of its closest surrounding op
that has the trait AutomaticAllocationScope.
diff --git a/mlir/docs/Tutorials/Toy/Ch-5.md b/mlir/docs/Tutorials/Toy/Ch-5.md
index d3c054ee16db..b8964f93e1a3 100644
--- a/mlir/docs/Tutorials/Toy/Ch-5.md
+++ b/mlir/docs/Tutorials/Toy/Ch-5.md
@@ -50,8 +50,9 @@ framework, we need to provide two things (and an optional third):
## Conversion Target
For our purposes, we want to convert the compute-intensive `Toy` operations into
-a combination of operations from the `Affine` `Standard` dialects for further
-optimization. To start off the lowering, we first define our conversion target:
+a combination of operations from the `Affine`, `MemRef` and `Standard` dialects
+for further optimization. To start off the lowering, we first define our
+conversion target:
```c++
void ToyToAffineLoweringPass::runOnFunction() {
@@ -61,8 +62,9 @@ void ToyToAffineLoweringPass::runOnFunction() {
// We define the specific operations, or dialects, that are legal targets for
// this lowering. In our case, we are lowering to a combination of the
- // `Affine` and `Standard` dialects.
- target.addLegalDialect<mlir::AffineDialect, mlir::StandardOpsDialect>();
+ // `Affine`, `MemRef` and `Standard` dialects.
+ target.addLegalDialect<mlir::AffineDialect, mlir::memref::MemRefDialect,
+ mlir::StandardOpsDialect>();
// We also define the Toy dialect as Illegal so that the conversion will fail
// if any of these operations are *not* converted. Given that we actually want
diff --git a/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp
index 46d8fe40e5d4..f2332fd5437f 100644
--- a/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp
+++ b/mlir/examples/toy/Ch5/mlir/LowerToAffineLoops.cpp
@@ -7,8 +7,8 @@
//===----------------------------------------------------------------------===//
//
// This file implements a partial lowering of Toy operations to a combination of
-// affine loops and standard operations. This lowering expects that all calls
-// have been inlined, and all shapes have been resolved.
+// affine loops, memref operations and standard operations. This lowering
+// expects that all calls have been inlined, and all shapes have been resolved.
//
//===----------------------------------------------------------------------===//
@@ -16,6 +16,7 @@
#include "toy/Passes.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/DialectConversion.h"
@@ -36,7 +37,7 @@ static MemRefType convertTensorToMemRef(TensorType type) {
/// Insert an allocation and deallocation for the given MemRefType.
static Value insertAllocAndDealloc(MemRefType type, Location loc,
PatternRewriter &rewriter) {
- auto alloc = rewriter.create<AllocOp>(loc, type);
+ auto alloc = rewriter.create<memref::AllocOp>(loc, type);
// Make sure to allocate at the beginning of the block.
auto *parentBlock = alloc->getBlock();
@@ -44,7 +45,7 @@ static Value insertAllocAndDealloc(MemRefType type, Location loc,
// Make sure to deallocate this alloc at the end of the block. This is fine
// as toy functions have no control flow.
- auto dealloc = rewriter.create<DeallocOp>(loc, alloc);
+ auto dealloc = rewriter.create<memref::DeallocOp>(loc, alloc);
dealloc->moveBefore(&parentBlock->back());
return alloc;
}
@@ -152,8 +153,8 @@ struct ConstantOpLowering : public OpRewritePattern<toy::ConstantOp> {
if (!valueShape.empty()) {
for (auto i : llvm::seq<int64_t>(
- 0, *std::max_element(valueShape.begin(), valueShape.end())))
- constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
+ 0, *std::max_element(valueShape.begin(), valueShape.end())))
+ constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
} else {
// This is the case of a tensor of rank 0.
constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, 0));
@@ -257,7 +258,7 @@ namespace {
struct ToyToAffineLoweringPass
: public PassWrapper<ToyToAffineLoweringPass, FunctionPass> {
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, StandardOpsDialect>();
+ registry.insert<AffineDialect, memref::MemRefDialect, StandardOpsDialect>();
}
void runOnFunction() final;
};
@@ -283,8 +284,9 @@ void ToyToAffineLoweringPass::runOnFunction() {
// We define the specific operations, or dialects, that are legal targets for
// this lowering. In our case, we are lowering to a combination of the
- // `Affine` and `Standard` dialects.
- target.addLegalDialect<AffineDialect, StandardOpsDialect>();
+ // `Affine`, `MemRef` and `Standard` dialects.
+ target.addLegalDialect<AffineDialect, memref::MemRefDialect,
+ StandardOpsDialect>();
// We also define the Toy dialect as Illegal so that the conversion will fail
// if any of these operations are *not* converted. Given that we actually want
diff --git a/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp
index 2089b6579ccb..250ad387594f 100644
--- a/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp
+++ b/mlir/examples/toy/Ch6/mlir/LowerToAffineLoops.cpp
@@ -7,8 +7,8 @@
//===----------------------------------------------------------------------===//
//
// This file implements a partial lowering of Toy operations to a combination of
-// affine loops and standard operations. This lowering expects that all calls
-// have been inlined, and all shapes have been resolved.
+// affine loops, memref operations and standard operations. This lowering
+// expects that all calls have been inlined, and all shapes have been resolved.
//
//===----------------------------------------------------------------------===//
@@ -16,6 +16,7 @@
#include "toy/Passes.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/DialectConversion.h"
@@ -36,7 +37,7 @@ static MemRefType convertTensorToMemRef(TensorType type) {
/// Insert an allocation and deallocation for the given MemRefType.
static Value insertAllocAndDealloc(MemRefType type, Location loc,
PatternRewriter &rewriter) {
- auto alloc = rewriter.create<AllocOp>(loc, type);
+ auto alloc = rewriter.create<memref::AllocOp>(loc, type);
// Make sure to allocate at the beginning of the block.
auto *parentBlock = alloc->getBlock();
@@ -44,7 +45,7 @@ static Value insertAllocAndDealloc(MemRefType type, Location loc,
// Make sure to deallocate this alloc at the end of the block. This is fine
// as toy functions have no control flow.
- auto dealloc = rewriter.create<DeallocOp>(loc, alloc);
+ auto dealloc = rewriter.create<memref::DeallocOp>(loc, alloc);
dealloc->moveBefore(&parentBlock->back());
return alloc;
}
@@ -152,8 +153,8 @@ struct ConstantOpLowering : public OpRewritePattern<toy::ConstantOp> {
if (!valueShape.empty()) {
for (auto i : llvm::seq<int64_t>(
- 0, *std::max_element(valueShape.begin(), valueShape.end())))
- constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
+ 0, *std::max_element(valueShape.begin(), valueShape.end())))
+ constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
} else {
// This is the case of a tensor of rank 0.
constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, 0));
@@ -256,7 +257,7 @@ namespace {
struct ToyToAffineLoweringPass
: public PassWrapper<ToyToAffineLoweringPass, FunctionPass> {
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, StandardOpsDialect>();
+ registry.insert<AffineDialect, memref::MemRefDialect, StandardOpsDialect>();
}
void runOnFunction() final;
};
@@ -282,8 +283,9 @@ void ToyToAffineLoweringPass::runOnFunction() {
// We define the specific operations, or dialects, that are legal targets for
// this lowering. In our case, we are lowering to a combination of the
- // `Affine` and `Standard` dialects.
- target.addLegalDialect<AffineDialect, StandardOpsDialect>();
+ // `Affine`, `MemRef` and `Standard` dialects.
+ target.addLegalDialect<AffineDialect, memref::MemRefDialect,
+ StandardOpsDialect>();
// We also define the Toy dialect as Illegal so that the conversion will fail
// if any of these operations are *not* converted. Given that we actually want
diff --git a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp
index 8523a8371192..bbe3e54f8ac5 100644
--- a/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp
+++ b/mlir/examples/toy/Ch6/mlir/LowerToLLVM.cpp
@@ -30,6 +30,7 @@
#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
@@ -91,7 +92,8 @@ class PrintOpLowering : public ConversionPattern {
// Generate a call to printf for the current element of the loop.
auto printOp = cast<toy::PrintOp>(op);
- auto elementLoad = rewriter.create<LoadOp>(loc, printOp.input(), loopIvs);
+ auto elementLoad =
+ rewriter.create<memref::LoadOp>(loc, printOp.input(), loopIvs);
rewriter.create<CallOp>(loc, printfRef, rewriter.getIntegerType(32),
ArrayRef<Value>({formatSpecifierCst, elementLoad}));
diff --git a/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp b/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp
index 46d8fe40e5d4..f2332fd5437f 100644
--- a/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp
+++ b/mlir/examples/toy/Ch7/mlir/LowerToAffineLoops.cpp
@@ -7,8 +7,8 @@
//===----------------------------------------------------------------------===//
//
// This file implements a partial lowering of Toy operations to a combination of
-// affine loops and standard operations. This lowering expects that all calls
-// have been inlined, and all shapes have been resolved.
+// affine loops, memref operations and standard operations. This lowering
+// expects that all calls have been inlined, and all shapes have been resolved.
//
//===----------------------------------------------------------------------===//
@@ -16,6 +16,7 @@
#include "toy/Passes.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/DialectConversion.h"
@@ -36,7 +37,7 @@ static MemRefType convertTensorToMemRef(TensorType type) {
/// Insert an allocation and deallocation for the given MemRefType.
static Value insertAllocAndDealloc(MemRefType type, Location loc,
PatternRewriter &rewriter) {
- auto alloc = rewriter.create<AllocOp>(loc, type);
+ auto alloc = rewriter.create<memref::AllocOp>(loc, type);
// Make sure to allocate at the beginning of the block.
auto *parentBlock = alloc->getBlock();
@@ -44,7 +45,7 @@ static Value insertAllocAndDealloc(MemRefType type, Location loc,
// Make sure to deallocate this alloc at the end of the block. This is fine
// as toy functions have no control flow.
- auto dealloc = rewriter.create<DeallocOp>(loc, alloc);
+ auto dealloc = rewriter.create<memref::DeallocOp>(loc, alloc);
dealloc->moveBefore(&parentBlock->back());
return alloc;
}
@@ -152,8 +153,8 @@ struct ConstantOpLowering : public OpRewritePattern<toy::ConstantOp> {
if (!valueShape.empty()) {
for (auto i : llvm::seq<int64_t>(
- 0, *std::max_element(valueShape.begin(), valueShape.end())))
- constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
+ 0, *std::max_element(valueShape.begin(), valueShape.end())))
+ constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, i));
} else {
// This is the case of a tensor of rank 0.
constantIndices.push_back(rewriter.create<ConstantIndexOp>(loc, 0));
@@ -257,7 +258,7 @@ namespace {
struct ToyToAffineLoweringPass
: public PassWrapper<ToyToAffineLoweringPass, FunctionPass> {
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, StandardOpsDialect>();
+ registry.insert<AffineDialect, memref::MemRefDialect, StandardOpsDialect>();
}
void runOnFunction() final;
};
@@ -283,8 +284,9 @@ void ToyToAffineLoweringPass::runOnFunction() {
// We define the specific operations, or dialects, that are legal targets for
// this lowering. In our case, we are lowering to a combination of the
- // `Affine` and `Standard` dialects.
- target.addLegalDialect<AffineDialect, StandardOpsDialect>();
+ // `Affine`, `MemRef` and `Standard` dialects.
+ target.addLegalDialect<AffineDialect, memref::MemRefDialect,
+ StandardOpsDialect>();
// We also define the Toy dialect as Illegal so that the conversion will fail
// if any of these operations are *not* converted. Given that we actually want
diff --git a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp
index 8523a8371192..bbe3e54f8ac5 100644
--- a/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp
+++ b/mlir/examples/toy/Ch7/mlir/LowerToLLVM.cpp
@@ -30,6 +30,7 @@
#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
@@ -91,7 +92,8 @@ class PrintOpLowering : public ConversionPattern {
// Generate a call to printf for the current element of the loop.
auto printOp = cast<toy::PrintOp>(op);
- auto elementLoad = rewriter.create<LoadOp>(loc, printOp.input(), loopIvs);
+ auto elementLoad =
+ rewriter.create<memref::LoadOp>(loc, printOp.input(), loopIvs);
rewriter.create<CallOp>(loc, printfRef, rewriter.getIntegerType(32),
ArrayRef<Value>({formatSpecifierCst, elementLoad}));
diff --git a/mlir/include/mlir/Conversion/Passes.td b/mlir/include/mlir/Conversion/Passes.td
index b818c2b3ea23..f2e3fc3d3d24 100644
--- a/mlir/include/mlir/Conversion/Passes.td
+++ b/mlir/include/mlir/Conversion/Passes.td
@@ -121,7 +121,7 @@ def LowerHostCodeToLLVM : Pass<"lower-host-to-llvm", "ModuleOp"> {
def ConvertGpuOpsToNVVMOps : Pass<"convert-gpu-to-nvvm", "gpu::GPUModuleOp"> {
let summary = "Generate NVVM operations for gpu operations";
let constructor = "mlir::createLowerGpuOpsToNVVMOpsPass()";
- let dependentDialects = ["NVVM::NVVMDialect"];
+ let dependentDialects = ["NVVM::NVVMDialect", "memref::MemRefDialect"];
let options = [
Option<"indexBitwidth", "index-bitwidth", "unsigned",
/*default=kDeriveIndexBitwidthFromDataLayout*/"0",
@@ -210,7 +210,7 @@ def ConvertLinalgToStandard : Pass<"convert-linalg-to-std", "ModuleOp"> {
let summary = "Convert the operations from the linalg dialect into the "
"Standard dialect";
let constructor = "mlir::createConvertLinalgToStandardPass()";
- let dependentDialects = ["StandardOpsDialect"];
+ let dependentDialects = ["memref::MemRefDialect", "StandardOpsDialect"];
}
//===----------------------------------------------------------------------===//
@@ -316,7 +316,11 @@ def ConvertShapeToStandard : Pass<"convert-shape-to-std", "ModuleOp"> {
let summary = "Convert operations from the shape dialect into the standard "
"dialect";
let constructor = "mlir::createConvertShapeToStandardPass()";
- let dependentDialects = ["StandardOpsDialect", "scf::SCFDialect"];
+ let dependentDialects = [
+ "memref::MemRefDialect",
+ "StandardOpsDialect",
+ "scf::SCFDialect"
+ ];
}
def ConvertShapeConstraints: Pass<"convert-shape-constraints", "FuncOp"> {
@@ -474,7 +478,11 @@ def ConvertVectorToSCF : FunctionPass<"convert-vector-to-scf"> {
let summary = "Lower the operations from the vector dialect into the SCF "
"dialect";
let constructor = "mlir::createConvertVectorToSCFPass()";
- let dependentDialects = ["AffineDialect", "scf::SCFDialect"];
+ let dependentDialects = [
+ "AffineDialect",
+ "memref::MemRefDialect",
+ "scf::SCFDialect"
+ ];
let options = [
Option<"fullUnroll", "full-unroll", "bool", /*default=*/"false",
"Perform full unrolling when converting vector transfers to SCF">,
diff --git a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h b/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h
index 02fefc689bac..ca623a3c050e 100644
--- a/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h
+++ b/mlir/include/mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h
@@ -72,7 +72,8 @@ void populateStdToLLVMConversionPatterns(LLVMTypeConverter &converter,
/// Creates a pass to convert the Standard dialect into the LLVMIR dialect.
/// stdlib malloc/free is used by default for allocating memrefs allocated with
-/// std.alloc, while LLVM's alloca is used for those allocated with std.alloca.
+/// memref.alloc, while LLVM's alloca is used for those allocated with
+/// memref.alloca.
std::unique_ptr<OperationPass<ModuleOp>>
createLowerToLLVMPass(const LowerToLLVMOptions &options =
LowerToLLVMOptions::getDefaultOptions());
diff --git a/mlir/include/mlir/Dialect/Affine/Passes.td b/mlir/include/mlir/Dialect/Affine/Passes.td
index 9f2aac7b685a..77ba06483304 100644
--- a/mlir/include/mlir/Dialect/Affine/Passes.td
+++ b/mlir/include/mlir/Dialect/Affine/Passes.td
@@ -18,6 +18,7 @@ include "mlir/Pass/PassBase.td"
def AffineDataCopyGeneration : FunctionPass<"affine-data-copy-generate"> {
let summary = "Generate explicit copying for affine memory operations";
let constructor = "mlir::createAffineDataCopyGenerationPass()";
+ let dependentDialects = ["memref::MemRefDialect"];
let options = [
Option<"fastMemoryCapacity", "fast-mem-capacity", "uint64_t",
/*default=*/"std::numeric_limits<uint64_t>::max()",
diff --git a/mlir/include/mlir/Dialect/CMakeLists.txt b/mlir/include/mlir/Dialect/CMakeLists.txt
index f9fab7937fc4..8abad863ba32 100644
--- a/mlir/include/mlir/Dialect/CMakeLists.txt
+++ b/mlir/include/mlir/Dialect/CMakeLists.txt
@@ -9,6 +9,7 @@ add_subdirectory(GPU)
add_subdirectory(Math)
add_subdirectory(Linalg)
add_subdirectory(LLVMIR)
+add_subdirectory(MemRef)
add_subdirectory(OpenACC)
add_subdirectory(OpenMP)
add_subdirectory(PDL)
diff --git a/mlir/include/mlir/Dialect/GPU/GPUOps.td b/mlir/include/mlir/Dialect/GPU/GPUOps.td
index b8b958dde11a..5fb0793030b0 100644
--- a/mlir/include/mlir/Dialect/GPU/GPUOps.td
+++ b/mlir/include/mlir/Dialect/GPU/GPUOps.td
@@ -480,7 +480,7 @@ def GPU_LaunchOp : GPU_Op<"launch">,
%num_bx : index, %num_by : index, %num_bz : index,
%num_tx : index, %num_ty : index, %num_tz : index)
"some_op"(%bx, %tx) : (index, index) -> ()
- %3 = "std.load"(%val1, %bx) : (memref<?xf32, 1>, index) -> f32
+ %3 = "memref.load"(%val1, %bx) : (memref<?xf32, 1>, index) -> f32
}
```
@@ -812,7 +812,7 @@ def GPU_AllocOp : GPU_Op<"alloc", [
let summary = "GPU memory allocation operation.";
let description = [{
The `gpu.alloc` operation allocates a region of memory on the GPU. It is
- similar to the `std.alloc` op, but supports asynchronous GPU execution.
+ similar to the `memref.alloc` op, but supports asynchronous GPU execution.
The op does not execute before all async dependencies have finished
executing.
@@ -850,7 +850,7 @@ def GPU_DeallocOp : GPU_Op<"dealloc", [GPU_AsyncOpInterface]> {
let description = [{
The `gpu.dealloc` operation frees the region of memory referenced by a
memref which was originally created by the `gpu.alloc` operation. It is
- similar to the `std.dealloc` op, but supports asynchronous GPU execution.
+ similar to the `memref.dealloc` op, but supports asynchronous GPU execution.
The op does not execute before all async dependencies have finished
executing.
diff --git a/mlir/include/mlir/Dialect/Linalg/EDSC/FoldedIntrinsics.h b/mlir/include/mlir/Dialect/Linalg/EDSC/FoldedIntrinsics.h
index d98ec90a47e0..afabcb92ae1e 100644
--- a/mlir/include/mlir/Dialect/Linalg/EDSC/FoldedIntrinsics.h
+++ b/mlir/include/mlir/Dialect/Linalg/EDSC/FoldedIntrinsics.h
@@ -11,6 +11,7 @@
#include "mlir/Dialect/Linalg/EDSC/Builders.h"
#include "mlir/Dialect/Linalg/EDSC/Intrinsics.h"
#include "mlir/Dialect/Math/IR/Math.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
#include "mlir/Transforms/FoldUtils.h"
@@ -35,30 +36,25 @@ struct FoldedValueBuilder {
};
using folded_math_tanh = FoldedValueBuilder<math::TanhOp>;
-using folded_std_constant_index = FoldedValueBuilder<ConstantIndexOp>;
-using folded_std_constant_float = FoldedValueBuilder<ConstantFloatOp>;
-using folded_std_constant_int = FoldedValueBuilder<ConstantIntOp>;
-using folded_std_constant = FoldedValueBuilder<ConstantOp>;
-using folded_std_dim = FoldedValueBuilder<DimOp>;
+using folded_memref_alloc = FoldedValueBuilder<memref::AllocOp>;
+using folded_memref_cast = FoldedValueBuilder<memref::CastOp>;
+using folded_memref_dim = FoldedValueBuilder<memref::DimOp>;
+using folded_memref_load = FoldedValueBuilder<memref::LoadOp>;
+using folded_memref_sub_view = FoldedValueBuilder<memref::SubViewOp>;
+using folded_memref_tensor_load = FoldedValueBuilder<memref::TensorLoadOp>;
+using folded_memref_view = FoldedValueBuilder<memref::ViewOp>;
using folded_std_muli = FoldedValueBuilder<MulIOp>;
using folded_std_addi = FoldedValueBuilder<AddIOp>;
using folded_std_addf = FoldedValueBuilder<AddFOp>;
-using folded_std_alloc = FoldedValueBuilder<AllocOp>;
using folded_std_constant = FoldedValueBuilder<ConstantOp>;
using folded_std_constant_float = FoldedValueBuilder<ConstantFloatOp>;
using folded_std_constant_index = FoldedValueBuilder<ConstantIndexOp>;
using folded_std_constant_int = FoldedValueBuilder<ConstantIntOp>;
-using folded_std_dim = FoldedValueBuilder<DimOp>;
using folded_std_index_cast = FoldedValueBuilder<IndexCastOp>;
using folded_std_muli = FoldedValueBuilder<MulIOp>;
using folded_std_mulf = FoldedValueBuilder<MulFOp>;
-using folded_std_memref_cast = FoldedValueBuilder<MemRefCastOp>;
using folded_std_select = FoldedValueBuilder<SelectOp>;
-using folded_std_load = FoldedValueBuilder<LoadOp>;
using folded_std_subi = FoldedValueBuilder<SubIOp>;
-using folded_std_sub_view = FoldedValueBuilder<SubViewOp>;
-using folded_std_tensor_load = FoldedValueBuilder<TensorLoadOp>;
-using folded_std_view = FoldedValueBuilder<ViewOp>;
using folded_std_zero_extendi = FoldedValueBuilder<ZeroExtendIOp>;
using folded_std_sign_extendi = FoldedValueBuilder<SignExtendIOp>;
using folded_tensor_extract = FoldedValueBuilder<tensor::ExtractOp>;
diff --git a/mlir/include/mlir/Dialect/Linalg/IR/LinalgSparseOps.td b/mlir/include/mlir/Dialect/Linalg/IR/LinalgSparseOps.td
index 9d9402478495..f0653c462aae 100644
--- a/mlir/include/mlir/Dialect/Linalg/IR/LinalgSparseOps.td
+++ b/mlir/include/mlir/Dialect/Linalg/IR/LinalgSparseOps.td
@@ -18,7 +18,7 @@
//
// The other operations form the bridge between the opaque pointer and
// the actual storage of pointers, indices, and values. These operations
-// resemble 'tensor_to_memref' in the sense that they map tensors to
+// resemble 'buffer_cast' in the sense that they map tensors to
// their bufferized memrefs, but they lower into actual calls since
// sparse storage does not bufferize into a single memrefs, as dense
// tensors do, but into a hierarchical storage scheme where pointers
@@ -74,9 +74,9 @@ def Linalg_SparseTensorToPointersMemRefOp :
let description = [{
Returns the pointers array of the sparse storage scheme at the
given dimension for the given tensor. This is similar to the
- `tensor_to_memref` operation in the sense that it provides a bridge
+ `buffer_cast` operation in the sense that it provides a bridge
between a tensor world view and a bufferized world view. Unlike the
- `tensor_to_memref` operation, however, this sparse operation actually
+ `buffer_cast` operation, however, this sparse operation actually
lowers into a call into a support library to obtain access to the
pointers array.
@@ -98,9 +98,9 @@ def Linalg_SparseTensorToIndicesMemRefOp :
let description = [{
Returns the indices array of the sparse storage scheme at the
given dimension for the given tensor. This is similar to the
- `tensor_to_memref` operation in the sense that it provides a bridge
+ `buffer_cast` operation in the sense that it provides a bridge
between a tensor world view and a bufferized world view. Unlike the
- `tensor_to_memref` operation, however, this sparse operation actually
+ `buffer_cast` operation, however, this sparse operation actually
lowers into a call into a support library to obtain access to the
indices array.
@@ -122,9 +122,9 @@ def Linalg_SparseTensorToValuesMemRefOp :
let description = [{
Returns the values array of the sparse storage scheme for the given
tensor, independent of the actual dimension. This is similar to the
- `tensor_to_memref` operation in the sense that it provides a bridge
+ `buffer_cast` operation in the sense that it provides a bridge
between a tensor world view and a bufferized world view. Unlike the
- `tensor_to_memref` operation, however, this sparse operation actually
+ `buffer_cast` operation, however, this sparse operation actually
lowers into a call into a support library to obtain access to the
values array.
diff --git a/mlir/include/mlir/Dialect/Linalg/Passes.h b/mlir/include/mlir/Dialect/Linalg/Passes.h
index 7d93dd00d86a..34e25689763c 100644
--- a/mlir/include/mlir/Dialect/Linalg/Passes.h
+++ b/mlir/include/mlir/Dialect/Linalg/Passes.h
@@ -34,11 +34,11 @@ createLinalgPromotionPass(bool dynamicBuffers, bool useAlloca);
std::unique_ptr<OperationPass<FuncOp>> createLinalgPromotionPass();
/// Create a pass to convert Linalg operations to scf.for loops and
-/// std.load/std.store accesses.
+/// memref.load/memref.store accesses.
std::unique_ptr<OperationPass<FuncOp>> createConvertLinalgToLoopsPass();
/// Create a pass to convert Linalg operations to scf.parallel loops and
-/// std.load/std.store accesses.
+/// memref.load/memref.store accesses.
std::unique_ptr<OperationPass<FuncOp>> createConvertLinalgToParallelLoopsPass();
/// Create a pass to convert Linalg operations to affine.for loops and
diff --git a/mlir/include/mlir/Dialect/Linalg/Passes.td b/mlir/include/mlir/Dialect/Linalg/Passes.td
index ec54e93c988d..aad11179be69 100644
--- a/mlir/include/mlir/Dialect/Linalg/Passes.td
+++ b/mlir/include/mlir/Dialect/Linalg/Passes.td
@@ -19,7 +19,7 @@ def ConvertElementwiseToLinalg : FunctionPass<"convert-elementwise-to-linalg"> {
This pass only converts ops that operate on ranked tensors.
}];
let constructor = "mlir::createConvertElementwiseToLinalgPass()";
- let dependentDialects = ["linalg::LinalgDialect"];
+ let dependentDialects = ["linalg::LinalgDialect", "memref::MemRefDialect"];
}
def LinalgFoldUnitExtentDims : FunctionPass<"linalg-fold-unit-extent-dims"> {
@@ -70,13 +70,21 @@ def LinalgLowerToLoops : FunctionPass<"convert-linalg-to-loops"> {
"interchange vector",
"llvm::cl::ZeroOrMore, llvm::cl::MiscFlags::CommaSeparated">
];
- let dependentDialects = ["linalg::LinalgDialect", "scf::SCFDialect", "AffineDialect"];
+ let dependentDialects = [
+ "linalg::LinalgDialect",
+ "scf::SCFDialect",
+ "AffineDialect"
+ ];
}
def LinalgBufferize : Pass<"linalg-bufferize", "FuncOp"> {
let summary = "Bufferize the linalg dialect";
let constructor = "mlir::createLinalgBufferizePass()";
- let dependentDialects = ["linalg::LinalgDialect", "AffineDialect"];
+ let dependentDialects = [
+ "linalg::LinalgDialect",
+ "AffineDialect",
+ "memref::MemRefDialect"
+ ];
}
def LinalgLowerToParallelLoops
@@ -90,7 +98,12 @@ def LinalgLowerToParallelLoops
"interchange vector",
"llvm::cl::ZeroOrMore, llvm::cl::MiscFlags::CommaSeparated">
];
- let dependentDialects = ["AffineDialect", "linalg::LinalgDialect", "scf::SCFDialect"];
+ let dependentDialects = [
+ "AffineDialect",
+ "linalg::LinalgDialect",
+ "memref::MemRefDialect",
+ "scf::SCFDialect"
+ ];
}
def LinalgPromotion : FunctionPass<"linalg-promote-subviews"> {
@@ -109,7 +122,10 @@ def LinalgTiling : FunctionPass<"linalg-tile"> {
let summary = "Tile operations in the linalg dialect";
let constructor = "mlir::createLinalgTilingPass()";
let dependentDialects = [
- "AffineDialect", "linalg::LinalgDialect", "scf::SCFDialect"
+ "AffineDialect",
+ "linalg::LinalgDialect",
+ "memref::MemRefDialect",
+ "scf::SCFDialect"
];
let options = [
ListOption<"tileSizes", "linalg-tile-sizes", "int64_t",
@@ -127,7 +143,12 @@ def LinalgTilingToParallelLoops
"Test generation of dynamic promoted buffers",
"llvm::cl::ZeroOrMore, llvm::cl::MiscFlags::CommaSeparated">
];
- let dependentDialects = ["AffineDialect", "linalg::LinalgDialect", "scf::SCFDialect"];
+ let dependentDialects = [
+ "AffineDialect",
+ "linalg::LinalgDialect",
+ "memref::MemRefDialect",
+ "scf::SCFDialect"
+ ];
}
def LinalgGeneralization : FunctionPass<"linalg-generalize-named-ops"> {
diff --git a/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h b/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h
index 8f422d284df6..54a4aec9f867 100644
--- a/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h
+++ b/mlir/include/mlir/Dialect/Linalg/Transforms/Transforms.h
@@ -147,8 +147,8 @@ LinalgOp interchange(LinalgOp op, ArrayRef<unsigned> interchangeVector);
/// dimension. If that is not possible, contains the dynamic size of the
/// subview. The call back should return the buffer to use.
using AllocBufferCallbackFn = std::function<Optional<Value>(
- OpBuilder &b, SubViewOp subView, ArrayRef<Value> boundingSubViewSize,
- OperationFolder *folder)>;
+ OpBuilder &b, memref::SubViewOp subView,
+ ArrayRef<Value> boundingSubViewSize, OperationFolder *folder)>;
/// Callback function type used to deallocate the buffers used to hold the
/// promoted subview.
@@ -244,7 +244,7 @@ struct PromotionInfo {
Value partialLocalView;
};
Optional<PromotionInfo>
-promoteSubviewAsNewBuffer(OpBuilder &b, Location loc, SubViewOp subView,
+promoteSubviewAsNewBuffer(OpBuilder &b, Location loc, memref::SubViewOp subView,
AllocBufferCallbackFn allocationFn,
OperationFolder *folder = nullptr);
@@ -818,7 +818,7 @@ struct PadTensorOpVectorizationPattern : public OpRewritePattern<PadTensorOp> {
/// Match and rewrite for the pattern:
/// ```
/// %alloc = ...
-/// [optional] %view = std.view %alloc ...
+/// [optional] %view = memref.view %alloc ...
/// %subView = subview %allocOrView ...
/// [optional] linalg.fill(%allocOrView, %cst) ...
/// ...
@@ -828,7 +828,7 @@ struct PadTensorOpVectorizationPattern : public OpRewritePattern<PadTensorOp> {
/// into
/// ```
/// [unchanged] %alloc = ...
-/// [unchanged] [optional] %view = std.view %alloc ...
+/// [unchanged] [optional] %view = memref.view %alloc ...
/// [unchanged] [unchanged] %subView = subview %allocOrView ...
/// ...
/// vector.transfer_read %in[...], %cst ...
@@ -849,7 +849,7 @@ struct LinalgCopyVTRForwardingPattern
/// Match and rewrite for the pattern:
/// ```
/// %alloc = ...
-/// [optional] %view = std.view %alloc ...
+/// [optional] %view = memref.view %alloc ...
/// %subView = subview %allocOrView...
/// ...
/// vector.transfer_write %..., %allocOrView[...]
@@ -858,7 +858,7 @@ struct LinalgCopyVTRForwardingPattern
/// into
/// ```
/// [unchanged] %alloc = ...
-/// [unchanged] [optional] %view = std.view %alloc ...
+/// [unchanged] [optional] %view = memref.view %alloc ...
/// [unchanged] %subView = subview %allocOrView...
/// ...
/// vector.transfer_write %..., %out[...]
diff --git a/mlir/include/mlir/Dialect/Linalg/Utils/Utils.h b/mlir/include/mlir/Dialect/Linalg/Utils/Utils.h
index 21b7dcfe4319..2dc208f429f4 100644
--- a/mlir/include/mlir/Dialect/Linalg/Utils/Utils.h
+++ b/mlir/include/mlir/Dialect/Linalg/Utils/Utils.h
@@ -13,6 +13,7 @@
#include "mlir/Dialect/Linalg/Analysis/DependenceAnalysis.h"
#include "mlir/Dialect/Linalg/EDSC/Builders.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
@@ -21,7 +22,7 @@
#include "llvm/ADT/SetVector.h"
using mlir::edsc::intrinsics::AffineIndexedValue;
-using mlir::edsc::intrinsics::StdIndexedValue;
+using mlir::edsc::intrinsics::MemRefIndexedValue;
namespace mlir {
class AffineExpr;
@@ -213,7 +214,7 @@ template <typename LoopTy>
struct GenerateLoopNest {
using IndexedValueTy =
typename std::conditional<std::is_same<LoopTy, AffineForOp>::value,
- AffineIndexedValue, StdIndexedValue>::type;
+ AffineIndexedValue, MemRefIndexedValue>::type;
static void
doit(ArrayRef<Range> loopRanges, ValueRange iterArgInitValues,
diff --git a/mlir/include/mlir/Dialect/MemRef/CMakeLists.txt b/mlir/include/mlir/Dialect/MemRef/CMakeLists.txt
new file mode 100644
index 000000000000..f33061b2d87c
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/CMakeLists.txt
@@ -0,0 +1 @@
+add_subdirectory(IR)
diff --git a/mlir/include/mlir/Dialect/MemRef/EDSC/Intrinsics.h b/mlir/include/mlir/Dialect/MemRef/EDSC/Intrinsics.h
new file mode 100644
index 000000000000..fb2ee895c507
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/EDSC/Intrinsics.h
@@ -0,0 +1,37 @@
+//===- Intrinsics.h - MLIR EDSC Intrinsics for MemRefOps --------*- 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_DIALECT_MEMREF_EDSC_INTRINSICS_H_
+#define MLIR_DIALECT_MEMREF_EDSC_INTRINSICS_H_
+
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
+#include "mlir/EDSC/Builders.h"
+
+namespace mlir {
+namespace edsc {
+namespace intrinsics {
+
+using memref_alloc = ValueBuilder<memref::AllocOp>;
+using memref_alloca = ValueBuilder<memref::AllocaOp>;
+using memref_cast = ValueBuilder<memref::CastOp>;
+using memref_dealloc = OperationBuilder<memref::DeallocOp>;
+using memref_dim = ValueBuilder<memref::DimOp>;
+using memref_load = ValueBuilder<memref::LoadOp>;
+using memref_store = OperationBuilder<memref::StoreOp>;
+using memref_sub_view = ValueBuilder<memref::SubViewOp>;
+using memref_tensor_load = ValueBuilder<memref::TensorLoadOp>;
+using memref_tensor_store = OperationBuilder<memref::TensorStoreOp>;
+using memref_view = ValueBuilder<memref::ViewOp>;
+
+/// Provide an index notation around memref_load and memref_store.
+using MemRefIndexedValue =
+ TemplatedIndexedValue<intrinsics::memref_load, intrinsics::memref_store>;
+} // namespace intrinsics
+} // namespace edsc
+} // namespace mlir
+
+#endif // MLIR_DIALECT_MEMREF_EDSC_INTRINSICS_H_
diff --git a/mlir/include/mlir/Dialect/MemRef/IR/CMakeLists.txt b/mlir/include/mlir/Dialect/MemRef/IR/CMakeLists.txt
new file mode 100644
index 000000000000..762f8dd133cd
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/IR/CMakeLists.txt
@@ -0,0 +1,2 @@
+add_mlir_dialect(MemRefOps memref)
+add_mlir_doc(MemRefOps -gen-dialect-doc MemRefOps Dialects/)
diff --git a/mlir/include/mlir/Dialect/MemRef/IR/MemRef.h b/mlir/include/mlir/Dialect/MemRef/IR/MemRef.h
new file mode 100644
index 000000000000..503a6583dbb1
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/IR/MemRef.h
@@ -0,0 +1,239 @@
+//===- MemRef.h - MemRef dialect --------------------------------*- 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_DIALECT_MEMREF_IR_MEMREF_H_
+#define MLIR_DIALECT_MEMREF_IR_MEMREF_H_
+
+#include "mlir/IR/Dialect.h"
+#include "mlir/Interfaces/CallInterfaces.h"
+#include "mlir/Interfaces/CastInterfaces.h"
+#include "mlir/Interfaces/SideEffectInterfaces.h"
+#include "mlir/Interfaces/ViewLikeInterface.h"
+
+namespace mlir {
+raw_ostream &operator<<(raw_ostream &os, Range &range);
+
+/// Return the list of Range (i.e. offset, size, stride). Each Range
+/// entry contains either the dynamic value or a ConstantIndexOp constructed
+/// with `b` at location `loc`.
+SmallVector<Range, 8> getOrCreateRanges(OffsetSizeAndStrideOpInterface op,
+ OpBuilder &b, Location loc);
+} // namespace mlir
+
+//===----------------------------------------------------------------------===//
+// MemRef Dialect
+//===----------------------------------------------------------------------===//
+
+#include "mlir/Dialect/MemRef/IR/MemRefOpsDialect.h.inc"
+
+//===----------------------------------------------------------------------===//
+// MemRef Dialect Operations
+//===----------------------------------------------------------------------===//
+
+#define GET_OP_CLASSES
+#include "mlir/Dialect/MemRef/IR/MemRefOps.h.inc"
+
+namespace mlir {
+namespace memref {
+// DmaStartOp starts a non-blocking DMA operation that transfers data from a
+// source memref to a destination memref. The source and destination memref need
+// not be of the same dimensionality, but need to have the same elemental type.
+// The operands include the source and destination memref's each followed by its
+// indices, size of the data transfer in terms of the number of elements (of the
+// elemental type of the memref), a tag memref with its indices, and optionally
+// at the end, a stride and a number_of_elements_per_stride arguments. The tag
+// location is used by a DmaWaitOp to check for completion. The indices of the
+// source memref, destination memref, and the tag memref have the same
+// restrictions as any load/store. The optional stride arguments should be of
+// 'index' type, and specify a stride for the slower memory space (memory space
+// with a lower memory space id), transferring chunks of
+// number_of_elements_per_stride every stride until %num_elements are
+// transferred. Either both or no stride arguments should be specified.
+//
+// For example, a DmaStartOp operation that transfers 256 elements of a memref
+// '%src' in memory space 0 at indices [%i, %j] to memref '%dst' in memory space
+// 1 at indices [%k, %l], would be specified as follows:
+//
+// %num_elements = constant 256
+// %idx = constant 0 : index
+// %tag = alloc() : memref<1 x i32, (d0) -> (d0), 4>
+// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%idx] :
+// memref<40 x 128 x f32>, (d0) -> (d0), 0>,
+// memref<2 x 1024 x f32>, (d0) -> (d0), 1>,
+// memref<1 x i32>, (d0) -> (d0), 2>
+//
+// If %stride and %num_elt_per_stride are specified, the DMA is expected to
+// transfer %num_elt_per_stride elements every %stride elements apart from
+// memory space 0 until %num_elements are transferred.
+//
+// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%idx], %stride,
+// %num_elt_per_stride :
+//
+// TODO: add additional operands to allow source and destination striding, and
+// multiple stride levels.
+// TODO: Consider replacing src/dst memref indices with view memrefs.
+class DmaStartOp
+ : public Op<DmaStartOp, OpTrait::VariadicOperands, OpTrait::ZeroResult> {
+public:
+ using Op::Op;
+
+ static void build(OpBuilder &builder, OperationState &result, Value srcMemRef,
+ ValueRange srcIndices, Value destMemRef,
+ ValueRange destIndices, Value numElements, Value tagMemRef,
+ ValueRange tagIndices, Value stride = nullptr,
+ Value elementsPerStride = nullptr);
+
+ // Returns the source MemRefType for this DMA operation.
+ Value getSrcMemRef() { return getOperand(0); }
+ // Returns the rank (number of indices) of the source MemRefType.
+ unsigned getSrcMemRefRank() {
+ return getSrcMemRef().getType().cast<MemRefType>().getRank();
+ }
+ // Returns the source memref indices for this DMA operation.
+ operand_range getSrcIndices() {
+ return {(*this)->operand_begin() + 1,
+ (*this)->operand_begin() + 1 + getSrcMemRefRank()};
+ }
+
+ // Returns the destination MemRefType for this DMA operations.
+ Value getDstMemRef() { return getOperand(1 + getSrcMemRefRank()); }
+ // Returns the rank (number of indices) of the destination MemRefType.
+ unsigned getDstMemRefRank() {
+ return getDstMemRef().getType().cast<MemRefType>().getRank();
+ }
+ unsigned getSrcMemorySpace() {
+ return getSrcMemRef().getType().cast<MemRefType>().getMemorySpaceAsInt();
+ }
+ unsigned getDstMemorySpace() {
+ return getDstMemRef().getType().cast<MemRefType>().getMemorySpaceAsInt();
+ }
+
+ // Returns the destination memref indices for this DMA operation.
+ operand_range getDstIndices() {
+ return {(*this)->operand_begin() + 1 + getSrcMemRefRank() + 1,
+ (*this)->operand_begin() + 1 + getSrcMemRefRank() + 1 +
+ getDstMemRefRank()};
+ }
+
+ // Returns the number of elements being transferred by this DMA operation.
+ Value getNumElements() {
+ return getOperand(1 + getSrcMemRefRank() + 1 + getDstMemRefRank());
+ }
+
+ // Returns the Tag MemRef for this DMA operation.
+ Value getTagMemRef() {
+ return getOperand(1 + getSrcMemRefRank() + 1 + getDstMemRefRank() + 1);
+ }
+ // Returns the rank (number of indices) of the tag MemRefType.
+ unsigned getTagMemRefRank() {
+ return getTagMemRef().getType().cast<MemRefType>().getRank();
+ }
+
+ // Returns the tag memref index for this DMA operation.
+ operand_range getTagIndices() {
+ unsigned tagIndexStartPos =
+ 1 + getSrcMemRefRank() + 1 + getDstMemRefRank() + 1 + 1;
+ return {(*this)->operand_begin() + tagIndexStartPos,
+ (*this)->operand_begin() + tagIndexStartPos + getTagMemRefRank()};
+ }
+
+ /// Returns true if this is a DMA from a faster memory space to a slower one.
+ bool isDestMemorySpaceFaster() {
+ return (getSrcMemorySpace() < getDstMemorySpace());
+ }
+
+ /// Returns true if this is a DMA from a slower memory space to a faster one.
+ bool isSrcMemorySpaceFaster() {
+ // Assumes that a lower number is for a slower memory space.
+ return (getDstMemorySpace() < getSrcMemorySpace());
+ }
+
+ /// Given a DMA start operation, returns the operand position of either the
+ /// source or destination memref depending on the one that is at the higher
+ /// level of the memory hierarchy. Asserts failure if neither is true.
+ unsigned getFasterMemPos() {
+ assert(isSrcMemorySpaceFaster() || isDestMemorySpaceFaster());
+ return isSrcMemorySpaceFaster() ? 0 : getSrcMemRefRank() + 1;
+ }
+
+ static StringRef getOperationName() { return "memref.dma_start"; }
+ static ParseResult parse(OpAsmParser &parser, OperationState &result);
+ void print(OpAsmPrinter &p);
+ LogicalResult verify();
+
+ LogicalResult fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results);
+
+ bool isStrided() {
+ return getNumOperands() != 1 + getSrcMemRefRank() + 1 + getDstMemRefRank() +
+ 1 + 1 + getTagMemRefRank();
+ }
+
+ Value getStride() {
+ if (!isStrided())
+ return nullptr;
+ return getOperand(getNumOperands() - 1 - 1);
+ }
+
+ Value getNumElementsPerStride() {
+ if (!isStrided())
+ return nullptr;
+ return getOperand(getNumOperands() - 1);
+ }
+};
+
+// DmaWaitOp blocks until the completion of a DMA operation associated with the
+// tag element '%tag[%index]'. %tag is a memref, and %index has to be an index
+// with the same restrictions as any load/store index. %num_elements is the
+// number of elements associated with the DMA operation. For example:
+//
+// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%index] :
+// memref<2048 x f32>, (d0) -> (d0), 0>,
+// memref<256 x f32>, (d0) -> (d0), 1>
+// memref<1 x i32>, (d0) -> (d0), 2>
+// ...
+// ...
+// dma_wait %tag[%index], %num_elements : memref<1 x i32, (d0) -> (d0), 2>
+//
+class DmaWaitOp
+ : public Op<DmaWaitOp, OpTrait::VariadicOperands, OpTrait::ZeroResult> {
+public:
+ using Op::Op;
+
+ static void build(OpBuilder &builder, OperationState &result, Value tagMemRef,
+ ValueRange tagIndices, Value numElements);
+
+ static StringRef getOperationName() { return "memref.dma_wait"; }
+
+ // Returns the Tag MemRef associated with the DMA operation being waited on.
+ Value getTagMemRef() { return getOperand(0); }
+
+ // Returns the tag memref index for this DMA operation.
+ operand_range getTagIndices() {
+ return {(*this)->operand_begin() + 1,
+ (*this)->operand_begin() + 1 + getTagMemRefRank()};
+ }
+
+ // Returns the rank (number of indices) of the tag memref.
+ unsigned getTagMemRefRank() {
+ return getTagMemRef().getType().cast<MemRefType>().getRank();
+ }
+
+ // Returns the number of elements transferred in the associated DMA operation.
+ Value getNumElements() { return getOperand(1 + getTagMemRefRank()); }
+
+ static ParseResult parse(OpAsmParser &parser, OperationState &result);
+ void print(OpAsmPrinter &p);
+ LogicalResult fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results);
+ LogicalResult verify();
+};
+} // namespace memref
+} // namespace mlir
+
+#endif // MLIR_DIALECT_MEMREF_IR_MEMREF_H_
diff --git a/mlir/include/mlir/Dialect/MemRef/IR/MemRefBase.td b/mlir/include/mlir/Dialect/MemRef/IR/MemRefBase.td
new file mode 100644
index 000000000000..ada1e526ca42
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/IR/MemRefBase.td
@@ -0,0 +1,25 @@
+//===- MemRefBase.td - Base definitions for memref dialect -*- tablegen -*-===//
+//
+// 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 MEMREF_BASE
+#define MEMREF_BASE
+
+include "mlir/IR/OpBase.td"
+
+def MemRef_Dialect : Dialect {
+ let name = "memref";
+ let cppNamespace = "::mlir::memref";
+ let description = [{
+ The `memref` dialect is intended to hold core memref creation and
+ manipulation ops, which are not strongly associated with any particular
+ other dialect or domain abstraction.
+ }];
+ let hasConstantMaterializer = 1;
+}
+
+#endif // MEMREF_BASE
diff --git a/mlir/include/mlir/Dialect/MemRef/IR/MemRefOps.td b/mlir/include/mlir/Dialect/MemRef/IR/MemRefOps.td
new file mode 100644
index 000000000000..804c850d3bd2
--- /dev/null
+++ b/mlir/include/mlir/Dialect/MemRef/IR/MemRefOps.td
@@ -0,0 +1,1270 @@
+//===- MemRefOps.td - MemRef op definitions ----------------*- tablegen -*-===//
+//
+// 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 MEMREF_OPS
+#define MEMREF_OPS
+
+include "mlir/Dialect/MemRef/IR/MemRefBase.td"
+include "mlir/Interfaces/CastInterfaces.td"
+include "mlir/Interfaces/SideEffectInterfaces.td"
+include "mlir/Interfaces/ViewLikeInterface.td"
+include "mlir/IR/SymbolInterfaces.td"
+
+class MemRef_Op<string mnemonic, list<OpTrait> traits = []>
+ : Op<MemRef_Dialect, mnemonic, traits> {
+ let printer = [{ return ::print(p, *this); }];
+ let verifier = [{ return ::verify(*this); }];
+ let parser = [{ return ::parse$cppClass(parser, result); }];
+}
+
+//===----------------------------------------------------------------------===//
+// AllocLikeOp
+//===----------------------------------------------------------------------===//
+
+// Base class for memref allocating ops: alloca and alloc.
+//
+// %0 = alloclike(%m)[%s] : memref<8x?xf32, (d0, d1)[s0] -> ((d0 + s0), d1)>
+//
+class AllocLikeOp<string mnemonic,
+ Resource resource,
+ list<OpTrait> traits = []> :
+ MemRef_Op<mnemonic,
+ !listconcat([
+ AttrSizedOperandSegments
+ ], traits)> {
+
+ let arguments = (ins Variadic<Index>:$dynamicSizes,
+ // The symbolic operands (the ones in square brackets) bind
+ // to the symbols of the memref's layout map.
+ Variadic<Index>:$symbolOperands,
+ Confined<OptionalAttr<I64Attr>, [IntMinValue<0>]>:$alignment);
+ let results = (outs Res<AnyMemRef, "", [MemAlloc<resource>]>:$memref);
+
+ let builders = [
+ OpBuilder<(ins "MemRefType":$memrefType,
+ CArg<"IntegerAttr", "IntegerAttr()">:$alignment), [{
+ return build($_builder, $_state, memrefType, {}, alignment);
+ }]>,
+ OpBuilder<(ins "MemRefType":$memrefType, "ValueRange":$dynamicSizes,
+ CArg<"IntegerAttr", "IntegerAttr()">:$alignment), [{
+ return build($_builder, $_state, memrefType, dynamicSizes, {}, alignment);
+ }]>,
+ OpBuilder<(ins "MemRefType":$memrefType, "ValueRange":$dynamicSizes,
+ "ValueRange":$symbolOperands,
+ CArg<"IntegerAttr", "{}">:$alignment), [{
+ $_state.types.push_back(memrefType);
+ $_state.addOperands(dynamicSizes);
+ $_state.addOperands(symbolOperands);
+ $_state.addAttribute(getOperandSegmentSizeAttr(),
+ $_builder.getI32VectorAttr({
+ static_cast<int32_t>(dynamicSizes.size()),
+ static_cast<int32_t>(symbolOperands.size())}));
+ if (alignment)
+ $_state.addAttribute(getAlignmentAttrName(), alignment);
+ }]>];
+
+ let extraClassDeclaration = [{
+ static StringRef getAlignmentAttrName() { return "alignment"; }
+
+ MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
+
+ /// Returns the dynamic sizes for this alloc operation if specified.
+ operand_range getDynamicSizes() { return dynamicSizes(); }
+ }];
+
+ let assemblyFormat = [{
+ `(`$dynamicSizes`)` (`` `[` $symbolOperands^ `]`)? attr-dict `:` type($memref)
+ }];
+
+ let hasCanonicalizer = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// AssumeAlignmentOp
+//===----------------------------------------------------------------------===//
+
+def AssumeAlignmentOp : MemRef_Op<"assume_alignment"> {
+ let summary =
+ "assertion that gives alignment information to the input memref";
+ let description = [{
+ The `assume_alignment` operation takes a memref and an integer of alignment
+ value, and internally annotates the buffer with the given alignment. If
+ the buffer isn't aligned to the given alignment, the behavior is undefined.
+
+ This operation doesn't affect the semantics of a correct program. It's for
+ optimization only, and the optimization is best-effort.
+ }];
+ let arguments = (ins AnyMemRef:$memref,
+ Confined<I32Attr, [IntPositive]>:$alignment);
+ let results = (outs);
+
+ let assemblyFormat = "$memref `,` $alignment attr-dict `:` type($memref)";
+}
+
+//===----------------------------------------------------------------------===//
+// BaseOpWithOffsetSizesAndStrides
+//===----------------------------------------------------------------------===//
+
+// Base class for ops with static/dynamic offset, sizes and strides
+// attributes/arguments.
+class BaseOpWithOffsetSizesAndStrides<string mnemonic, list<OpTrait> traits = []> :
+ MemRef_Op<mnemonic,
+ !listconcat(traits, [NoSideEffect, AttrSizedOperandSegments])> {
+ code extraBaseClassDeclaration = [{
+ /// Returns the dynamic sizes for this subview operation if specified.
+ operand_range getDynamicSizes() { return sizes(); }
+
+ /// Return the list of Range (i.e. offset, size, stride). Each
+ /// Range entry contains either the dynamic value or a ConstantIndexOp
+ /// constructed with `b` at location `loc`.
+ SmallVector<Range, 8> getOrCreateRanges(OpBuilder &b, Location loc) {
+ return mlir::getOrCreateRanges(*this, b, loc);
+ }
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// AllocOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_AllocOp : AllocLikeOp<"alloc", DefaultResource> {
+ let summary = "memory allocation operation";
+ let description = [{
+ The `alloc` operation allocates a region of memory, as specified by its
+ memref type.
+
+ Example:
+
+ ```mlir
+ %0 = memref.alloc() : memref<8x64xf32, 1>
+ ```
+
+ The optional list of dimension operands are bound to the dynamic dimensions
+ specified in its memref type. In the example below, the ssa value '%d' is
+ bound to the second dimension of the memref (which is dynamic).
+
+ ```mlir
+ %0 = memref.alloc(%d) : memref<8x?xf32, 1>
+ ```
+
+ The optional list of symbol operands are bound to the symbols of the
+ memrefs affine map. In the example below, the ssa value '%s' is bound to
+ the symbol 's0' in the affine map specified in the allocs memref type.
+
+ ```mlir
+ %0 = memref.alloc()[%s] : memref<8x64xf32,
+ affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ ```
+
+ This operation returns a single ssa value of memref type, which can be used
+ by subsequent load and store operations.
+
+ The optional `alignment` attribute may be specified to ensure that the
+ region of memory that will be indexed is aligned at the specified byte
+ boundary.
+
+ ```mlir
+ %0 = memref.alloc()[%s] {alignment = 8} :
+ memref<8x64xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ ```
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// AllocaOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_AllocaOp : AllocLikeOp<"alloca", AutomaticAllocationScopeResource> {
+ let summary = "stack memory allocation operation";
+ let description = [{
+ The `alloca` operation allocates memory on the stack, to be automatically
+ released when control transfers back from the region of its closest
+ surrounding operation with an
+ [`AutomaticAllocationScope`](../Traits.md#automaticallocationscope) trait.
+ The amount of memory allocated is specified by its memref and additional
+ operands. For example:
+
+ ```mlir
+ %0 = memref.alloca() : memref<8x64xf32>
+ ```
+
+ The optional list of dimension operands are bound to the dynamic dimensions
+ specified in its memref type. In the example below, the SSA value '%d' is
+ bound to the second dimension of the memref (which is dynamic).
+
+ ```mlir
+ %0 = memref.alloca(%d) : memref<8x?xf32>
+ ```
+
+ The optional list of symbol operands are bound to the symbols of the
+ memref's affine map. In the example below, the SSA value '%s' is bound to
+ the symbol 's0' in the affine map specified in the allocs memref type.
+
+ ```mlir
+ %0 = memref.alloca()[%s] : memref<8x64xf32,
+ affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>>
+ ```
+
+ This operation returns a single SSA value of memref type, which can be used
+ by subsequent load and store operations. An optional alignment attribute, if
+ specified, guarantees alignment at least to that boundary. If not specified,
+ an alignment on any convenient boundary compatible with the type will be
+ chosen.
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// BufferCastOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_BufferCastOp : MemRef_Op<"buffer_cast",
+ [SameOperandsAndResultShape, SameOperandsAndResultElementType,
+ TypesMatchWith<"type of 'tensor' is the tensor equivalent of 'memref'",
+ "memref", "tensor",
+ "getTensorTypeFromMemRefType($_self)">]> {
+ let summary = "tensor to memref cast operation";
+ let description = [{
+ Casts a tensor to a memref.
+
+ ```mlir
+ // Result type is tensor<4x?xf32>
+ %12 = memref.buffer_cast %10 : memref<4x?xf32, #map0, 42>
+ ```
+ }];
+
+ let arguments = (ins AnyTensor:$tensor);
+ let results = (outs AnyRankedOrUnrankedMemRef:$memref);
+ // This op is fully verified by traits.
+ let verifier = ?;
+
+ let assemblyFormat = "$tensor attr-dict `:` type($memref)";
+
+ let hasFolder = 1;
+ let hasCanonicalizer = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// CastOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_CastOp : MemRef_Op<"cast", [
+ NoSideEffect, SameOperandsAndResultShape,
+ DeclareOpInterfaceMethods<CastOpInterface>
+ ]> {
+ let summary = "memref cast operation";
+ let description = [{
+ Syntax:
+
+ ```
+ operation ::= ssa-id `=` `memref.cast` ssa-use `:` type `to` type
+ ```
+
+ The `memref.cast` operation converts a memref from one type to an equivalent
+ type with a compatible shape. The source and destination types are
+ compatible if:
+
+ a. Both are ranked memref types with the same element type, address space,
+ and rank and:
+ 1. Both have the same layout or both have compatible strided layouts.
+ 2. The individual sizes (resp. offset and strides in the case of strided
+ memrefs) may convert constant dimensions to dynamic dimensions and
+ vice-versa.
+
+ If the cast converts any dimensions from an unknown to a known size, then it
+ acts as an assertion that fails at runtime if the dynamic dimensions
+ disagree with resultant destination size.
+
+ Example:
+
+ ```mlir
+ // Assert that the input dynamic shape matches the destination static shape.
+ %2 = memref.cast %1 : memref<?x?xf32> to memref<4x4xf32>
+ // Erase static shape information, replacing it with dynamic information.
+ %3 = memref.cast %1 : memref<4xf32> to memref<?xf32>
+
+ // The same holds true for offsets and strides.
+
+ // Assert that the input dynamic shape matches the destination static stride.
+ %4 = memref.cast %1 : memref<12x4xf32, offset:?, strides: [?, ?]> to
+ memref<12x4xf32, offset:5, strides: [4, 1]>
+ // Erase static offset and stride information, replacing it with
+ // dynamic information.
+ %5 = memref.cast %1 : memref<12x4xf32, offset:5, strides: [4, 1]> to
+ memref<12x4xf32, offset:?, strides: [?, ?]>
+ ```
+
+ b. Either or both memref types are unranked with the same element type, and
+ address space.
+
+ Example:
+
+ ```mlir
+ Cast to concrete shape.
+ %4 = memref.cast %1 : memref<*xf32> to memref<4x?xf32>
+
+ Erase rank information.
+ %5 = memref.cast %1 : memref<4x?xf32> to memref<*xf32>
+ ```
+ }];
+
+ let arguments = (ins AnyRankedOrUnrankedMemRef:$source);
+ let results = (outs AnyRankedOrUnrankedMemRef:$dest);
+ let assemblyFormat = "$source attr-dict `:` type($source) `to` type($dest)";
+ let verifier = "return impl::verifyCastOp(*this, areCastCompatible);";
+ let builders = [
+ OpBuilder<(ins "Value":$source, "Type":$destType), [{
+ impl::buildCastOp($_builder, $_state, source, destType);
+ }]>
+ ];
+
+ let extraClassDeclaration = [{
+ /// Fold the given CastOp into consumer op.
+ static bool canFoldIntoConsumerOp(CastOp castOp);
+ }];
+
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// DeallocOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_DeallocOp : MemRef_Op<"dealloc", [MemRefsNormalizable]> {
+ let summary = "memory deallocation operation";
+ let description = [{
+ The `dealloc` operation frees the region of memory referenced by a memref
+ which was originally created by the `alloc` operation.
+ The `dealloc` operation should not be called on memrefs which alias an
+ alloc'd memref (e.g. memrefs returned by `view` operations).
+
+ Example:
+
+ ```mlir
+ %0 = memref.alloc() : memref<8x64xf32, (d0, d1) -> (d0, d1), 1>
+ memref.dealloc %0 : memref<8x64xf32, (d0, d1) -> (d0, d1), 1>
+ ```
+ }];
+
+ let arguments = (ins Arg<AnyMemRef, "", [MemFree]>:$memref);
+
+ let hasCanonicalizer = 1;
+ let hasFolder = 1;
+ let assemblyFormat = "$memref attr-dict `:` type($memref)";
+}
+
+//===----------------------------------------------------------------------===//
+// DimOp
+//===----------------------------------------------------------------------===//
+
+def DimOp : MemRef_Op<"dim", [NoSideEffect]> {
+ let summary = "dimension index operation";
+ let description = [{
+ The `dim` operation takes a memref and a dimension operand of type `index`.
+ It returns the size of the requested dimension of the given memref.
+ If the dimension index is out of bounds the behavior is undefined.
+
+ The specified memref type is that of the first operand.
+
+ Example:
+
+ ```mlir
+ // Always returns 4, can be constant folded:
+ %c0 = constant 0 : index
+ %x = memref.dim %A, %c0 : memref<4 x ? x f32>
+
+ // Returns the dynamic dimension of %A.
+ %c1 = constant 1 : index
+ %y = memref.dim %A, %c1 : memref<4 x ? x f32>
+
+ // Equivalent generic form:
+ %x = "memref.dim"(%A, %c0) : (memref<4 x ? x f32>, index) -> index
+ %y = "memref.dim"(%A, %c1) : (memref<4 x ? x f32>, index) -> index
+ ```
+ }];
+
+ let arguments = (ins AnyTypeOf<[AnyTensor, AnyRankedOrUnrankedMemRef],
+ "any memref or tensor type">:$memrefOrTensor,
+ Index:$index);
+ let results = (outs Index:$result);
+
+ let assemblyFormat = [{
+ attr-dict $memrefOrTensor `,` $index `:` type($memrefOrTensor)
+ }];
+
+ let builders = [
+ OpBuilder<(ins "Value":$memrefOrTensor, "int64_t":$index)>,
+ OpBuilder<(ins "Value":$memrefOrTensor, "Value":$index)>
+ ];
+
+ let extraClassDeclaration = [{
+ /// Helper function to get the index as a simple integer if it is constant.
+ Optional<int64_t> getConstantIndex();
+ }];
+
+ let hasCanonicalizer = 1;
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// GetGlobalOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_GetGlobalOp : MemRef_Op<"get_global",
+ [NoSideEffect, DeclareOpInterfaceMethods<SymbolUserOpInterface>]> {
+ let summary = "get the memref pointing to a global variable";
+ let description = [{
+ The `memref.get_global` operation retrieves the memref pointing to a
+ named global variable. If the global variable is marked constant, writing
+ to the result memref (such as through a `memref.store` operation) is
+ undefined.
+
+ Example:
+
+ ```mlir
+ %x = memref.get_global @foo : memref<2xf32>
+ ```
+ }];
+
+ let arguments = (ins FlatSymbolRefAttr:$name);
+ let results = (outs AnyStaticShapeMemRef:$result);
+ let assemblyFormat = "$name `:` type($result) attr-dict";
+
+ // `GetGlobalOp` is fully verified by its traits.
+ let verifier = ?;
+}
+
+//===----------------------------------------------------------------------===//
+// GlobalOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_GlobalOp : MemRef_Op<"global", [Symbol]> {
+ let summary = "declare or define a global memref variable";
+ let description = [{
+ The `memref.global` operation declares or defines a named global variable.
+ The backing memory for the variable is allocated statically and is described
+ by the type of the variable (which should be a statically shaped memref
+ type). The operation is a declaration if no `inital_value` is specified,
+ else it is a definition. The `initial_value` can either be a unit attribute
+ to represent a definition of an uninitialized global variable, or an
+ elements attribute to represent the definition of a global variable with an
+ initial value. The global variable can also be marked constant using the
+ `constant` unit attribute. Writing to such constant global variables is
+ undefined.
+
+ The global variable can be accessed by using the `memref.get_global` to
+ retrieve the memref for the global variable. Note that the memref
+ for such global variable itself is immutable (i.e., memref.get_global for a
+ given global variable will always return the same memref descriptor).
+
+ Example:
+
+ ```mlir
+ // Private variable with an initial value.
+ memref.global "private" @x : memref<2xf32> = dense<0.0,2.0>
+
+ // Declaration of an external variable.
+ memref.global "private" @y : memref<4xi32>
+
+ // Uninitialized externally visible variable.
+ memref.global @z : memref<3xf16> = uninitialized
+
+ // Externally visible constant variable.
+ memref.global constant @c : memref<2xi32> = dense<1, 4>
+ ```
+ }];
+
+ let arguments = (ins
+ SymbolNameAttr:$sym_name,
+ OptionalAttr<StrAttr>:$sym_visibility,
+ TypeAttr:$type,
+ OptionalAttr<AnyAttr>:$initial_value,
+ UnitAttr:$constant
+ );
+
+ let assemblyFormat = [{
+ ($sym_visibility^)?
+ (`constant` $constant^)?
+ $sym_name `:`
+ custom<GlobalMemrefOpTypeAndInitialValue>($type, $initial_value)
+ attr-dict
+ }];
+
+ let extraClassDeclaration = [{
+ bool isExternal() { return !initial_value(); }
+ bool isUninitialized() {
+ return !isExternal() && initial_value().getValue().isa<UnitAttr>();
+ }
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// LoadOp
+//===----------------------------------------------------------------------===//
+
+def LoadOp : MemRef_Op<"load",
+ [TypesMatchWith<"result type matches element type of 'memref'",
+ "memref", "result",
+ "$_self.cast<MemRefType>().getElementType()">,
+ MemRefsNormalizable]> {
+ let summary = "load operation";
+ let description = [{
+ The `load` op reads an element from a memref specified by an index list. The
+ output of load is a new value with the same type as the elements of the
+ memref. The arity of indices is the rank of the memref (i.e., if the memref
+ loaded from is of rank 3, then 3 indices are required for the load following
+ the memref identifier).
+
+ In an `affine.if` or `affine.for` body, the indices of a load are restricted
+ to SSA values bound to surrounding loop induction variables,
+ [symbols](Affine.md#dimensions-and-symbols), results of a
+ [`constant` operation](#stdconstant-constantop), or the result of an
+ `affine.apply` operation that can in turn take as arguments all of the
+ aforementioned SSA values or the recursively result of such an
+ `affine.apply` operation.
+
+ Example:
+
+ ```mlir
+ %1 = affine.apply affine_map<(d0, d1) -> (3*d0)> (%i, %j)
+ %2 = affine.apply affine_map<(d0, d1) -> (d1+1)> (%i, %j)
+ %12 = memref.load %A[%1, %2] : memref<8x?xi32, #layout, memspace0>
+
+ // Example of an indirect load (treated as non-affine)
+ %3 = affine.apply affine_map<(d0) -> (2*d0 + 1)>(%12)
+ %13 = memref.load %A[%3, %2] : memref<4x?xi32, #layout, memspace0>
+ ```
+
+ **Context:** The `load` and `store` operations are specifically crafted to
+ fully resolve a reference to an element of a memref, and (in affine
+ `affine.if` and `affine.for` operations) the compiler can follow use-def
+ chains (e.g. through [`affine.apply`](Affine.md#affineapply-affineapplyop)
+ operations) to precisely analyze references at compile-time using polyhedral
+ techniques. This is possible because of the
+ [restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
+ in these contexts.
+ }];
+
+ let arguments = (ins Arg<AnyMemRef, "the reference to load from",
+ [MemRead]>:$memref,
+ Variadic<Index>:$indices);
+ let results = (outs AnyType:$result);
+
+ let builders = [
+ OpBuilder<(ins "Value":$memref, CArg<"ValueRange", "{}">:$indices), [{
+ auto memrefType = memref.getType().cast<MemRefType>();
+ $_state.addOperands(memref);
+ $_state.addOperands(indices);
+ $_state.types.push_back(memrefType.getElementType());
+ }]>];
+
+ let extraClassDeclaration = [{
+ Value getMemRef() { return getOperand(0); }
+ void setMemRef(Value value) { setOperand(0, value); }
+ MemRefType getMemRefType() {
+ return getMemRef().getType().cast<MemRefType>();
+ }
+
+ operand_range getIndices() { return {operand_begin() + 1, operand_end()}; }
+ }];
+
+ let hasCanonicalizer = 1;
+ let hasFolder = 1;
+
+ let assemblyFormat = "$memref `[` $indices `]` attr-dict `:` type($memref)";
+}
+
+//===----------------------------------------------------------------------===//
+// PrefetchOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_PrefetchOp : MemRef_Op<"prefetch"> {
+ let summary = "prefetch operation";
+ let description = [{
+ The "prefetch" op prefetches data from a memref location described with
+ subscript indices similar to memref.load, and with three attributes: a
+ read/write specifier, a locality hint, and a cache type specifier as shown
+ below:
+
+ ```mlir
+ memref.prefetch %0[%i, %j], read, locality<3>, data : memref<400x400xi32>
+ ```
+
+ The read/write specifier is either 'read' or 'write', the locality hint
+ ranges from locality<0> (no locality) to locality<3> (extremely local keep
+ in cache). The cache type specifier is either 'data' or 'instr'
+ and specifies whether the prefetch is performed on data cache or on
+ instruction cache.
+ }];
+
+ let arguments = (ins AnyMemRef:$memref, Variadic<Index>:$indices,
+ BoolAttr:$isWrite,
+ Confined<I32Attr, [IntMinValue<0>,
+ IntMaxValue<3>]>:$localityHint,
+ BoolAttr:$isDataCache);
+
+ let extraClassDeclaration = [{
+ MemRefType getMemRefType() {
+ return memref().getType().cast<MemRefType>();
+ }
+ static StringRef getLocalityHintAttrName() { return "localityHint"; }
+ static StringRef getIsWriteAttrName() { return "isWrite"; }
+ static StringRef getIsDataCacheAttrName() { return "isDataCache"; }
+ }];
+
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// ReinterpretCastOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_ReinterpretCastOp:
+ BaseOpWithOffsetSizesAndStrides<"reinterpret_cast", [
+ NoSideEffect, ViewLikeOpInterface, OffsetSizeAndStrideOpInterface
+ ]> {
+ let summary = "memref reinterpret cast operation";
+ let description = [{
+ Modify offset, sizes and strides of an unranked/ranked memref.
+
+ Example:
+ ```mlir
+ memref.reinterpret_cast %ranked to
+ offset: [0],
+ sizes: [%size0, 10],
+ strides: [1, %stride1]
+ : memref<?x?xf32> to memref<?x10xf32, offset: 0, strides: [1, ?]>
+
+ memref.reinterpret_cast %unranked to
+ offset: [%offset],
+ sizes: [%size0, %size1],
+ strides: [%stride0, %stride1]
+ : memref<*xf32> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ ```
+ }];
+
+ let arguments = (ins
+ Arg<AnyRankedOrUnrankedMemRef, "", []>:$source,
+ Variadic<Index>:$offsets,
+ Variadic<Index>:$sizes,
+ Variadic<Index>:$strides,
+ I64ArrayAttr:$static_offsets,
+ I64ArrayAttr:$static_sizes,
+ I64ArrayAttr:$static_strides
+ );
+ let results = (outs AnyMemRef:$result);
+
+ let assemblyFormat = [{
+ $source `to` `offset` `` `:`
+ custom<OperandsOrIntegersOffsetsOrStridesList>($offsets, $static_offsets)
+ `` `,` `sizes` `` `:`
+ custom<OperandsOrIntegersSizesList>($sizes, $static_sizes) `` `,` `strides`
+ `` `:`
+ custom<OperandsOrIntegersOffsetsOrStridesList>($strides, $static_strides)
+ attr-dict `:` type($source) `to` type($result)
+ }];
+
+ let parser = ?;
+ let printer = ?;
+
+ let builders = [
+ // Build a ReinterpretCastOp with mixed static and dynamic entries.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "OpFoldResult":$offset, "ArrayRef<OpFoldResult>":$sizes,
+ "ArrayRef<OpFoldResult>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a ReinterpretCastOp with static entries.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "int64_t":$offset, "ArrayRef<int64_t>":$sizes,
+ "ArrayRef<int64_t>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a ReinterpretCastOp with dynamic entries.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "Value":$offset, "ValueRange":$sizes,
+ "ValueRange":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>
+ ];
+
+ let extraClassDeclaration = extraBaseClassDeclaration # [{
+ // The result of the op is always a ranked memref.
+ MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
+ Value getViewSource() { return source(); }
+
+ /// Return the rank of the source ShapedType.
+ unsigned getResultRank() {
+ return getResult().getType().cast<ShapedType>().getRank();
+ }
+
+ /// Return the expected rank of each of the`static_offsets`, `static_sizes`
+ /// and `static_strides` attributes.
+ std::array<unsigned, 3> getArrayAttrMaxRanks() {
+ unsigned resultRank = getResult().getType().cast<ShapedType>().getRank();
+ return {1, resultRank, resultRank};
+ }
+
+ /// Return the number of leading operands before the `offsets`, `sizes` and
+ /// and `strides` operands.
+ static unsigned getOffsetSizeAndStrideStartOperandIndex() { return 1; }
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// ReshapeOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_ReshapeOp: MemRef_Op<"reshape", [
+ ViewLikeOpInterface, NoSideEffect]> {
+ let summary = "memref reshape operation";
+ let description = [{
+ The `reshape` operation converts a memref from one type to an
+ equivalent type with a provided shape. The data is never copied or
+ modified. The source and destination types are compatible if both have the
+ same element type, same number of elements, address space and identity
+ layout map. The following combinations are possible:
+
+ a. Source type is ranked or unranked. Shape argument has static size.
+ Result type is ranked.
+
+ ```mlir
+ // Reshape statically-shaped memref.
+ %dst = memref.reshape %src(%shape)
+ : (memref<4x1xf32>, memref<1xi32>) to memref<4xf32>
+ %dst0 = memref.reshape %src(%shape0)
+ : (memref<4x1xf32>, memref<2xi32>) to memref<2x2xf32>
+ // Flatten unranked memref.
+ %dst = memref.reshape %src(%shape)
+ : (memref<*xf32>, memref<1xi32>) to memref<?xf32>
+ ```
+
+ b. Source type is ranked or unranked. Shape argument has dynamic size.
+ Result type is unranked.
+
+ ```mlir
+ // Reshape dynamically-shaped 1D memref.
+ %dst = memref.reshape %src(%shape)
+ : (memref<?xf32>, memref<?xi32>) to memref<*xf32>
+ // Reshape unranked memref.
+ %dst = memref.reshape %src(%shape)
+ : (memref<*xf32>, memref<?xi32>) to memref<*xf32>
+ ```
+ }];
+
+ let arguments = (ins
+ AnyRankedOrUnrankedMemRef:$source,
+ MemRefRankOf<[AnySignlessInteger, Index], [1]>:$shape
+ );
+ let results = (outs AnyRankedOrUnrankedMemRef:$result);
+
+ let builders = [OpBuilder<
+ (ins "MemRefType":$resultType, "Value":$operand, "Value":$shape), [{
+ $_state.addOperands(operand);
+ $_state.addOperands(shape);
+ $_state.addTypes(resultType);
+ }]>];
+
+ let extraClassDeclaration = [{
+ MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
+ Value getViewSource() { return source(); }
+ }];
+
+ let assemblyFormat = [{
+ $source `(` $shape `)` attr-dict `:` functional-type(operands, results)
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// StoreOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_StoreOp : MemRef_Op<"store",
+ [TypesMatchWith<"type of 'value' matches element type of 'memref'",
+ "memref", "value",
+ "$_self.cast<MemRefType>().getElementType()">,
+ MemRefsNormalizable]> {
+ let summary = "store operation";
+ let description = [{
+ Store a value to a memref location given by indices. The value stored should
+ have the same type as the elemental type of the memref. The number of
+ arguments provided within brackets need to match the rank of the memref.
+
+ In an affine context, the indices of a store are restricted to SSA values
+ bound to surrounding loop induction variables,
+ [symbols](Affine.md#restrictions-on-dimensions-and-symbols), results of a
+ [`constant` operation](#stdconstant-constantop), or the result of an
+ [`affine.apply`](Affine.md#affineapply-affineapplyop) operation that can in
+ turn take as arguments all of the aforementioned SSA values or the
+ recursively result of such an `affine.apply` operation.
+
+ Example:
+
+ ```mlir
+ memref.store %100, %A[%1, 1023] : memref<4x?xf32, #layout, memspace0>
+ ```
+
+ **Context:** The `load` and `store` operations are specifically crafted to
+ fully resolve a reference to an element of a memref, and (in polyhedral
+ `affine.if` and `affine.for` operations) the compiler can follow use-def
+ chains (e.g. through [`affine.apply`](Affine.md#affineapply-affineapplyop)
+ operations) to precisely analyze references at compile-time using polyhedral
+ techniques. This is possible because of the
+ [restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
+ in these contexts.
+ }];
+
+ let arguments = (ins AnyType:$value,
+ Arg<AnyMemRef, "the reference to store to",
+ [MemWrite]>:$memref,
+ Variadic<Index>:$indices);
+
+ let builders = [
+ OpBuilder<(ins "Value":$valueToStore, "Value":$memref), [{
+ $_state.addOperands(valueToStore);
+ $_state.addOperands(memref);
+ }]>];
+
+ let extraClassDeclaration = [{
+ Value getValueToStore() { return getOperand(0); }
+
+ Value getMemRef() { return getOperand(1); }
+ void setMemRef(Value value) { setOperand(1, value); }
+ MemRefType getMemRefType() {
+ return getMemRef().getType().cast<MemRefType>();
+ }
+
+ operand_range getIndices() {
+ return {operand_begin() + 2, operand_end()};
+ }
+ }];
+
+ let hasFolder = 1;
+
+ let assemblyFormat = [{
+ $value `,` $memref `[` $indices `]` attr-dict `:` type($memref)
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// SubViewOp
+//===----------------------------------------------------------------------===//
+
+def SubViewOp : BaseOpWithOffsetSizesAndStrides<
+ "subview", [DeclareOpInterfaceMethods<ViewLikeOpInterface>,
+ NoSideEffect, OffsetSizeAndStrideOpInterface] > {
+ let summary = "memref subview operation";
+ let description = [{
+ The "subview" operation converts a memref type to another memref type
+ which represents a reduced-size view of the original memref as specified by
+ the operation's offsets, sizes and strides arguments.
+
+ The SubView operation supports the following arguments:
+
+ * source: the "base" memref on which to create a "view" memref.
+ * offsets: memref-rank number of offsets into the "base" memref at which to
+ create the "view" memref.
+ * sizes: memref-rank number of sizes which specify the sizes of the result
+ "view" memref type.
+ * strides: memref-rank number of strides that compose multiplicatively with
+ the base memref strides in each dimension.
+
+ The representation based on offsets, sizes and strides support a
+ partially-static specification via attributes specified through the
+ `static_offsets`, `static_sizes` and `static_strides` arguments. A special
+ sentinel value ShapedType::kDynamicSize and
+ ShapedType::kDynamicStrideOrOffset encodes that the corresponding entry has
+ a dynamic value.
+
+ A subview operation may additionally reduce the rank of the resulting view
+ by removing dimensions that are statically known to be of size 1.
+
+ Example 1:
+
+ ```mlir
+ %0 = memref.alloc() : memref<64x4xf32, (d0, d1) -> (d0 * 4 + d1)>
+
+ // Create a sub-view of "base" memref '%0' with offset arguments '%c0',
+ // dynamic sizes for each dimension, and stride arguments '%c1'.
+ %1 = memref.subview %0[%c0, %c0][%size0, %size1][%c1, %c1]
+ : memref<64x4xf32, (d0, d1) -> (d0 * 4 + d1) > to
+ memref<?x?xf32, (d0, d1)[s0, s1] -> (d0 * s1 + d1 + s0)>
+ ```
+
+ Example 2:
+
+ ```mlir
+ %0 = memref.alloc() : memref<8x16x4xf32, (d0, d1, d1) -> (d0 * 64 + d1 * 4 + d2)>
+
+ // Create a sub-view of "base" memref '%0' with dynamic offsets, sizes,
+ // and strides.
+ // Note that dynamic offsets are represented by the linearized dynamic
+ // offset symbol 's0' in the subview memref layout map, and that the
+ // dynamic strides operands, after being applied to the base memref
+ // strides in each dimension, are represented in the view memref layout
+ // map as symbols 's1', 's2' and 's3'.
+ %1 = memref.subview %0[%i, %j, %k][%size0, %size1, %size2][%x, %y, %z]
+ : memref<8x16x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)> to
+ memref<?x?x?xf32,
+ (d0, d1, d2)[s0, s1, s2, s3] -> (d0 * s1 + d1 * s2 + d2 * s3 + s0)>
+ ```
+
+ Example 3:
+
+ ```mlir
+ %0 = memref.alloc() : memref<8x16x4xf32, (d0, d1, d1) -> (d0 * 64 + d1 * 4 + d2)>
+
+ // Subview with constant offsets, sizes and strides.
+ %1 = memref.subview %0[0, 2, 0][4, 4, 4][64, 4, 1]
+ : memref<8x16x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)> to
+ memref<4x4x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2 + 8)>
+ ```
+
+ Example 4:
+
+ ```mlir
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+
+ // Subview with constant size, but dynamic offsets and
+ // strides. The resulting memref has a static shape, but if the
+ // base memref has an affine map to describe the layout, the result
+ // memref also uses an affine map to describe the layout. The
+ // strides of the result memref is computed as follows:
+ //
+ // Let #map1 represents the layout of the base memref, and #map2
+ // represents the layout of the result memref. A #mapsubview can be
+ // constructed to map an index from the result memref to the base
+ // memref (note that the description below uses more convenient
+ // naming for symbols, while in affine maps, symbols are
+ // represented as unsigned numbers that identify that symbol in the
+ // given affine map.
+ //
+ // #mapsubview = (d0, d1)[o0, o1, t0, t1] -> (d0 * t0 + o0, d1 * t1 + o1)
+ //
+ // where, o0, o1, ... are offsets, and t0, t1, ... are strides. Then,
+ //
+ // #map2 = #map1.compose(#mapsubview)
+ //
+ // If the layout map is represented as
+ //
+ // #map1 = (d0, d1)[s0, s1, s2] -> (d0 * s1 + d1 * s2 + s0)
+ //
+ // then,
+ //
+ // #map2 = (d0, d1)[s0, s1, s2, o0, o1, t0, t1] ->
+ // (d0 * s1 * t0 + d1 * s2 * t1 + o0 * s1 + o1 * s2 + s0)
+ //
+ // Representing this canonically
+ //
+ // #map2 = (d0, d1)[r0, r1, r2] -> (d0 * r1 + d1 * r2 + r0)
+ //
+ // where, r0 = o0 * s1 + o1 * s2 + s0, r1 = s1 * t0, r2 = s2 * t1.
+ %1 = memref.subview %0[%i, %j][4, 4][%x, %y] :
+ : memref<?x?xf32, (d0, d1)[s0, s1, s2] -> (d0 * s1 + d1 * s2 + s0)> to
+ memref<4x4xf32, (d0, d1)[r0, r1, r2] -> (d0 * r1 + d1 * r2 + r0)>
+
+ // Note that the subview op does not guarantee that the result
+ // memref is "inbounds" w.r.t to base memref. It is upto the client
+ // to ensure that the subview is accessed in a manner that is
+ // in-bounds.
+ ```
+
+ Example 5:
+
+ ```mlir
+ // Rank-reducing subview.
+ %1 = memref.subview %0[0, 0, 0][1, 16, 4][1, 1, 1] :
+ memref<8x16x4xf32> to memref<16x4xf32>
+ %3 = memref.subview %2[3, 4, 2][1, 6, 3][1, 1, 1] :
+ memref<8x16x4xf32> to memref<6x3xf32, offset: 210, strides: [4, 1]>
+ ```
+ }
+ }];
+
+ let arguments = (ins
+ AnyMemRef:$source,
+ Variadic<Index>:$offsets,
+ Variadic<Index>:$sizes,
+ Variadic<Index>:$strides,
+ I64ArrayAttr:$static_offsets,
+ I64ArrayAttr:$static_sizes,
+ I64ArrayAttr:$static_strides
+ );
+ let results = (outs AnyMemRef:$result);
+
+ let assemblyFormat = [{
+ $source ``
+ custom<OperandsOrIntegersOffsetsOrStridesList>($offsets, $static_offsets)
+ custom<OperandsOrIntegersSizesList>($sizes, $static_sizes)
+ custom<OperandsOrIntegersOffsetsOrStridesList>($strides, $static_strides)
+ attr-dict `:` type($source) `to` type($result)
+ }];
+
+ let builders = [
+ // Build a SubViewOp with mixed static and dynamic entries and custom
+ // result type. If the type passed is nullptr, it is inferred.
+ OpBuilder<(ins "Value":$source, "ArrayRef<OpFoldResult>":$offsets,
+ "ArrayRef<OpFoldResult>":$sizes, "ArrayRef<OpFoldResult>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a SubViewOp with mixed static and dynamic entries and inferred
+ // result type.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "ArrayRef<OpFoldResult>":$offsets, "ArrayRef<OpFoldResult>":$sizes,
+ "ArrayRef<OpFoldResult>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a SubViewOp with static entries and custom result type. If the
+ // type passed is nullptr, it is inferred.
+ OpBuilder<(ins "Value":$source, "ArrayRef<int64_t>":$offsets,
+ "ArrayRef<int64_t>":$sizes, "ArrayRef<int64_t>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a SubViewOp with static entries and inferred result type.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "ArrayRef<int64_t>":$offsets, "ArrayRef<int64_t>":$sizes,
+ "ArrayRef<int64_t>":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a SubViewOp with dynamic entries and custom result type. If the
+ // type passed is nullptr, it is inferred.
+ OpBuilder<(ins "Value":$source, "ValueRange":$offsets,
+ "ValueRange":$sizes, "ValueRange":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
+ // Build a SubViewOp with dynamic entries and inferred result type.
+ OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
+ "ValueRange":$offsets, "ValueRange":$sizes, "ValueRange":$strides,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>
+ ];
+
+ let extraClassDeclaration = extraBaseClassDeclaration # [{
+ /// Returns the type of the base memref operand.
+ MemRefType getSourceType() {
+ return source().getType().cast<MemRefType>();
+ }
+
+ /// The result of a subview is always a memref.
+ MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
+
+ /// A subview result type can be fully inferred from the source type and the
+ /// static representation of offsets, sizes and strides. Special sentinels
+ /// encode the dynamic case.
+ static Type inferResultType(MemRefType sourceMemRefType,
+ ArrayRef<int64_t> staticOffsets,
+ ArrayRef<int64_t> staticSizes,
+ ArrayRef<int64_t> staticStrides);
+ static Type inferResultType(MemRefType sourceMemRefType,
+ ArrayRef<OpFoldResult> staticOffsets,
+ ArrayRef<OpFoldResult> staticSizes,
+ ArrayRef<OpFoldResult> staticStrides);
+ static Type inferRankReducedResultType(unsigned resultRank,
+ MemRefType sourceMemRefType,
+ ArrayRef<int64_t> staticOffsets,
+ ArrayRef<int64_t> staticSizes,
+ ArrayRef<int64_t> staticStrides);
+ static Type inferRankReducedResultType(unsigned resultRank,
+ MemRefType sourceMemRefType,
+ ArrayRef<OpFoldResult> staticOffsets,
+ ArrayRef<OpFoldResult> staticSizes,
+ ArrayRef<OpFoldResult> staticStrides);
+
+ /// Return the expected rank of each of the`static_offsets`, `static_sizes`
+ /// and `static_strides` attributes.
+ std::array<unsigned, 3> getArrayAttrMaxRanks() {
+ unsigned rank = getSourceType().getRank();
+ return {rank, rank, rank};
+ }
+
+ /// Return the number of leading operands before the `offsets`, `sizes` and
+ /// and `strides` operands.
+ static unsigned getOffsetSizeAndStrideStartOperandIndex() { return 1; }
+ }];
+
+ let hasCanonicalizer = 1;
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// TensorLoadOp
+//===----------------------------------------------------------------------===//
+
+def TensorLoadOp : MemRef_Op<"tensor_load",
+ [SameOperandsAndResultShape, SameOperandsAndResultElementType,
+ TypesMatchWith<"result type matches tensor equivalent of 'memref'",
+ "memref", "result",
+ "getTensorTypeFromMemRefType($_self)">]> {
+ let summary = "tensor load operation";
+ let description = [{
+ Create a tensor from a memref, making an independent copy of the element
+ data. The result value is a tensor whose shape and element type match the
+ memref operand.
+
+ The opposite of this op is buffer_cast. Together, these two ops are
+ useful for source/target materializations when doing type conversions
+ involving tensors and memrefs.
+
+ Example:
+
+ ```mlir
+ // Produces a value of tensor<4x?xf32> type.
+ %12 = memref.tensor_load %10 : memref<4x?xf32, #layout, memspace0>
+ ```
+ }];
+
+ let arguments = (ins Arg<AnyRankedOrUnrankedMemRef,
+ "the reference to load from", [MemRead]>:$memref);
+ let results = (outs AnyTensor:$result);
+ // TensorLoadOp is fully verified by traits.
+ let verifier = ?;
+
+ let builders = [
+ OpBuilder<(ins "Value":$memref), [{
+ $_state.addOperands(memref);
+ $_state.addTypes(getTensorTypeFromMemRefType(memref.getType()));
+ }]>];
+
+ let extraClassDeclaration = [{
+ /// The result of a tensor_load is always a tensor.
+ TensorType getType() {
+ Type resultType = getResult().getType();
+ if (resultType.isa<TensorType>())
+ return resultType.cast<TensorType>();
+ return {};
+ }
+ }];
+
+ let assemblyFormat = "$memref attr-dict `:` type($memref)";
+
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// TensorStoreOp
+//===----------------------------------------------------------------------===//
+
+def TensorStoreOp : MemRef_Op<"tensor_store",
+ [SameOperandsShape, SameOperandsElementType,
+ TypesMatchWith<"type of 'value' matches tensor equivalent of 'memref'",
+ "memref", "tensor",
+ "getTensorTypeFromMemRefType($_self)">]> {
+ let summary = "tensor store operation";
+ let description = [{
+ Stores the contents of a tensor into a memref. The first operand is a value
+ of tensor type, the second operand is a value of memref type. The shapes and
+ element types of these must match, and are specified by the memref type.
+
+ Example:
+
+ ```mlir
+ %9 = dim %8, 1 : tensor<4x?xf32>
+ %10 = alloc(%9) : memref<4x?xf32, #layout, memspace0>
+ memref.tensor_store %8, %10 : memref<4x?xf32, #layout, memspace0>
+ ```
+ }];
+
+ let arguments = (ins AnyTensor:$tensor, Arg<AnyRankedOrUnrankedMemRef,
+ "the reference to store to", [MemWrite]>:$memref);
+ // TensorStoreOp is fully verified by traits.
+ let verifier = ?;
+
+ let assemblyFormat = "$tensor `,` $memref attr-dict `:` type($memref)";
+}
+
+//===----------------------------------------------------------------------===//
+// TransposeOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_TransposeOp : MemRef_Op<"transpose", [NoSideEffect]>,
+ Arguments<(ins AnyStridedMemRef:$in, AffineMapAttr:$permutation)>,
+ Results<(outs AnyStridedMemRef)> {
+ let summary = "`transpose` produces a new strided memref (metadata-only)";
+ let description = [{
+ The `transpose` op produces a strided memref whose sizes and strides
+ are a permutation of the original `in` memref. This is purely a metadata
+ transformation.
+
+ Example:
+
+ ```mlir
+ %1 = memref.transpose %0 (i, j) -> (j, i) : memref<?x?xf32> to memref<?x?xf32, affine_map<(d0, d1)[s0] -> (d1 * s0 + d0)>>
+ ```
+ }];
+
+ let builders = [
+ OpBuilder<(ins "Value":$in, "AffineMapAttr":$permutation,
+ CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>];
+
+ let extraClassDeclaration = [{
+ static StringRef getPermutationAttrName() { return "permutation"; }
+ ShapedType getShapedType() { return in().getType().cast<ShapedType>(); }
+ }];
+
+ let hasFolder = 1;
+}
+
+//===----------------------------------------------------------------------===//
+// ViewOp
+//===----------------------------------------------------------------------===//
+
+def MemRef_ViewOp : MemRef_Op<"view", [
+ DeclareOpInterfaceMethods<ViewLikeOpInterface>, NoSideEffect]> {
+ let summary = "memref view operation";
+ let description = [{
+ The "view" operation extracts an N-D contiguous memref with empty layout map
+ with arbitrary element type from a 1-D contiguous memref with empty layout
+ map of i8 element type. The ViewOp supports the following arguments:
+
+ * A single dynamic byte-shift operand must be specified which represents a
+ a shift of the base 1-D memref pointer from which to create the resulting
+ contiguous memref view with identity layout.
+ * A dynamic size operand that must be specified for each dynamic dimension
+ in the resulting view memref type.
+
+ The "view" operation gives a structured indexing form to a flat 1-D buffer.
+ Unlike "subview" it can perform a type change. The type change behavior
+ requires the op to have special semantics because, e.g. a byte shift of 3
+ cannot be represented as an offset on f64.
+ For now, a "view" op:
+
+ 1. Only takes a contiguous source memref with 0 offset and empty layout.
+ 2. Must specify a byte_shift operand (in the future, a special integer
+ attribute may be added to support the folded case).
+ 3. Returns a contiguous memref with 0 offset and empty layout.
+
+ Example:
+
+ ```mlir
+ // Allocate a flat 1D/i8 memref.
+ %0 = memref.alloc() : memref<2048xi8>
+
+ // ViewOp with dynamic offset and static sizes.
+ %1 = memref.view %0[%offset_1024][] : memref<2048xi8> to memref<64x4xf32>
+
+ // ViewOp with dynamic offset and two dynamic size.
+ %2 = memref.view %0[%offset_1024][%size0, %size1] :
+ memref<2048xi8> to memref<?x4x?xf32>
+ ```
+ }];
+
+ let arguments = (ins MemRefRankOf<[I8], [1]>:$source,
+ Index:$byte_shift,
+ Variadic<Index>:$sizes);
+ let results = (outs AnyMemRef);
+
+ let extraClassDeclaration = [{
+ /// The result of a view is always a memref.
+ MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
+
+ /// Returns the dynamic sizes for this view operation. This is redundant
+ /// with `sizes` but needed in template implementations. More specifically:
+ /// ```
+ /// template <typename AnyMemRefDefOp>
+ /// bool isMemRefSizeValidSymbol(AnyMemRefDefOp memrefDefOp, unsigned index,
+ /// Region *region)
+ /// ```
+ operand_range getDynamicSizes() {
+ return {sizes().begin(), sizes().end()};
+ }
+ }];
+
+ let hasCanonicalizer = 1;
+}
+
+#endif // MEMREF_OPS
diff --git a/mlir/include/mlir/Dialect/SCF/Passes.td b/mlir/include/mlir/Dialect/SCF/Passes.td
index 611869466214..458cf769bfc9 100644
--- a/mlir/include/mlir/Dialect/SCF/Passes.td
+++ b/mlir/include/mlir/Dialect/SCF/Passes.td
@@ -14,6 +14,7 @@ include "mlir/Pass/PassBase.td"
def SCFBufferize : FunctionPass<"scf-bufferize"> {
let summary = "Bufferize the scf dialect.";
let constructor = "mlir::createSCFBufferizePass()";
+ let dependentDialects = ["memref::MemRefDialect"];
}
def SCFForLoopSpecialization
diff --git a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
index c4ccdbd0c2d3..685f0d0da120 100644
--- a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
@@ -25,5 +25,6 @@ def ShapeToShapeLowering : FunctionPass<"shape-to-shape-lowering"> {
def ShapeBufferize : FunctionPass<"shape-bufferize"> {
let summary = "Bufferize the shape dialect.";
let constructor = "mlir::createShapeBufferizePass()";
+ let dependentDialects = ["memref::MemRefDialect"];
}
#endif // MLIR_DIALECT_SHAPE_TRANSFORMS_PASSES
diff --git a/mlir/include/mlir/Dialect/StandardOps/EDSC/Intrinsics.h b/mlir/include/mlir/Dialect/StandardOps/EDSC/Intrinsics.h
index 8ba9dcc3c030..d100cf7d9bc2 100644
--- a/mlir/include/mlir/Dialect/StandardOps/EDSC/Intrinsics.h
+++ b/mlir/include/mlir/Dialect/StandardOps/EDSC/Intrinsics.h
@@ -17,35 +17,24 @@ namespace intrinsics {
using std_addi = ValueBuilder<AddIOp>;
using std_addf = ValueBuilder<AddFOp>;
-using std_alloc = ValueBuilder<AllocOp>;
-using std_alloca = ValueBuilder<AllocaOp>;
using std_call = OperationBuilder<CallOp>;
using std_constant = ValueBuilder<ConstantOp>;
using std_constant_float = ValueBuilder<ConstantFloatOp>;
using std_constant_index = ValueBuilder<ConstantIndexOp>;
using std_constant_int = ValueBuilder<ConstantIntOp>;
-using std_dealloc = OperationBuilder<DeallocOp>;
using std_divis = ValueBuilder<SignedDivIOp>;
using std_diviu = ValueBuilder<UnsignedDivIOp>;
-using std_dim = ValueBuilder<DimOp>;
using std_fpext = ValueBuilder<FPExtOp>;
using std_fptrunc = ValueBuilder<FPTruncOp>;
using std_index_cast = ValueBuilder<IndexCastOp>;
using std_muli = ValueBuilder<MulIOp>;
using std_mulf = ValueBuilder<MulFOp>;
-using std_memref_cast = ValueBuilder<MemRefCastOp>;
using std_ret = OperationBuilder<ReturnOp>;
using std_select = ValueBuilder<SelectOp>;
-using std_load = ValueBuilder<LoadOp>;
using std_sign_extendi = ValueBuilder<SignExtendIOp>;
using std_splat = ValueBuilder<SplatOp>;
-using std_store = OperationBuilder<StoreOp>;
using std_subf = ValueBuilder<SubFOp>;
using std_subi = ValueBuilder<SubIOp>;
-using std_sub_view = ValueBuilder<SubViewOp>;
-using std_tensor_load = ValueBuilder<TensorLoadOp>;
-using std_tensor_store = OperationBuilder<TensorStoreOp>;
-using std_view = ValueBuilder<ViewOp>;
using std_zero_extendi = ValueBuilder<ZeroExtendIOp>;
using tensor_extract = ValueBuilder<tensor::ExtractOp>;
@@ -77,10 +66,6 @@ BranchOp std_br(Block *block, ValueRange operands);
/// or to `falseBranch` and `falseOperand` if `cond` evaluates to `false`.
CondBranchOp std_cond_br(Value cond, Block *trueBranch, ValueRange trueOperands,
Block *falseBranch, ValueRange falseOperands);
-
-/// Provide an index notation around sdt_load and std_store.
-using StdIndexedValue =
- TemplatedIndexedValue<intrinsics::std_load, intrinsics::std_store>;
} // namespace intrinsics
} // namespace edsc
} // namespace mlir
diff --git a/mlir/include/mlir/Dialect/StandardOps/IR/Ops.h b/mlir/include/mlir/Dialect/StandardOps/IR/Ops.h
index 241f4ed9fa84..f6f8d46cef28 100644
--- a/mlir/include/mlir/Dialect/StandardOps/IR/Ops.h
+++ b/mlir/include/mlir/Dialect/StandardOps/IR/Ops.h
@@ -34,8 +34,6 @@ class Builder;
class FuncOp;
class OpBuilder;
-raw_ostream &operator<<(raw_ostream &os, Range &range);
-
/// Return the list of Range (i.e. offset, size, stride). Each Range
/// entry contains either the dynamic value or a ConstantIndexOp constructed
/// with `b` at location `loc`.
@@ -110,200 +108,6 @@ class ConstantIndexOp : public ConstantOp {
static bool classof(Operation *op);
};
-// DmaStartOp starts a non-blocking DMA operation that transfers data from a
-// source memref to a destination memref. The source and destination memref need
-// not be of the same dimensionality, but need to have the same elemental type.
-// The operands include the source and destination memref's each followed by its
-// indices, size of the data transfer in terms of the number of elements (of the
-// elemental type of the memref), a tag memref with its indices, and optionally
-// at the end, a stride and a number_of_elements_per_stride arguments. The tag
-// location is used by a DmaWaitOp to check for completion. The indices of the
-// source memref, destination memref, and the tag memref have the same
-// restrictions as any load/store. The optional stride arguments should be of
-// 'index' type, and specify a stride for the slower memory space (memory space
-// with a lower memory space id), transferring chunks of
-// number_of_elements_per_stride every stride until %num_elements are
-// transferred. Either both or no stride arguments should be specified.
-//
-// For example, a DmaStartOp operation that transfers 256 elements of a memref
-// '%src' in memory space 0 at indices [%i, %j] to memref '%dst' in memory space
-// 1 at indices [%k, %l], would be specified as follows:
-//
-// %num_elements = constant 256
-// %idx = constant 0 : index
-// %tag = alloc() : memref<1 x i32, (d0) -> (d0), 4>
-// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%idx] :
-// memref<40 x 128 x f32>, (d0) -> (d0), 0>,
-// memref<2 x 1024 x f32>, (d0) -> (d0), 1>,
-// memref<1 x i32>, (d0) -> (d0), 2>
-//
-// If %stride and %num_elt_per_stride are specified, the DMA is expected to
-// transfer %num_elt_per_stride elements every %stride elements apart from
-// memory space 0 until %num_elements are transferred.
-//
-// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%idx], %stride,
-// %num_elt_per_stride :
-//
-// TODO: add additional operands to allow source and destination striding, and
-// multiple stride levels.
-// TODO: Consider replacing src/dst memref indices with view memrefs.
-class DmaStartOp
- : public Op<DmaStartOp, OpTrait::VariadicOperands, OpTrait::ZeroResult> {
-public:
- using Op::Op;
-
- static void build(OpBuilder &builder, OperationState &result, Value srcMemRef,
- ValueRange srcIndices, Value destMemRef,
- ValueRange destIndices, Value numElements, Value tagMemRef,
- ValueRange tagIndices, Value stride = nullptr,
- Value elementsPerStride = nullptr);
-
- // Returns the source MemRefType for this DMA operation.
- Value getSrcMemRef() { return getOperand(0); }
- // Returns the rank (number of indices) of the source MemRefType.
- unsigned getSrcMemRefRank() {
- return getSrcMemRef().getType().cast<MemRefType>().getRank();
- }
- // Returns the source memref indices for this DMA operation.
- operand_range getSrcIndices() {
- return {(*this)->operand_begin() + 1,
- (*this)->operand_begin() + 1 + getSrcMemRefRank()};
- }
-
- // Returns the destination MemRefType for this DMA operations.
- Value getDstMemRef() { return getOperand(1 + getSrcMemRefRank()); }
- // Returns the rank (number of indices) of the destination MemRefType.
- unsigned getDstMemRefRank() {
- return getDstMemRef().getType().cast<MemRefType>().getRank();
- }
- unsigned getSrcMemorySpace() {
- return getSrcMemRef().getType().cast<MemRefType>().getMemorySpaceAsInt();
- }
- unsigned getDstMemorySpace() {
- return getDstMemRef().getType().cast<MemRefType>().getMemorySpaceAsInt();
- }
-
- // Returns the destination memref indices for this DMA operation.
- operand_range getDstIndices() {
- return {(*this)->operand_begin() + 1 + getSrcMemRefRank() + 1,
- (*this)->operand_begin() + 1 + getSrcMemRefRank() + 1 +
- getDstMemRefRank()};
- }
-
- // Returns the number of elements being transferred by this DMA operation.
- Value getNumElements() {
- return getOperand(1 + getSrcMemRefRank() + 1 + getDstMemRefRank());
- }
-
- // Returns the Tag MemRef for this DMA operation.
- Value getTagMemRef() {
- return getOperand(1 + getSrcMemRefRank() + 1 + getDstMemRefRank() + 1);
- }
- // Returns the rank (number of indices) of the tag MemRefType.
- unsigned getTagMemRefRank() {
- return getTagMemRef().getType().cast<MemRefType>().getRank();
- }
-
- // Returns the tag memref index for this DMA operation.
- operand_range getTagIndices() {
- unsigned tagIndexStartPos =
- 1 + getSrcMemRefRank() + 1 + getDstMemRefRank() + 1 + 1;
- return {(*this)->operand_begin() + tagIndexStartPos,
- (*this)->operand_begin() + tagIndexStartPos + getTagMemRefRank()};
- }
-
- /// Returns true if this is a DMA from a faster memory space to a slower one.
- bool isDestMemorySpaceFaster() {
- return (getSrcMemorySpace() < getDstMemorySpace());
- }
-
- /// Returns true if this is a DMA from a slower memory space to a faster one.
- bool isSrcMemorySpaceFaster() {
- // Assumes that a lower number is for a slower memory space.
- return (getDstMemorySpace() < getSrcMemorySpace());
- }
-
- /// Given a DMA start operation, returns the operand position of either the
- /// source or destination memref depending on the one that is at the higher
- /// level of the memory hierarchy. Asserts failure if neither is true.
- unsigned getFasterMemPos() {
- assert(isSrcMemorySpaceFaster() || isDestMemorySpaceFaster());
- return isSrcMemorySpaceFaster() ? 0 : getSrcMemRefRank() + 1;
- }
-
- static StringRef getOperationName() { return "std.dma_start"; }
- static ParseResult parse(OpAsmParser &parser, OperationState &result);
- void print(OpAsmPrinter &p);
- LogicalResult verify();
-
- LogicalResult fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results);
-
- bool isStrided() {
- return getNumOperands() != 1 + getSrcMemRefRank() + 1 + getDstMemRefRank() +
- 1 + 1 + getTagMemRefRank();
- }
-
- Value getStride() {
- if (!isStrided())
- return nullptr;
- return getOperand(getNumOperands() - 1 - 1);
- }
-
- Value getNumElementsPerStride() {
- if (!isStrided())
- return nullptr;
- return getOperand(getNumOperands() - 1);
- }
-};
-
-// DmaWaitOp blocks until the completion of a DMA operation associated with the
-// tag element '%tag[%index]'. %tag is a memref, and %index has to be an index
-// with the same restrictions as any load/store index. %num_elements is the
-// number of elements associated with the DMA operation. For example:
-//
-// dma_start %src[%i, %j], %dst[%k, %l], %num_elements, %tag[%index] :
-// memref<2048 x f32>, (d0) -> (d0), 0>,
-// memref<256 x f32>, (d0) -> (d0), 1>
-// memref<1 x i32>, (d0) -> (d0), 2>
-// ...
-// ...
-// dma_wait %tag[%index], %num_elements : memref<1 x i32, (d0) -> (d0), 2>
-//
-class DmaWaitOp
- : public Op<DmaWaitOp, OpTrait::VariadicOperands, OpTrait::ZeroResult> {
-public:
- using Op::Op;
-
- static void build(OpBuilder &builder, OperationState &result, Value tagMemRef,
- ValueRange tagIndices, Value numElements);
-
- static StringRef getOperationName() { return "std.dma_wait"; }
-
- // Returns the Tag MemRef associated with the DMA operation being waited on.
- Value getTagMemRef() { return getOperand(0); }
-
- // Returns the tag memref index for this DMA operation.
- operand_range getTagIndices() {
- return {(*this)->operand_begin() + 1,
- (*this)->operand_begin() + 1 + getTagMemRefRank()};
- }
-
- // Returns the rank (number of indices) of the tag memref.
- unsigned getTagMemRefRank() {
- return getTagMemRef().getType().cast<MemRefType>().getRank();
- }
-
- // Returns the number of elements transferred in the associated DMA operation.
- Value getNumElements() { return getOperand(1 + getTagMemRefRank()); }
-
- static ParseResult parse(OpAsmParser &parser, OperationState &result);
- void print(OpAsmPrinter &p);
- LogicalResult fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results);
- LogicalResult verify();
-};
-
/// Given an `originalShape` and a `reducedShape` assumed to be a subset of
/// `originalShape` with some `1` entries erased, return the set of indices
/// that specifies which of the entries of `originalShape` are dropped to obtain
@@ -316,45 +120,6 @@ llvm::Optional<llvm::SmallDenseSet<unsigned>>
computeRankReductionMask(ArrayRef<int64_t> originalShape,
ArrayRef<int64_t> reducedShape);
-/// Determines whether MemRefCastOp casts to a more dynamic version of the
-/// source memref. This is useful to to fold a memref_cast into a consuming op
-/// and implement canonicalization patterns for ops in
diff erent dialects that
-/// may consume the results of memref_cast operations. Such foldable memref_cast
-/// operations are typically inserted as `view` and `subview` ops and are
-/// canonicalized, to preserve the type compatibility of their uses.
-///
-/// Returns true when all conditions are met:
-/// 1. source and result are ranked memrefs with strided semantics and same
-/// element type and rank.
-/// 2. each of the source's size, offset or stride has more static information
-/// than the corresponding result's size, offset or stride.
-///
-/// Example 1:
-/// ```mlir
-/// %1 = memref_cast %0 : memref<8x16xf32> to memref<?x?xf32>
-/// %2 = consumer %1 ... : memref<?x?xf32> ...
-/// ```
-///
-/// may fold into:
-///
-/// ```mlir
-/// %2 = consumer %0 ... : memref<8x16xf32> ...
-/// ```
-///
-/// Example 2:
-/// ```
-/// %1 = memref_cast %0 : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
-/// to memref<?x?xf32>
-/// consumer %1 : memref<?x?xf32> ...
-/// ```
-///
-/// may fold into:
-///
-/// ```
-/// consumer %0 ... : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
-/// ```
-bool canFoldIntoConsumerOp(MemRefCastOp castOp);
-
/// Compute `lhs` `pred` `rhs`, where `pred` is one of the known integer
/// comparison predicates.
bool applyCmpPredicate(CmpIPredicate predicate, const APInt &lhs,
diff --git a/mlir/include/mlir/Dialect/StandardOps/IR/Ops.td b/mlir/include/mlir/Dialect/StandardOps/IR/Ops.td
index 47d2bf12e4db..403335ad90e5 100644
--- a/mlir/include/mlir/Dialect/StandardOps/IR/Ops.td
+++ b/mlir/include/mlir/Dialect/StandardOps/IR/Ops.td
@@ -43,13 +43,13 @@ class Std_Op<string mnemonic, list<OpTrait> traits = []> :
let parser = [{ return ::parse$cppClass(parser, result); }];
}
-// Base class for standard cast operations. Requires single operand and result,
-// but does not constrain them to specific types.
-class CastOp<string mnemonic, list<OpTrait> traits = []> :
- Std_Op<mnemonic, traits # [
+// Base class for arithmetic cast operations. Requires single operand and
+// result, but does not constrain them to specific types.
+class ArithmeticCastOp<string mnemonic, list<OpTrait> traits = []> :
+ Std_Op<mnemonic, traits # ElementwiseMappable.traits # [
+ DeclareOpInterfaceMethods<VectorUnrollOpInterface>,
NoSideEffect, SameOperandsAndResultShape,
- DeclareOpInterfaceMethods<CastOpInterface>
- ]> {
+ DeclareOpInterfaceMethods<CastOpInterface>]> {
let results = (outs AnyType);
let builders = [
@@ -69,13 +69,6 @@ class CastOp<string mnemonic, list<OpTrait> traits = []> :
let verifier = ?;
}
-// Base class for arithmetic cast operations.
-class ArithmeticCastOp<string mnemonic, list<OpTrait> traits = []> :
- CastOp<mnemonic, traits # [
- DeclareOpInterfaceMethods<VectorUnrollOpInterface>] #
- ElementwiseMappable.traits> {
-}
-
// Base class for unary ops. Requires single operand and result. Individual
// classes will have `operand` accessor.
class UnaryOp<string mnemonic, list<OpTrait> traits = []> :
@@ -189,64 +182,6 @@ class FloatTernaryOp<string mnemonic, list<OpTrait> traits = []> :
[DeclareOpInterfaceMethods<VectorUnrollOpInterface>])>,
Arguments<(ins FloatLike:$a, FloatLike:$b, FloatLike:$c)>;
-// Base class for memref allocating ops: alloca and alloc.
-//
-// %0 = alloclike(%m)[%s] : memref<8x?xf32, (d0, d1)[s0] -> ((d0 + s0), d1)>
-//
-class AllocLikeOp<string mnemonic,
- Resource resource,
- list<OpTrait> traits = []> :
- Std_Op<mnemonic,
- !listconcat([
- AttrSizedOperandSegments
- ], traits)> {
-
- let arguments = (ins Variadic<Index>:$dynamicSizes,
- // The symbolic operands (the ones in square brackets) bind
- // to the symbols of the memref's layout map.
- Variadic<Index>:$symbolOperands,
- Confined<OptionalAttr<I64Attr>, [IntMinValue<0>]>:$alignment);
- let results = (outs Res<AnyMemRef, "", [MemAlloc<resource>]>:$memref);
-
- let builders = [
- OpBuilder<(ins "MemRefType":$memrefType,
- CArg<"IntegerAttr", "IntegerAttr()">:$alignment), [{
- return build($_builder, $_state, memrefType, {}, alignment);
- }]>,
- OpBuilder<(ins "MemRefType":$memrefType, "ValueRange":$dynamicSizes,
- CArg<"IntegerAttr", "IntegerAttr()">:$alignment), [{
- return build($_builder, $_state, memrefType, dynamicSizes, {}, alignment);
- }]>,
- OpBuilder<(ins "MemRefType":$memrefType, "ValueRange":$dynamicSizes,
- "ValueRange":$symbolOperands,
- CArg<"IntegerAttr", "{}">:$alignment), [{
- $_state.types.push_back(memrefType);
- $_state.addOperands(dynamicSizes);
- $_state.addOperands(symbolOperands);
- $_state.addAttribute(getOperandSegmentSizeAttr(),
- $_builder.getI32VectorAttr({
- static_cast<int32_t>(dynamicSizes.size()),
- static_cast<int32_t>(symbolOperands.size())}));
- if (alignment)
- $_state.addAttribute(getAlignmentAttrName(), alignment);
- }]>];
-
- let extraClassDeclaration = [{
- static StringRef getAlignmentAttrName() { return "alignment"; }
-
- MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
-
- /// Returns the dynamic sizes for this alloc operation if specified.
- operand_range getDynamicSizes() { return dynamicSizes(); }
- }];
-
- let assemblyFormat = [{
- `(`$dynamicSizes`)` (`` `[` $symbolOperands^ `]`)? attr-dict `:` type($memref)
- }];
-
- let hasCanonicalizer = 1;
-}
-
// Base class for ops with static/dynamic offset, sizes and strides
// attributes/arguments.
class BaseOpWithOffsetSizesAndStrides<string mnemonic, list<OpTrait> traits = []> :
@@ -362,96 +297,6 @@ def AddIOp : IntBinaryOp<"addi", [Commutative]> {
let hasFolder = 1;
}
-//===----------------------------------------------------------------------===//
-// AllocOp
-//===----------------------------------------------------------------------===//
-
-def AllocOp : AllocLikeOp<"alloc", DefaultResource> {
- let summary = "memory allocation operation";
- let description = [{
- The `alloc` operation allocates a region of memory, as specified by its
- memref type.
-
- Example:
-
- ```mlir
- %0 = alloc() : memref<8x64xf32, 1>
- ```
-
- The optional list of dimension operands are bound to the dynamic dimensions
- specified in its memref type. In the example below, the ssa value '%d' is
- bound to the second dimension of the memref (which is dynamic).
-
- ```mlir
- %0 = alloc(%d) : memref<8x?xf32, 1>
- ```
-
- The optional list of symbol operands are bound to the symbols of the
- memrefs affine map. In the example below, the ssa value '%s' is bound to
- the symbol 's0' in the affine map specified in the allocs memref type.
-
- ```mlir
- %0 = alloc()[%s] : memref<8x64xf32,
- affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
- ```
-
- This operation returns a single ssa value of memref type, which can be used
- by subsequent load and store operations.
-
- The optional `alignment` attribute may be specified to ensure that the
- region of memory that will be indexed is aligned at the specified byte
- boundary.
-
- ```mlir
- %0 = alloc()[%s] {alignment = 8} :
- memref<8x64xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
- ```
- }];
-}
-
-//===----------------------------------------------------------------------===//
-// AllocaOp
-//===----------------------------------------------------------------------===//
-
-def AllocaOp : AllocLikeOp<"alloca", AutomaticAllocationScopeResource> {
- let summary = "stack memory allocation operation";
- let description = [{
- The `alloca` operation allocates memory on the stack, to be automatically
- released when control transfers back from the region of its closest
- surrounding operation with an
- [`AutomaticAllocationScope`](../Traits.md#automaticallocationscope) trait.
- The amount of memory allocated is specified by its memref and additional
- operands. For example:
-
- ```mlir
- %0 = alloca() : memref<8x64xf32>
- ```
-
- The optional list of dimension operands are bound to the dynamic dimensions
- specified in its memref type. In the example below, the SSA value '%d' is
- bound to the second dimension of the memref (which is dynamic).
-
- ```mlir
- %0 = alloca(%d) : memref<8x?xf32>
- ```
-
- The optional list of symbol operands are bound to the symbols of the
- memref's affine map. In the example below, the SSA value '%s' is bound to
- the symbol 's0' in the affine map specified in the allocs memref type.
-
- ```mlir
- %0 = alloca()[%s] : memref<8x64xf32,
- affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>>
- ```
-
- This operation returns a single SSA value of memref type, which can be used
- by subsequent load and store operations. An optional alignment attribute, if
- specified, guarantees alignment at least to that boundary. If not specified,
- an alignment on any convenient boundary compatible with the type will be
- chosen.
- }];
-}
-
//===----------------------------------------------------------------------===//
// AndOp
//===----------------------------------------------------------------------===//
@@ -515,28 +360,6 @@ def AssertOp : Std_Op<"assert"> {
let hasCanonicalizer = 1;
}
-//===----------------------------------------------------------------------===//
-// AssumeAlignmentOp
-//===----------------------------------------------------------------------===//
-
-def AssumeAlignmentOp : Std_Op<"assume_alignment"> {
- let summary =
- "assertion that gives alignment information to the input memref";
- let description = [{
- The `assume_alignment` operation takes a memref and an integer of alignment
- value, and internally annotates the buffer with the given alignment. If
- the buffer isn't aligned to the given alignment, the behavior is undefined.
-
- This operation doesn't affect the semantics of a correct program. It's for
- optimization only, and the optimization is best-effort.
- }];
- let arguments = (ins AnyMemRef:$memref,
- Confined<I32Attr, [IntPositive]>:$alignment);
- let results = (outs);
-
- let assemblyFormat = "$memref `,` $alignment attr-dict `:` type($memref)";
-}
-
//===----------------------------------------------------------------------===//
// AtomicRMWOp
//===----------------------------------------------------------------------===//
@@ -1333,87 +1156,6 @@ def CopySignOp : FloatBinaryOp<"copysign"> {
}];
}
-//===----------------------------------------------------------------------===//
-// DeallocOp
-//===----------------------------------------------------------------------===//
-
-def DeallocOp : Std_Op<"dealloc", [MemRefsNormalizable]> {
- let summary = "memory deallocation operation";
- let description = [{
- The `dealloc` operation frees the region of memory referenced by a memref
- which was originally created by the `alloc` operation.
- The `dealloc` operation should not be called on memrefs which alias an
- alloc'd memref (e.g. memrefs returned by `view` operations).
-
- Example:
-
- ```mlir
- %0 = alloc() : memref<8x64xf32, (d0, d1) -> (d0, d1), 1>
- dealloc %0 : memref<8x64xf32, (d0, d1) -> (d0, d1), 1>
- ```
- }];
-
- let arguments = (ins Arg<AnyMemRef, "", [MemFree]>:$memref);
-
- let hasCanonicalizer = 1;
- let hasFolder = 1;
- let assemblyFormat = "$memref attr-dict `:` type($memref)";
-}
-
-//===----------------------------------------------------------------------===//
-// DimOp
-//===----------------------------------------------------------------------===//
-
-def DimOp : Std_Op<"dim", [NoSideEffect]> {
- let summary = "dimension index operation";
- let description = [{
- The `dim` operation takes a memref/tensor and a dimension operand of type
- `index`.
- It returns the size of the requested dimension of the given memref/tensor.
- If the dimension index is out of bounds the behavior is undefined.
-
- The specified memref or tensor type is that of the first operand.
-
- Example:
-
- ```mlir
- // Always returns 4, can be constant folded:
- %c0 = constant 0 : index
- %x = = dim %A, %c0 : tensor<4 x ? x f32>
-
- // Returns the dynamic dimension of %A.
- %c1 = constant 1 : index
- %y = dim %A, %c1 : tensor<4 x ? x f32>
-
- // Equivalent generic form:
- %x = "std.dim"(%A, %c0) : (tensor<4 x ? x f32>, index) -> index
- %y = "std.dim"(%A, %c1) : (tensor<4 x ? x f32>, index) -> index
- ```
- }];
-
- let arguments = (ins AnyTypeOf<[AnyRankedOrUnrankedMemRef, AnyTensor],
- "any tensor or memref type">:$memrefOrTensor,
- Index:$index);
- let results = (outs Index:$result);
-
- let assemblyFormat = [{
- attr-dict $memrefOrTensor `,` $index `:` type($memrefOrTensor)
- }];
-
- let builders = [
- OpBuilder<(ins "Value":$memrefOrTensor, "int64_t":$index)>,
- OpBuilder<(ins "Value":$memrefOrTensor, "Value":$index)>
- ];
-
- let extraClassDeclaration = [{
- /// Helper function to get the index as a simple integer if it is constant.
- Optional<int64_t> getConstantIndex();
- }];
-
- let hasCanonicalizer = 1;
- let hasFolder = 1;
-}
-
//===----------------------------------------------------------------------===//
// DivFOp
//===----------------------------------------------------------------------===//
@@ -1512,98 +1254,6 @@ def FPTruncOp : ArithmeticCastOp<"fptrunc">, Arguments<(ins AnyType:$in)> {
}];
}
-//===----------------------------------------------------------------------===//
-// GlobalMemrefOp
-//===----------------------------------------------------------------------===//
-
-def GlobalMemrefOp : Std_Op<"global_memref", [Symbol]> {
- let summary = "declare or define a global memref variable";
- let description = [{
- The `global_memref` operation declares or defines a named global variable.
- The backing memory for the variable is allocated statically and is described
- by the type of the variable (which should be a statically shaped memref
- type). The operation is a declaration if no `inital_value` is specified,
- else it is a definition. The `initial_value` can either be a unit attribute
- to represent a definition of an uninitialized global variable, or an
- elements attribute to represent the definition of a global variable with an
- initial value. The global variable can also be marked constant using the
- `constant` unit attribute. Writing to such constant global variables is
- undefined.
-
- The global variable can be accessed by using the `get_global_memref` to
- retrieve the memref for the global variable. Note that the memref
- for such global variable itself is immutable (i.e., get_global_memref for a
- given global variable will always return the same memref descriptor).
-
- Example:
-
- ```mlir
- // Private variable with an initial value.
- global_memref "private" @x : memref<2xf32> = dense<0.0,2.0>
-
- // Declaration of an external variable.
- global_memref "private" @y : memref<4xi32>
-
- // Uninitialized externally visible variable.
- global_memref @z : memref<3xf16> = uninitialized
-
- // Externally visible constant variable.
- global_memref constant @c : memref<2xi32> = dense<1, 4>
- ```
- }];
-
- let arguments = (ins
- SymbolNameAttr:$sym_name,
- OptionalAttr<StrAttr>:$sym_visibility,
- TypeAttr:$type,
- OptionalAttr<AnyAttr>:$initial_value,
- UnitAttr:$constant
- );
-
- let assemblyFormat = [{
- ($sym_visibility^)?
- (`constant` $constant^)?
- $sym_name `:`
- custom<GlobalMemrefOpTypeAndInitialValue>($type, $initial_value)
- attr-dict
- }];
-
- let extraClassDeclaration = [{
- bool isExternal() { return !initial_value(); }
- bool isUninitialized() {
- return !isExternal() && initial_value().getValue().isa<UnitAttr>();
- }
- }];
-}
-
-//===----------------------------------------------------------------------===//
-// GetGlobalMemrefOp
-//===----------------------------------------------------------------------===//
-
-def GetGlobalMemrefOp : Std_Op<"get_global_memref",
- [NoSideEffect, DeclareOpInterfaceMethods<SymbolUserOpInterface>]> {
- let summary = "get the memref pointing to a global variable";
- let description = [{
- The `get_global_memref` operation retrieves the memref pointing to a
- named global variable. If the global variable is marked constant, writing
- to the result memref (such as through a `std.store` operation) is
- undefined.
-
- Example:
-
- ```mlir
- %x = get_global_memref @foo : memref<2xf32>
- ```
- }];
-
- let arguments = (ins FlatSymbolRefAttr:$name);
- let results = (outs AnyStaticShapeMemRef:$result);
- let assemblyFormat = "$name `:` type($result) attr-dict";
-
- // `GetGlobalMemrefOp` is fully verified by its traits.
- let verifier = ?;
-}
-
//===----------------------------------------------------------------------===//
// IndexCastOp
//===----------------------------------------------------------------------===//
@@ -1620,470 +1270,126 @@ def IndexCastOp : ArithmeticCastOp<"index_cast">, Arguments<(ins AnyType:$in)> {
}
//===----------------------------------------------------------------------===//
-// LoadOp
+// MulFOp
//===----------------------------------------------------------------------===//
-def LoadOp : Std_Op<"load",
- [TypesMatchWith<"result type matches element type of 'memref'",
- "memref", "result",
- "$_self.cast<MemRefType>().getElementType()">,
- MemRefsNormalizable]> {
- let summary = "load operation";
+def MulFOp : FloatBinaryOp<"mulf"> {
+ let summary = "floating point multiplication operation";
let description = [{
- The `load` op reads an element from a memref specified by an index list. The
- output of load is a new value with the same type as the elements of the
- memref. The arity of indices is the rank of the memref (i.e., if the memref
- loaded from is of rank 3, then 3 indices are required for the load following
- the memref identifier).
-
- In an `affine.if` or `affine.for` body, the indices of a load are restricted
- to SSA values bound to surrounding loop induction variables,
- [symbols](Affine.md#dimensions-and-symbols), results of a
- [`constant` operation](#stdconstant-constantop), or the result of an
- `affine.apply` operation that can in turn take as arguments all of the
- aforementioned SSA values or the recursively result of such an
- `affine.apply` operation.
+ Syntax:
+
+ ```
+ operation ::= ssa-id `=` `std.mulf` ssa-use `,` ssa-use `:` type
+ ```
+
+ The `mulf` operation takes two operands and returns one result, each of
+ these is required to be the same type. This type may be a floating point
+ scalar type, a vector whose element type is a floating point type, or a
+ floating point tensor.
Example:
```mlir
- %1 = affine.apply affine_map<(d0, d1) -> (3*d0)> (%i, %j)
- %2 = affine.apply affine_map<(d0, d1) -> (d1+1)> (%i, %j)
- %12 = load %A[%1, %2] : memref<8x?xi32, #layout, memspace0>
+ // Scalar multiplication.
+ %a = mulf %b, %c : f64
+
+ // SIMD pointwise vector multiplication, e.g. for Intel SSE.
+ %f = mulf %g, %h : vector<4xf32>
- // Example of an indirect load (treated as non-affine)
- %3 = affine.apply affine_map<(d0) -> (2*d0 + 1)>(%12)
- %13 = load %A[%3, %2] : memref<4x?xi32, #layout, memspace0>
+ // Tensor pointwise multiplication.
+ %x = mulf %y, %z : tensor<4x?xbf16>
```
- **Context:** The `load` and `store` operations are specifically crafted to
- fully resolve a reference to an element of a memref, and (in affine
- `affine.if` and `affine.for` operations) the compiler can follow use-def
- chains (e.g. through [`affine.apply`](Affine.md#affineapply-affineapplyop)
- operations) to precisely analyze references at compile-time using polyhedral
- techniques. This is possible because of the
- [restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
- in these contexts.
+ TODO: In the distant future, this will accept optional attributes for fast
+ math, contraction, rounding mode, and other controls.
}];
+ let hasFolder = 1;
+}
- let arguments = (ins Arg<AnyMemRef, "the reference to load from",
- [MemRead]>:$memref,
- Variadic<Index>:$indices);
- let results = (outs AnyType:$result);
-
- let builders = [
- OpBuilder<(ins "Value":$memref, CArg<"ValueRange", "{}">:$indices), [{
- auto memrefType = memref.getType().cast<MemRefType>();
- $_state.addOperands(memref);
- $_state.addOperands(indices);
- $_state.types.push_back(memrefType.getElementType());
- }]>];
-
- let extraClassDeclaration = [{
- Value getMemRef() { return getOperand(0); }
- void setMemRef(Value value) { setOperand(0, value); }
- MemRefType getMemRefType() {
- return getMemRef().getType().cast<MemRefType>();
- }
-
- operand_range getIndices() { return {operand_begin() + 1, operand_end()}; }
- }];
+//===----------------------------------------------------------------------===//
+// MulIOp
+//===----------------------------------------------------------------------===//
- let hasCanonicalizer = 1;
+def MulIOp : IntBinaryOp<"muli", [Commutative]> {
+ let summary = "integer multiplication operation";
let hasFolder = 1;
-
- let assemblyFormat = "$memref `[` $indices `]` attr-dict `:` type($memref)";
}
//===----------------------------------------------------------------------===//
-// MemRefCastOp
+// NegFOp
//===----------------------------------------------------------------------===//
-def MemRefCastOp : CastOp<"memref_cast", [
- DeclareOpInterfaceMethods<ViewLikeOpInterface>
- ]> {
- let summary = "memref cast operation";
+def NegFOp : FloatUnaryOp<"negf"> {
+ let summary = "floating point negation";
let description = [{
Syntax:
```
- operation ::= ssa-id `=` `std.memref_cast` ssa-use `:` type `to` type
+ operation ::= ssa-id `=` `negf` ssa-use `:` type
```
- The `memref_cast` operation converts a memref from one type to an equivalent
- type with a compatible shape. The source and destination types are
- compatible if:
-
- a. Both are ranked memref types with the same element type, address space,
- and rank and:
- 1. Both have the same layout or both have compatible strided layouts.
- 2. The individual sizes (resp. offset and strides in the case of strided
- memrefs) may convert constant dimensions to dynamic dimensions and
- vice-versa.
-
- If the cast converts any dimensions from an unknown to a known size, then it
- acts as an assertion that fails at runtime if the dynamic dimensions
- disagree with resultant destination size.
+ The `negf` operation computes the negation of a given value. It takes one
+ operand and returns one result of the same type. This type may be a float
+ scalar type, a vector whose element type is float, or a tensor of floats.
+ It has no standard attributes.
Example:
```mlir
- // Assert that the input dynamic shape matches the destination static shape.
- %2 = memref_cast %1 : memref<?x?xf32> to memref<4x4xf32>
- // Erase static shape information, replacing it with dynamic information.
- %3 = memref_cast %1 : memref<4xf32> to memref<?xf32>
+ // Scalar negation value.
+ %a = negf %b : f64
+
+ // SIMD vector element-wise negation value.
+ %f = negf %g : vector<4xf32>
+
+ // Tensor element-wise negation value.
+ %x = negf %y : tensor<4x?xf8>
+ ```
+ }];
+}
+
+//===----------------------------------------------------------------------===//
+// OrOp
+//===----------------------------------------------------------------------===//
- // The same holds true for offsets and strides.
+def OrOp : IntBinaryOp<"or", [Commutative]> {
+ let summary = "integer binary or";
+ let description = [{
+ Syntax:
- // Assert that the input dynamic shape matches the destination static stride.
- %4 = memref_cast %1 : memref<12x4xf32, offset:?, strides: [?, ?]> to
- memref<12x4xf32, offset:5, strides: [4, 1]>
- // Erase static offset and stride information, replacing it with
- // dynamic information.
- %5 = memref_cast %1 : memref<12x4xf32, offset:5, strides: [4, 1]> to
- memref<12x4xf32, offset:?, strides: [?, ?]>
+ ```
+ operation ::= ssa-id `=` `or` ssa-use `,` ssa-use `:` type
```
- b. Either or both memref types are unranked with the same element type, and
- address space.
+ The `or` operation takes two operands and returns one result, each of these
+ is required to be the same type. This type may be an integer scalar type, a
+ vector whose element type is integer, or a tensor of integers. It has no
+ standard attributes.
Example:
```mlir
- Cast to concrete shape.
- %4 = memref_cast %1 : memref<*xf32> to memref<4x?xf32>
+ // Scalar integer bitwise or.
+ %a = or %b, %c : i64
+
+ // SIMD vector element-wise bitwise integer or.
+ %f = or %g, %h : vector<4xi32>
- Erase rank information.
- %5 = memref_cast %1 : memref<4x?xf32> to memref<*xf32>
+ // Tensor element-wise bitwise integer or.
+ %x = or %y, %z : tensor<4x?xi8>
```
}];
-
- let arguments = (ins AnyRankedOrUnrankedMemRef:$source);
- let results = (outs AnyRankedOrUnrankedMemRef);
-
let hasFolder = 1;
}
-
//===----------------------------------------------------------------------===//
-// MemRefReinterpretCastOp
+// RankOp
//===----------------------------------------------------------------------===//
-def MemRefReinterpretCastOp:
- BaseOpWithOffsetSizesAndStrides<"memref_reinterpret_cast", [
- NoSideEffect, ViewLikeOpInterface, OffsetSizeAndStrideOpInterface
- ]> {
- let summary = "memref reinterpret cast operation";
+def RankOp : Std_Op<"rank", [NoSideEffect]> {
+ let summary = "rank operation";
let description = [{
- Modify offset, sizes and strides of an unranked/ranked memref.
-
- Example:
- ```mlir
- memref_reinterpret_cast %ranked to
- offset: [0],
- sizes: [%size0, 10],
- strides: [1, %stride1]
- : memref<?x?xf32> to memref<?x10xf32, offset: 0, strides: [1, ?]>
-
- memref_reinterpret_cast %unranked to
- offset: [%offset],
- sizes: [%size0, %size1],
- strides: [%stride0, %stride1]
- : memref<*xf32> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- ```
- }];
-
- let arguments = (ins
- Arg<AnyRankedOrUnrankedMemRef, "", []>:$source,
- Variadic<Index>:$offsets,
- Variadic<Index>:$sizes,
- Variadic<Index>:$strides,
- I64ArrayAttr:$static_offsets,
- I64ArrayAttr:$static_sizes,
- I64ArrayAttr:$static_strides
- );
- let results = (outs AnyMemRef:$result);
-
- let assemblyFormat = [{
- $source `to` `offset` `` `:`
- custom<OperandsOrIntegersOffsetsOrStridesList>($offsets, $static_offsets)
- `` `,` `sizes` `` `:`
- custom<OperandsOrIntegersSizesList>($sizes, $static_sizes) `` `,` `strides`
- `` `:`
- custom<OperandsOrIntegersOffsetsOrStridesList>($strides, $static_strides)
- attr-dict `:` type($source) `to` type($result)
- }];
-
- let parser=?;
- let printer=?;
-
- let builders = [
- // Build a ReinterpretCastOp with mixed static and dynamic entries.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "OpFoldResult":$offset, "ArrayRef<OpFoldResult>":$sizes,
- "ArrayRef<OpFoldResult>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a ReinterpretCastOp with static entries.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "int64_t":$offset, "ArrayRef<int64_t>":$sizes,
- "ArrayRef<int64_t>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a ReinterpretCastOp with dynamic entries.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "Value":$offset, "ValueRange":$sizes,
- "ValueRange":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>
- ];
-
- let extraClassDeclaration = extraBaseClassDeclaration # [{
- // The result of the op is always a ranked memref.
- MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
- Value getViewSource() { return source(); }
-
- /// Return the rank of the source ShapedType.
- unsigned getResultRank() {
- return getResult().getType().cast<ShapedType>().getRank();
- }
-
- /// Return the expected rank of each of the`static_offsets`, `static_sizes`
- /// and `static_strides` attributes.
- std::array<unsigned, 3> getArrayAttrMaxRanks() {
- unsigned resultRank = getResult().getType().cast<ShapedType>().getRank();
- return {1, resultRank, resultRank};
- }
-
- /// Return the number of leading operands before the `offsets`, `sizes` and
- /// and `strides` operands.
- static unsigned getOffsetSizeAndStrideStartOperandIndex() { return 1; }
- }];
-}
-
-//===----------------------------------------------------------------------===//
-// MemRefReshapeOp
-//===----------------------------------------------------------------------===//
-
-def MemRefReshapeOp: Std_Op<"memref_reshape", [
- ViewLikeOpInterface, NoSideEffect]> {
- let summary = "memref reshape operation";
- let description = [{
- The `memref_reshape` operation converts a memref from one type to an
- equivalent type with a provided shape. The data is never copied or
- modified. The source and destination types are compatible if both have the
- same element type, same number of elements, address space and identity
- layout map. The following combinations are possible:
-
- a. Source type is ranked or unranked. Shape argument has static size.
- Result type is ranked.
-
- ```mlir
- // Reshape statically-shaped memref.
- %dst = memref_reshape %src(%shape)
- : (memref<4x1xf32>, memref<1xi32>) to memref<4xf32>
- %dst0 = memref_reshape %src(%shape0)
- : (memref<4x1xf32>, memref<2xi32>) to memref<2x2xf32>
- // Flatten unranked memref.
- %dst = memref_reshape %src(%shape)
- : (memref<*xf32>, memref<1xi32>) to memref<?xf32>
- ```
-
- a. Source type is ranked or unranked. Shape argument has dynamic size.
- Result type is unranked.
-
- ```mlir
- // Reshape dynamically-shaped 1D memref.
- %dst = memref_reshape %src(%shape)
- : (memref<?xf32>, memref<?xi32>) to memref<*xf32>
- // Reshape unranked memref.
- %dst = memref_reshape %src(%shape)
- : (memref<*xf32>, memref<?xi32>) to memref<*xf32>
- ```
- }];
-
- let arguments = (ins
- AnyRankedOrUnrankedMemRef:$source,
- MemRefRankOf<[AnySignlessInteger, Index], [1]>:$shape
- );
- let results = (outs AnyRankedOrUnrankedMemRef:$result);
-
- let builders = [OpBuilder<
- (ins "MemRefType":$resultType, "Value":$operand, "Value":$shape), [{
- $_state.addOperands(operand);
- $_state.addOperands(shape);
- $_state.addTypes(resultType);
- }]>];
-
- let extraClassDeclaration = [{
- MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
- Value getViewSource() { return source(); }
- }];
-
- let assemblyFormat = [{
- $source `(` $shape `)` attr-dict `:` functional-type(operands, results)
- }];
-}
-
-//===----------------------------------------------------------------------===//
-// MulFOp
-//===----------------------------------------------------------------------===//
-
-def MulFOp : FloatBinaryOp<"mulf"> {
- let summary = "floating point multiplication operation";
- let description = [{
- Syntax:
-
- ```
- operation ::= ssa-id `=` `std.mulf` ssa-use `,` ssa-use `:` type
- ```
-
- The `mulf` operation takes two operands and returns one result, each of
- these is required to be the same type. This type may be a floating point
- scalar type, a vector whose element type is a floating point type, or a
- floating point tensor.
-
- Example:
-
- ```mlir
- // Scalar multiplication.
- %a = mulf %b, %c : f64
-
- // SIMD pointwise vector multiplication, e.g. for Intel SSE.
- %f = mulf %g, %h : vector<4xf32>
-
- // Tensor pointwise multiplication.
- %x = mulf %y, %z : tensor<4x?xbf16>
- ```
-
- TODO: In the distant future, this will accept optional attributes for fast
- math, contraction, rounding mode, and other controls.
- }];
- let hasFolder = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// MulIOp
-//===----------------------------------------------------------------------===//
-
-def MulIOp : IntBinaryOp<"muli", [Commutative]> {
- let summary = "integer multiplication operation";
- let hasFolder = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// NegFOp
-//===----------------------------------------------------------------------===//
-
-def NegFOp : FloatUnaryOp<"negf"> {
- let summary = "floating point negation";
- let description = [{
- Syntax:
-
- ```
- operation ::= ssa-id `=` `negf` ssa-use `:` type
- ```
-
- The `negf` operation computes the negation of a given value. It takes one
- operand and returns one result of the same type. This type may be a float
- scalar type, a vector whose element type is float, or a tensor of floats.
- It has no standard attributes.
-
- Example:
-
- ```mlir
- // Scalar negation value.
- %a = negf %b : f64
-
- // SIMD vector element-wise negation value.
- %f = negf %g : vector<4xf32>
-
- // Tensor element-wise negation value.
- %x = negf %y : tensor<4x?xf8>
- ```
- }];
-}
-
-//===----------------------------------------------------------------------===//
-// OrOp
-//===----------------------------------------------------------------------===//
-
-def OrOp : IntBinaryOp<"or", [Commutative]> {
- let summary = "integer binary or";
- let description = [{
- Syntax:
-
- ```
- operation ::= ssa-id `=` `or` ssa-use `,` ssa-use `:` type
- ```
-
- The `or` operation takes two operands and returns one result, each of these
- is required to be the same type. This type may be an integer scalar type, a
- vector whose element type is integer, or a tensor of integers. It has no
- standard attributes.
-
- Example:
-
- ```mlir
- // Scalar integer bitwise or.
- %a = or %b, %c : i64
-
- // SIMD vector element-wise bitwise integer or.
- %f = or %g, %h : vector<4xi32>
-
- // Tensor element-wise bitwise integer or.
- %x = or %y, %z : tensor<4x?xi8>
- ```
- }];
- let hasFolder = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// PrefetchOp
-//===----------------------------------------------------------------------===//
-
-def PrefetchOp : Std_Op<"prefetch"> {
- let summary = "prefetch operation";
- let description = [{
- The "prefetch" op prefetches data from a memref location described with
- subscript indices similar to std.load, and with three attributes: a
- read/write specifier, a locality hint, and a cache type specifier as shown
- below:
-
- ```mlir
- prefetch %0[%i, %j], read, locality<3>, data : memref<400x400xi32>
- ```
-
- The read/write specifier is either 'read' or 'write', the locality hint
- ranges from locality<0> (no locality) to locality<3> (extremely local keep
- in cache). The cache type specifier is either 'data' or 'instr'
- and specifies whether the prefetch is performed on data cache or on
- instruction cache.
- }];
-
- let arguments = (ins AnyMemRef:$memref, Variadic<Index>:$indices,
- BoolAttr:$isWrite,
- Confined<I32Attr, [IntMinValue<0>,
- IntMaxValue<3>]>:$localityHint,
- BoolAttr:$isDataCache);
-
- let extraClassDeclaration = [{
- MemRefType getMemRefType() {
- return memref().getType().cast<MemRefType>();
- }
- static StringRef getLocalityHintAttrName() { return "localityHint"; }
- static StringRef getIsWriteAttrName() { return "isWrite"; }
- static StringRef getIsDataCacheAttrName() { return "isDataCache"; }
- }];
-
- let hasFolder = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// RankOp
-//===----------------------------------------------------------------------===//
-
-def RankOp : Std_Op<"rank", [NoSideEffect]> {
- let summary = "rank operation";
- let description = [{
- The `rank` operation takes a memref/tensor operand and returns its rank.
+ The `rank` operation takes a memref/tensor operand and returns its rank.
Example:
@@ -2094,7 +1400,7 @@ def RankOp : Std_Op<"rank", [NoSideEffect]> {
}];
let arguments = (ins AnyTypeOf<[AnyRankedOrUnrankedMemRef, AnyTensor],
- "any tensor or memref type">:$memrefOrTensor);
+ "any memref or tensor type">:$memrefOrTensor);
let results = (outs Index);
let verifier = ?;
@@ -2486,77 +1792,6 @@ def SplatOp : Std_Op<"splat", [NoSideEffect,
let assemblyFormat = "$input attr-dict `:` type($aggregate)";
}
-//===----------------------------------------------------------------------===//
-// StoreOp
-//===----------------------------------------------------------------------===//
-
-def StoreOp : Std_Op<"store",
- [TypesMatchWith<"type of 'value' matches element type of 'memref'",
- "memref", "value",
- "$_self.cast<MemRefType>().getElementType()">,
- MemRefsNormalizable]> {
- let summary = "store operation";
- let description = [{
- Store a value to a memref location given by indices. The value stored should
- have the same type as the elemental type of the memref. The number of
- arguments provided within brackets need to match the rank of the memref.
-
- In an affine context, the indices of a store are restricted to SSA values
- bound to surrounding loop induction variables,
- [symbols](Affine.md#restrictions-on-dimensions-and-symbols), results of a
- [`constant` operation](#stdconstant-constantop), or the result of an
- [`affine.apply`](Affine.md#affineapply-affineapplyop) operation that can in turn
- take as arguments all of the aforementioned SSA values or the recursively
- result of such an `affine.apply` operation.
-
- Example:
-
- ```mlir
- store %100, %A[%1, 1023] : memref<4x?xf32, #layout, memspace0>
- ```
-
- **Context:** The `load` and `store` operations are specifically crafted to
- fully resolve a reference to an element of a memref, and (in polyhedral
- `affine.if` and `affine.for` operations) the compiler can follow use-def
- chains (e.g. through [`affine.apply`](Affine.md#affineapply-affineapplyop)
- operations) to precisely analyze references at compile-time using polyhedral
- techniques. This is possible because of the
- [restrictions on dimensions and symbols](Affine.md#restrictions-on-dimensions-and-symbols)
- in these contexts.
- }];
-
- let arguments = (ins AnyType:$value,
- Arg<AnyMemRef, "the reference to store to",
- [MemWrite]>:$memref,
- Variadic<Index>:$indices);
-
- let builders = [
- OpBuilder<(ins "Value":$valueToStore, "Value":$memref), [{
- $_state.addOperands(valueToStore);
- $_state.addOperands(memref);
- }]>];
-
- let extraClassDeclaration = [{
- Value getValueToStore() { return getOperand(0); }
-
- Value getMemRef() { return getOperand(1); }
- void setMemRef(Value value) { setOperand(1, value); }
- MemRefType getMemRefType() {
- return getMemRef().getType().cast<MemRefType>();
- }
-
- operand_range getIndices() {
- return {operand_begin() + 2, operand_end()};
- }
- }];
-
- let hasFolder = 1;
-
- let assemblyFormat = [{
- $value `,` $memref `[` $indices `]` attr-dict `:` type($memref)
- }];
-}
-
//===----------------------------------------------------------------------===//
// SubFOp
//===----------------------------------------------------------------------===//
@@ -2575,240 +1810,6 @@ def SubIOp : IntBinaryOp<"subi"> {
let hasFolder = 1;
}
-//===----------------------------------------------------------------------===//
-// SubViewOp
-//===----------------------------------------------------------------------===//
-
-def SubViewOp : BaseOpWithOffsetSizesAndStrides<
- "subview", [DeclareOpInterfaceMethods<ViewLikeOpInterface>,
- NoSideEffect, OffsetSizeAndStrideOpInterface] > {
- let summary = "memref subview operation";
- let description = [{
- The "subview" operation converts a memref type to another memref type
- which represents a reduced-size view of the original memref as specified by
- the operation's offsets, sizes and strides arguments.
-
- The SubView operation supports the following arguments:
-
- * source: the "base" memref on which to create a "view" memref.
- * offsets: memref-rank number of offsets into the "base" memref at which to
- create the "view" memref.
- * sizes: memref-rank number of sizes which specify the sizes of the result
- "view" memref type.
- * strides: memref-rank number of strides that compose multiplicatively with
- the base memref strides in each dimension.
-
- The representation based on offsets, sizes and strides support a
- partially-static specification via attributes specified through the
- `static_offsets`, `static_sizes` and `static_strides` arguments. A special
- sentinel value ShapedType::kDynamicSize and
- ShapedType::kDynamicStrideOrOffset encodes that the corresponding entry has
- a dynamic value.
-
- A subview operation may additionally reduce the rank of the resulting view
- by removing dimensions that are statically known to be of size 1.
-
- Example 1:
-
- ```mlir
- %0 = alloc() : memref<64x4xf32, (d0, d1) -> (d0 * 4 + d1)>
-
- // Create a sub-view of "base" memref '%0' with offset arguments '%c0',
- // dynamic sizes for each dimension, and stride arguments '%c1'.
- %1 = subview %0[%c0, %c0][%size0, %size1][%c1, %c1]
- : memref<64x4xf32, (d0, d1) -> (d0 * 4 + d1) > to
- memref<?x?xf32, (d0, d1)[s0, s1] -> (d0 * s1 + d1 + s0)>
- ```
-
- Example 2:
-
- ```mlir
- %0 = alloc() : memref<8x16x4xf32, (d0, d1, d1) -> (d0 * 64 + d1 * 4 + d2)>
-
- // Create a sub-view of "base" memref '%0' with dynamic offsets, sizes,
- // and strides.
- // Note that dynamic offsets are represented by the linearized dynamic
- // offset symbol 's0' in the subview memref layout map, and that the
- // dynamic strides operands, after being applied to the base memref
- // strides in each dimension, are represented in the view memref layout
- // map as symbols 's1', 's2' and 's3'.
- %1 = subview %0[%i, %j, %k][%size0, %size1, %size2][%x, %y, %z]
- : memref<8x16x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)> to
- memref<?x?x?xf32,
- (d0, d1, d2)[s0, s1, s2, s3] -> (d0 * s1 + d1 * s2 + d2 * s3 + s0)>
- ```
-
- Example 3:
-
- ```mlir
- %0 = alloc() : memref<8x16x4xf32, (d0, d1, d1) -> (d0 * 64 + d1 * 4 + d2)>
-
- // Subview with constant offsets, sizes and strides.
- %1 = subview %0[0, 2, 0][4, 4, 4][64, 4, 1]
- : memref<8x16x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)> to
- memref<4x4x4xf32, (d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2 + 8)>
- ```
-
- Example 4:
-
- ```mlir
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
-
- // Subview with constant size, but dynamic offsets and
- // strides. The resulting memref has a static shape, but if the
- // base memref has an affine map to describe the layout, the result
- // memref also uses an affine map to describe the layout. The
- // strides of the result memref is computed as follows:
- //
- // Let #map1 represents the layout of the base memref, and #map2
- // represents the layout of the result memref. A #mapsubview can be
- // constructed to map an index from the result memref to the base
- // memref (note that the description below uses more convenient
- // naming for symbols, while in affine maps, symbols are
- // represented as unsigned numbers that identify that symbol in the
- // given affine map.
- //
- // #mapsubview = (d0, d1)[o0, o1, t0, t1] -> (d0 * t0 + o0, d1 * t1 + o1)
- //
- // where, o0, o1, ... are offsets, and t0, t1, ... are strides. Then,
- //
- // #map2 = #map1.compose(#mapsubview)
- //
- // If the layout map is represented as
- //
- // #map1 = (d0, d1)[s0, s1, s2] -> (d0 * s1 + d1 * s2 + s0)
- //
- // then,
- //
- // #map2 = (d0, d1)[s0, s1, s2, o0, o1, t0, t1] ->
- // (d0 * s1 * t0 + d1 * s2 * t1 + o0 * s1 + o1 * s2 + s0)
- //
- // Representing this canonically
- //
- // #map2 = (d0, d1)[r0, r1, r2] -> (d0 * r1 + d1 * r2 + r0)
- //
- // where, r0 = o0 * s1 + o1 * s2 + s0, r1 = s1 * t0, r2 = s2 * t1.
- %1 = subview %0[%i, %j][4, 4][%x, %y] :
- : memref<?x?xf32, (d0, d1)[s0, s1, s2] -> (d0 * s1 + d1 * s2 + s0)> to
- memref<4x4xf32, (d0, d1)[r0, r1, r2] -> (d0 * r1 + d1 * r2 + r0)>
-
- // Note that the subview op does not guarantee that the result
- // memref is "inbounds" w.r.t to base memref. It is upto the client
- // to ensure that the subview is accessed in a manner that is
- // in-bounds.
- ```
-
- Example 5:
-
- ```mlir
- // Rank-reducing subview.
- %1 = subview %0[0, 0, 0][1, 16, 4][1, 1, 1] :
- memref<8x16x4xf32> to memref<16x4xf32>
- %3 = subview %2[3, 4, 2][1, 6, 3][1, 1, 1] :
- memref<8x16x4xf32> to memref<6x3xf32, offset: 210, strides: [4, 1]>
- ```
- }
- }];
-
- let arguments = (ins
- AnyMemRef:$source,
- Variadic<Index>:$offsets,
- Variadic<Index>:$sizes,
- Variadic<Index>:$strides,
- I64ArrayAttr:$static_offsets,
- I64ArrayAttr:$static_sizes,
- I64ArrayAttr:$static_strides
- );
- let results = (outs AnyMemRef:$result);
-
- let assemblyFormat = [{
- $source ``
- custom<OperandsOrIntegersOffsetsOrStridesList>($offsets, $static_offsets)
- custom<OperandsOrIntegersSizesList>($sizes, $static_sizes)
- custom<OperandsOrIntegersOffsetsOrStridesList>($strides, $static_strides)
- attr-dict `:` type($source) `to` type($result)
- }];
-
- let builders = [
- // Build a SubViewOp with mixed static and dynamic entries and custom
- // result type. If the type passed is nullptr, it is inferred.
- OpBuilder<(ins "Value":$source, "ArrayRef<OpFoldResult>":$offsets,
- "ArrayRef<OpFoldResult>":$sizes, "ArrayRef<OpFoldResult>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a SubViewOp with mixed static and dynamic entries and inferred
- // result type.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "ArrayRef<OpFoldResult>":$offsets, "ArrayRef<OpFoldResult>":$sizes,
- "ArrayRef<OpFoldResult>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a SubViewOp with static entries and custom result type. If the
- // type passed is nullptr, it is inferred.
- OpBuilder<(ins "Value":$source, "ArrayRef<int64_t>":$offsets,
- "ArrayRef<int64_t>":$sizes, "ArrayRef<int64_t>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a SubViewOp with static entries and inferred result type.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "ArrayRef<int64_t>":$offsets, "ArrayRef<int64_t>":$sizes,
- "ArrayRef<int64_t>":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a SubViewOp with dynamic entries and custom result type. If the
- // type passed is nullptr, it is inferred.
- OpBuilder<(ins "Value":$source, "ValueRange":$offsets,
- "ValueRange":$sizes, "ValueRange":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>,
- // Build a SubViewOp with dynamic entries and inferred result type.
- OpBuilder<(ins "MemRefType":$resultType, "Value":$source,
- "ValueRange":$offsets, "ValueRange":$sizes, "ValueRange":$strides,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>
- ];
-
- let extraClassDeclaration = extraBaseClassDeclaration # [{
- /// Returns the type of the base memref operand.
- MemRefType getSourceType() {
- return source().getType().cast<MemRefType>();
- }
-
- /// The result of a subview is always a memref.
- MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
-
- /// A subview result type can be fully inferred from the source type and the
- /// static representation of offsets, sizes and strides. Special sentinels
- /// encode the dynamic case.
- static Type inferResultType(MemRefType sourceMemRefType,
- ArrayRef<int64_t> staticOffsets,
- ArrayRef<int64_t> staticSizes,
- ArrayRef<int64_t> staticStrides);
- static Type inferResultType(MemRefType sourceMemRefType,
- ArrayRef<OpFoldResult> staticOffsets,
- ArrayRef<OpFoldResult> staticSizes,
- ArrayRef<OpFoldResult> staticStrides);
- static Type inferRankReducedResultType(unsigned resultRank,
- MemRefType sourceMemRefType,
- ArrayRef<int64_t> staticOffsets,
- ArrayRef<int64_t> staticSizes,
- ArrayRef<int64_t> staticStrides);
- static Type inferRankReducedResultType(unsigned resultRank,
- MemRefType sourceMemRefType,
- ArrayRef<OpFoldResult> staticOffsets,
- ArrayRef<OpFoldResult> staticSizes,
- ArrayRef<OpFoldResult> staticStrides);
-
- /// Return the expected rank of each of the`static_offsets`, `static_sizes`
- /// and `static_strides` attributes.
- std::array<unsigned, 3> getArrayAttrMaxRanks() {
- unsigned rank = getSourceType().getRank();
- return {rank, rank, rank};
- }
-
- /// Return the number of leading operands before the `offsets`, `sizes` and
- /// and `strides` operands.
- static unsigned getOffsetSizeAndStrideStartOperandIndex() { return 1; }
- }];
-
- let hasCanonicalizer = 1;
- let hasFolder = 1;
-}
-
//===----------------------------------------------------------------------===//
// SubTensorOp
//===----------------------------------------------------------------------===//
@@ -3046,170 +2047,6 @@ def SubTensorInsertOp : BaseOpWithOffsetSizesAndStrides<
let hasFolder = 1;
}
-//===----------------------------------------------------------------------===//
-// TensorLoadOp
-//===----------------------------------------------------------------------===//
-
-def TensorLoadOp : Std_Op<"tensor_load",
- [SameOperandsAndResultShape, SameOperandsAndResultElementType,
- TypesMatchWith<"result type matches tensor equivalent of 'memref'",
- "memref", "result",
- "getTensorTypeFromMemRefType($_self)">]> {
- let summary = "tensor load operation";
- let description = [{
- Create a tensor from a memref, making an independent copy of the element
- data. The result value is a tensor whose shape and element type match the
- memref operand.
-
- The opposite of this op is tensor_to_memref. Together, these two ops are
- useful for source/target materializations when doing type conversions
- involving tensors and memrefs.
-
- Example:
-
- ```mlir
- // Produces a value of tensor<4x?xf32> type.
- %12 = tensor_load %10 : memref<4x?xf32, #layout, memspace0>
- ```
- }];
-
- let arguments = (ins Arg<AnyRankedOrUnrankedMemRef,
- "the reference to load from", [MemRead]>:$memref);
- let results = (outs AnyTensor:$result);
- // TensorLoadOp is fully verified by traits.
- let verifier = ?;
-
- let builders = [
- OpBuilder<(ins "Value":$memref), [{
- $_state.addOperands(memref);
- $_state.addTypes(getTensorTypeFromMemRefType(memref.getType()));
- }]>];
-
- let extraClassDeclaration = [{
- /// The result of a tensor_load is always a tensor.
- TensorType getType() {
- Type resultType = getResult().getType();
- if (resultType.isa<TensorType>())
- return resultType.cast<TensorType>();
- return {};
- }
- }];
-
- let assemblyFormat = "$memref attr-dict `:` type($memref)";
-
- let hasFolder = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// TensorStoreOp
-//===----------------------------------------------------------------------===//
-
-def TensorStoreOp : Std_Op<"tensor_store",
- [SameOperandsShape, SameOperandsElementType,
- TypesMatchWith<"type of 'value' matches tensor equivalent of 'memref'",
- "memref", "tensor",
- "getTensorTypeFromMemRefType($_self)">]> {
- let summary = "tensor store operation";
- let description = [{
- Stores the contents of a tensor into a memref. The first operand is a value
- of tensor type, the second operand is a value of memref type. The shapes and
- element types of these must match, and are specified by the memref type.
-
- Example:
-
- ```mlir
- %9 = dim %8, 1 : tensor<4x?xf32>
- %10 = alloc(%9) : memref<4x?xf32, #layout, memspace0>
- tensor_store %8, %10 : memref<4x?xf32, #layout, memspace0>
- ```
- }];
-
- let arguments = (ins AnyTensor:$tensor, Arg<AnyRankedOrUnrankedMemRef,
- "the reference to store to", [MemWrite]>:$memref);
- // TensorStoreOp is fully verified by traits.
- let verifier = ?;
-
- let assemblyFormat = "$tensor `,` $memref attr-dict `:` type($memref)";
-}
-
-//===----------------------------------------------------------------------===//
-// TensorToMemrefOp
-//===----------------------------------------------------------------------===//
-
-def TensorToMemrefOp : Std_Op<"tensor_to_memref",
- [SameOperandsAndResultShape, SameOperandsAndResultElementType,
- TypesMatchWith<"type of 'tensor' is the tensor equivalent of 'memref'",
- "memref", "tensor",
- "getTensorTypeFromMemRefType($_self)">]> {
- let summary = "tensor to memref operation";
- let description = [{
- Create a memref from a tensor. This is a transient op created as a
- materialization during type conversions between tensors and memrefs.
-
- The opposite of this op is tensor_load. Together, these two ops are useful
- for source/target materializations when doing type conversions involving
- tensors and memrefs.
-
- This op is defined by the fold
- `tensor_to_memref(tensor_load(%memref)) -> %memref`, which is the property
- that makes it a valid materialization in the type conversion framework.
- This implies that one cannot assume that this op allocates a new memref for
- its result.
-
- Note: This op takes the memref type in its pretty form because the tensor
- type can always be inferred from the memref type, but the reverse is not
- true. For example, the memref might have a layout map or memory space which
- cannot be inferred from the tensor type.
-
- ```mlir
- // Result type is tensor<4x?xf32>
- %12 = tensor_to_memref %10 : memref<4x?xf32, #map0, 42>
- ```
- }];
-
- let arguments = (ins AnyTensor:$tensor);
- let results = (outs AnyRankedOrUnrankedMemRef:$memref);
- // This op is fully verified by traits.
- let verifier = ?;
-
- let assemblyFormat = "$tensor attr-dict `:` type($memref)";
-
- let hasFolder = 1;
- let hasCanonicalizer = 1;
-}
-
-//===----------------------------------------------------------------------===//
-// TransposeOp
-//===----------------------------------------------------------------------===//
-
-def TransposeOp : Std_Op<"transpose", [NoSideEffect]>,
- Arguments<(ins AnyStridedMemRef:$in, AffineMapAttr:$permutation)>,
- Results<(outs AnyStridedMemRef)> {
- let summary = "`transpose` produces a new strided memref (metadata-only)";
- let description = [{
- The `transpose` op produces a strided memref whose sizes and strides
- are a permutation of the original `in` memref. This is purely a metadata
- transformation.
-
- Example:
-
- ```mlir
- %1 = transpose %0 (i, j) -> (j, i) : memref<?x?xf32> to memref<?x?xf32, affine_map<(d0, d1)[s0] -> (d1 * s0 + d0)>>
- ```
- }];
-
- let builders = [
- OpBuilder<(ins "Value":$in, "AffineMapAttr":$permutation,
- CArg<"ArrayRef<NamedAttribute>", "{}">:$attrs)>];
-
- let extraClassDeclaration = [{
- static StringRef getPermutationAttrName() { return "permutation"; }
- ShapedType getShapedType() { return in().getType().cast<ShapedType>(); }
- }];
-
- let hasFolder = 1;
-}
-
//===----------------------------------------------------------------------===//
// TruncateIOp
//===----------------------------------------------------------------------===//
@@ -3357,74 +2194,6 @@ def UnsignedShiftRightOp : IntBinaryOp<"shift_right_unsigned"> {
}];
}
-//===----------------------------------------------------------------------===//
-// ViewOp
-//===----------------------------------------------------------------------===//
-
-def ViewOp : Std_Op<"view", [
- DeclareOpInterfaceMethods<ViewLikeOpInterface>, NoSideEffect]> {
- let summary = "memref view operation";
- let description = [{
- The "view" operation extracts an N-D contiguous memref with empty layout map
- with arbitrary element type from a 1-D contiguous memref with empty layout
- map of i8 element type. The ViewOp supports the following arguments:
-
- * A single dynamic byte-shift operand must be specified which represents a
- a shift of the base 1-D memref pointer from which to create the resulting
- contiguous memref view with identity layout.
- * A dynamic size operand that must be specified for each dynamic dimension
- in the resulting view memref type.
-
- The "view" operation gives a structured indexing form to a flat 1-D buffer.
- Unlike "subview" it can perform a type change. The type change behavior
- requires the op to have special semantics because, e.g. a byte shift of 3
- cannot be represented as an offset on f64.
- For now, a "view" op:
-
- 1. Only takes a contiguous source memref with 0 offset and empty layout.
- 2. Must specify a byte_shift operand (in the future, a special integer
- attribute may be added to support the folded case).
- 3. Returns a contiguous memref with 0 offset and empty layout.
-
- Example:
-
- ```mlir
- // Allocate a flat 1D/i8 memref.
- %0 = alloc() : memref<2048xi8>
-
- // ViewOp with dynamic offset and static sizes.
- %1 = view %0[%offset_1024][] : memref<2048xi8> to memref<64x4xf32>
-
- // ViewOp with dynamic offset and two dynamic size.
- %2 = view %0[%offset_1024][%size0, %size1] :
- memref<2048xi8> to memref<?x4x?xf32>
- ```
- }];
-
- let arguments = (ins MemRefRankOf<[I8], [1]>:$source,
- Index:$byte_shift,
- Variadic<Index>:$sizes);
- let results = (outs AnyMemRef);
-
- let extraClassDeclaration = [{
- /// The result of a view is always a memref.
- MemRefType getType() { return getResult().getType().cast<MemRefType>(); }
-
- /// Returns the dynamic sizes for this view operation. This is redundant
- /// with `sizes` but needed in template implementations. More specifically:
- /// ```
- /// template <typename AnyMemRefDefOp>
- /// bool isMemRefSizeValidSymbol(AnyMemRefDefOp memrefDefOp, unsigned index,
- /// Region *region)
- /// ```
- operand_range getDynamicSizes() {
- return {sizes().begin(), sizes().end()};
- }
- }];
-
- let hasCanonicalizer = 1;
-}
-
//===----------------------------------------------------------------------===//
// XOrOp
//===----------------------------------------------------------------------===//
diff --git a/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.h b/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.h
index f5ca215509c8..a6fdca852c80 100644
--- a/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.h
@@ -37,7 +37,7 @@ std::unique_ptr<Pass> createTensorConstantBufferizePass();
/// Creates an instance of the StdExpand pass that legalizes Std
/// dialect ops to be convertible to LLVM. For example,
/// `std.ceildivi_signed` gets transformed to a number of std operations,
-/// which can be lowered to LLVM; `memref_reshape` gets converted to
+/// which can be lowered to LLVM; `memref.reshape` gets converted to
/// `memref_reinterpret_cast`.
std::unique_ptr<Pass> createStdExpandOpsPass();
diff --git a/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.td b/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.td
index 9b27f2f03677..923065377437 100644
--- a/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/StandardOps/Transforms/Passes.td
@@ -44,9 +44,10 @@ def FuncBufferize : Pass<"func-bufferize", "ModuleOp"> {
implement the `ReturnLike` trait are not rewritten in general, as they
require that the corresponding parent operation is also rewritten.
Finally, this pass fails for unknown terminators, as we cannot decide
- whether they need rewriting.
+ whether they need rewriting.
}];
let constructor = "mlir::createFuncBufferizePass()";
+ let dependentDialects = ["memref::MemRefDialect"];
}
def TensorConstantBufferize : Pass<"tensor-constant-bufferize", "ModuleOp"> {
@@ -54,12 +55,13 @@ def TensorConstantBufferize : Pass<"tensor-constant-bufferize", "ModuleOp"> {
let description = [{
This pass bufferizes tensor constants.
- This pass needs to be a module pass because it inserts std.global_memref
+ This pass needs to be a module pass because it inserts memref.global
ops into the module, which cannot be done safely from a function pass due to
multi-threading. Most other bufferization passes can run in parallel at
function granularity.
}];
let constructor = "mlir::createTensorConstantBufferizePass()";
+ let dependentDialects = ["memref::MemRefDialect"];
}
#endif // MLIR_DIALECT_STANDARD_TRANSFORMS_PASSES
diff --git a/mlir/include/mlir/Dialect/StandardOps/Utils/Utils.h b/mlir/include/mlir/Dialect/StandardOps/Utils/Utils.h
index 8d46d9bfd278..ae4fcd01609b 100644
--- a/mlir/include/mlir/Dialect/StandardOps/Utils/Utils.h
+++ b/mlir/include/mlir/Dialect/StandardOps/Utils/Utils.h
@@ -16,6 +16,9 @@
#ifndef MLIR_DIALECT_STANDARDOPS_UTILS_UTILS_H
#define MLIR_DIALECT_STANDARDOPS_UTILS_UTILS_H
+#include "mlir/Dialect/StandardOps/IR/Ops.h"
+#include "mlir/IR/Matchers.h"
+#include "mlir/IR/PatternMatch.h"
#include "mlir/IR/Value.h"
namespace mlir {
@@ -27,6 +30,51 @@ class OpBuilder;
/// constructing the necessary DimOp operators.
SmallVector<Value, 4> getDynOperands(Location loc, Value val, OpBuilder &b);
+/// Matches a ConstantIndexOp.
+detail::op_matcher<ConstantIndexOp> matchConstantIndex();
+
+/// Detects the `values` produced by a ConstantIndexOp and places the new
+/// constant in place of the corresponding sentinel value.
+void canonicalizeSubViewPart(SmallVectorImpl<OpFoldResult> &values,
+ function_ref<bool(int64_t)> isDynamic);
+
+void getPositionsOfShapeOne(unsigned rank, ArrayRef<int64_t> shape,
+ llvm::SmallDenseSet<unsigned> &dimsToProject);
+
+/// Pattern to rewrite a subview op with constant arguments.
+template <typename OpType, typename CastOpFunc>
+class OpWithOffsetSizesAndStridesConstantArgumentFolder final
+ : public OpRewritePattern<OpType> {
+public:
+ using OpRewritePattern<OpType>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(OpType op,
+ PatternRewriter &rewriter) const override {
+ // No constant operand, just return;
+ if (llvm::none_of(op.getOperands(), [](Value operand) {
+ return matchPattern(operand, matchConstantIndex());
+ }))
+ return failure();
+
+ // At least one of offsets/sizes/strides is a new constant.
+ // Form the new list of operands and constant attributes from the existing.
+ SmallVector<OpFoldResult> mixedOffsets(op.getMixedOffsets());
+ SmallVector<OpFoldResult> mixedSizes(op.getMixedSizes());
+ SmallVector<OpFoldResult> mixedStrides(op.getMixedStrides());
+ canonicalizeSubViewPart(mixedOffsets, ShapedType::isDynamicStrideOrOffset);
+ canonicalizeSubViewPart(mixedSizes, ShapedType::isDynamic);
+ canonicalizeSubViewPart(mixedStrides, ShapedType::isDynamicStrideOrOffset);
+
+ // Create the new op in canonical form.
+ auto newOp = rewriter.create<OpType>(op.getLoc(), op.source(), mixedOffsets,
+ mixedSizes, mixedStrides);
+ CastOpFunc func;
+ func(rewriter, op, newOp);
+
+ return success();
+ }
+};
+
} // end namespace mlir
#endif // MLIR_DIALECT_STANDARDOPS_UTILS_UTILS_H
diff --git a/mlir/include/mlir/Dialect/Vector/VectorTransforms.h b/mlir/include/mlir/Dialect/Vector/VectorTransforms.h
index a258903d5a3a..ff3dbfdcad1f 100644
--- a/mlir/include/mlir/Dialect/Vector/VectorTransforms.h
+++ b/mlir/include/mlir/Dialect/Vector/VectorTransforms.h
@@ -180,11 +180,11 @@ struct UnrollVectorPattern : public RewritePattern {
/// ```
/// %1:3 = scf.if (%inBounds) {
/// // fastpath, direct cast
-/// memref_cast %A: memref<A...> to compatibleMemRefType
+/// memref.cast %A: memref<A...> to compatibleMemRefType
/// scf.yield %view : compatibleMemRefType, index, index
/// } else {
/// // slowpath, masked vector.transfer or linalg.copy.
-/// memref_cast %alloc: memref<B...> to compatibleMemRefType
+/// memref.cast %alloc: memref<B...> to compatibleMemRefType
/// scf.yield %4 : compatibleMemRefType, index, index
// }
/// %0 = vector.transfer_read %1#0[%1#1, %1#2] {masked = [false ... false]}
diff --git a/mlir/include/mlir/IR/OpDefinition.h b/mlir/include/mlir/IR/OpDefinition.h
index 9fa7e44767f9..8b3b052590a9 100644
--- a/mlir/include/mlir/IR/OpDefinition.h
+++ b/mlir/include/mlir/IR/OpDefinition.h
@@ -1133,7 +1133,7 @@ class AffineScope : public TraitBase<ConcreteType, AffineScope> {
/// A trait of region holding operations that define a new scope for automatic
/// allocations, i.e., allocations that are freed when control is transferred
/// back from the operation's region. Any operations performing such allocations
-/// (for eg. std.alloca) will have their allocations automatically freed at
+/// (for eg. memref.alloca) will have their allocations automatically freed at
/// their closest enclosing operation with this trait.
template <typename ConcreteType>
class AutomaticAllocationScope
diff --git a/mlir/include/mlir/InitAllDialects.h b/mlir/include/mlir/InitAllDialects.h
index e19072493a20..92cb5a3c3bcc 100644
--- a/mlir/include/mlir/InitAllDialects.h
+++ b/mlir/include/mlir/InitAllDialects.h
@@ -28,6 +28,7 @@
#include "mlir/Dialect/LLVMIR/ROCDLDialect.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
#include "mlir/Dialect/Math/IR/Math.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/OpenACC/OpenACC.h"
#include "mlir/Dialect/OpenMP/OpenMPDialect.h"
#include "mlir/Dialect/PDL/IR/PDL.h"
@@ -60,6 +61,7 @@ inline void registerAllDialects(DialectRegistry ®istry) {
LLVM::LLVMArmSVEDialect,
linalg::LinalgDialect,
math::MathDialect,
+ memref::MemRefDialect,
scf::SCFDialect,
omp::OpenMPDialect,
pdl::PDLDialect,
diff --git a/mlir/include/mlir/Transforms/Bufferize.h b/mlir/include/mlir/Transforms/Bufferize.h
index 8e1f9c0fe1f7..29e16c2ecb1c 100644
--- a/mlir/include/mlir/Transforms/Bufferize.h
+++ b/mlir/include/mlir/Transforms/Bufferize.h
@@ -54,7 +54,7 @@ void populateBufferizeMaterializationLegality(ConversionTarget &target);
/// Populate patterns to eliminate bufferize materializations.
///
-/// In particular, these are the tensor_load/tensor_to_memref ops.
+/// In particular, these are the tensor_load/buffer_cast ops.
void populateEliminateBufferizeMaterializationsPatterns(
MLIRContext *context, BufferizeTypeConverter &typeConverter,
OwningRewritePatternList &patterns);
diff --git a/mlir/include/mlir/Transforms/Passes.h b/mlir/include/mlir/Transforms/Passes.h
index b9e8c48c0ccd..60ea4b188ae1 100644
--- a/mlir/include/mlir/Transforms/Passes.h
+++ b/mlir/include/mlir/Transforms/Passes.h
@@ -54,7 +54,7 @@ std::unique_ptr<Pass>
createPromoteBuffersToStackPass(std::function<bool(Value)> isSmallAlloc);
/// Creates a pass that finalizes a partial bufferization by removing remaining
-/// tensor_load and tensor_to_memref operations.
+/// tensor_load and buffer_cast operations.
std::unique_ptr<FunctionPass> createFinalizingBufferizePass();
/// Creates a pass that converts memref function results to out-params.
diff --git a/mlir/include/mlir/Transforms/Passes.td b/mlir/include/mlir/Transforms/Passes.td
index 925e7e79841d..2305c4a39191 100644
--- a/mlir/include/mlir/Transforms/Passes.td
+++ b/mlir/include/mlir/Transforms/Passes.td
@@ -352,7 +352,7 @@ def BufferResultsToOutParams : Pass<"buffer-results-to-out-params", "ModuleOp">
works for static shaped memrefs.
}];
let constructor = "mlir::createBufferResultsToOutParamsPass()";
- let dependentDialects = ["linalg::LinalgDialect"];
+ let dependentDialects = ["linalg::LinalgDialect", "memref::MemRefDialect"];
}
def Canonicalizer : Pass<"canonicalize"> {
@@ -363,6 +363,7 @@ def Canonicalizer : Pass<"canonicalize"> {
details.
}];
let constructor = "mlir::createCanonicalizerPass()";
+ let dependentDialects = ["memref::MemRefDialect"];
}
def CopyRemoval : FunctionPass<"copy-removal"> {
@@ -406,11 +407,11 @@ def FinalizingBufferize : FunctionPass<"finalizing-bufferize"> {
let summary = "Finalize a partial bufferization";
let description = [{
A bufferize pass that finalizes a partial bufferization by removing
- remaining `tensor_load` and `tensor_to_memref` operations.
+ remaining `memref.tensor_load` and `memref.buffer_cast` operations.
The removal of those operations is only possible if the operations only
- exist in pairs, i.e., all uses of `tensor_load` operations are
- `tensor_to_memref` operations.
+ exist in pairs, i.e., all uses of `memref.tensor_load` operations are
+ `memref.buffer_cast` operations.
This pass will fail if not all operations can be removed or if any operation
with tensor typed operands remains.
@@ -535,7 +536,7 @@ def NormalizeMemRefs : Pass<"normalize-memrefs", "ModuleOp"> {
contained in the op. Operations marked with the [MemRefsNormalizable]
(https://mlir.llvm.org/docs/Traits/#memrefsnormalizable) trait are
expected to be normalizable. Supported operations include affine
- operations, std.alloc, std.dealloc, and std.return.
+ operations, memref.alloc, memref.dealloc, and std.return.
Given an appropriate layout map specified in the code, this transformation
can express tiled or linearized access to multi-dimensional data
diff --git a/mlir/include/mlir/Transforms/Utils.h b/mlir/include/mlir/Transforms/Utils.h
index 81b4dfd0a01b..8bbde7c15b34 100644
--- a/mlir/include/mlir/Transforms/Utils.h
+++ b/mlir/include/mlir/Transforms/Utils.h
@@ -28,6 +28,10 @@ class AffineForOp;
class Location;
class OpBuilder;
+namespace memref {
+class AllocOp;
+} // end namespace memref
+
/// Replaces all "dereferencing" uses of `oldMemRef` with `newMemRef` while
/// optionally remapping the old memref's indices using the supplied affine map,
/// `indexRemap`. The new memref could be of a
diff erent shape or rank.
@@ -88,7 +92,7 @@ LogicalResult replaceAllMemRefUsesWith(Value oldMemRef, Value newMemRef,
/// Rewrites the memref defined by this alloc op to have an identity layout map
/// and updates all its indexing uses. Returns failure if any of its uses
/// escape (while leaving the IR in a valid state).
-LogicalResult normalizeMemRef(AllocOp op);
+LogicalResult normalizeMemRef(memref::AllocOp *op);
/// Uses the old memref type map layout and computes the new memref type to have
/// a new shape and a layout map, where the old layout map has been normalized
diff --git a/mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp b/mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp
index ef04f688ba39..c7b5183e07c9 100644
--- a/mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp
+++ b/mlir/lib/Conversion/AffineToStandard/AffineToStandard.cpp
@@ -15,6 +15,7 @@
#include "../PassDetail.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Vector/VectorOps.h"
@@ -44,7 +45,8 @@ class AffineApplyExpander
: builder(builder), dimValues(dimValues), symbolValues(symbolValues),
loc(loc) {}
- template <typename OpTy> Value buildBinaryExpr(AffineBinaryOpExpr expr) {
+ template <typename OpTy>
+ Value buildBinaryExpr(AffineBinaryOpExpr expr) {
auto lhs = visit(expr.getLHS());
auto rhs = visit(expr.getRHS());
if (!lhs || !rhs)
@@ -563,8 +565,8 @@ class AffineApplyLowering : public OpRewritePattern<AffineApplyOp> {
};
/// Apply the affine map from an 'affine.load' operation to its operands, and
-/// feed the results to a newly created 'std.load' operation (which replaces the
-/// original 'affine.load').
+/// feed the results to a newly created 'memref.load' operation (which replaces
+/// the original 'affine.load').
class AffineLoadLowering : public OpRewritePattern<AffineLoadOp> {
public:
using OpRewritePattern<AffineLoadOp>::OpRewritePattern;
@@ -579,14 +581,14 @@ class AffineLoadLowering : public OpRewritePattern<AffineLoadOp> {
return failure();
// Build vector.load memref[expandedMap.results].
- rewriter.replaceOpWithNewOp<mlir::LoadOp>(op, op.getMemRef(),
- *resultOperands);
+ rewriter.replaceOpWithNewOp<memref::LoadOp>(op, op.getMemRef(),
+ *resultOperands);
return success();
}
};
/// Apply the affine map from an 'affine.prefetch' operation to its operands,
-/// and feed the results to a newly created 'std.prefetch' operation (which
+/// and feed the results to a newly created 'memref.prefetch' operation (which
/// replaces the original 'affine.prefetch').
class AffinePrefetchLowering : public OpRewritePattern<AffinePrefetchOp> {
public:
@@ -601,16 +603,16 @@ class AffinePrefetchLowering : public OpRewritePattern<AffinePrefetchOp> {
if (!resultOperands)
return failure();
- // Build std.prefetch memref[expandedMap.results].
- rewriter.replaceOpWithNewOp<PrefetchOp>(op, op.memref(), *resultOperands,
- op.isWrite(), op.localityHint(),
- op.isDataCache());
+ // Build memref.prefetch memref[expandedMap.results].
+ rewriter.replaceOpWithNewOp<memref::PrefetchOp>(
+ op, op.memref(), *resultOperands, op.isWrite(), op.localityHint(),
+ op.isDataCache());
return success();
}
};
/// Apply the affine map from an 'affine.store' operation to its operands, and
-/// feed the results to a newly created 'std.store' operation (which replaces
+/// feed the results to a newly created 'memref.store' operation (which replaces
/// the original 'affine.store').
class AffineStoreLowering : public OpRewritePattern<AffineStoreOp> {
public:
@@ -625,8 +627,8 @@ class AffineStoreLowering : public OpRewritePattern<AffineStoreOp> {
if (!maybeExpandedMap)
return failure();
- // Build std.store valueToStore, memref[expandedMap.results].
- rewriter.replaceOpWithNewOp<mlir::StoreOp>(
+ // Build memref.store valueToStore, memref[expandedMap.results].
+ rewriter.replaceOpWithNewOp<memref::StoreOp>(
op, op.getValueToStore(), op.getMemRef(), *maybeExpandedMap);
return success();
}
@@ -634,7 +636,8 @@ class AffineStoreLowering : public OpRewritePattern<AffineStoreOp> {
/// Apply the affine maps from an 'affine.dma_start' operation to each of their
/// respective map operands, and feed the results to a newly created
-/// 'std.dma_start' operation (which replaces the original 'affine.dma_start').
+/// 'memref.dma_start' operation (which replaces the original
+/// 'affine.dma_start').
class AffineDmaStartLowering : public OpRewritePattern<AffineDmaStartOp> {
public:
using OpRewritePattern<AffineDmaStartOp>::OpRewritePattern;
@@ -663,8 +666,8 @@ class AffineDmaStartLowering : public OpRewritePattern<AffineDmaStartOp> {
if (!maybeExpandedTagMap)
return failure();
- // Build std.dma_start operation with affine map results.
- rewriter.replaceOpWithNewOp<DmaStartOp>(
+ // Build memref.dma_start operation with affine map results.
+ rewriter.replaceOpWithNewOp<memref::DmaStartOp>(
op, op.getSrcMemRef(), *maybeExpandedSrcMap, op.getDstMemRef(),
*maybeExpandedDstMap, op.getNumElements(), op.getTagMemRef(),
*maybeExpandedTagMap, op.getStride(), op.getNumElementsPerStride());
@@ -673,7 +676,7 @@ class AffineDmaStartLowering : public OpRewritePattern<AffineDmaStartOp> {
};
/// Apply the affine map from an 'affine.dma_wait' operation tag memref,
-/// and feed the results to a newly created 'std.dma_wait' operation (which
+/// and feed the results to a newly created 'memref.dma_wait' operation (which
/// replaces the original 'affine.dma_wait').
class AffineDmaWaitLowering : public OpRewritePattern<AffineDmaWaitOp> {
public:
@@ -688,8 +691,8 @@ class AffineDmaWaitLowering : public OpRewritePattern<AffineDmaWaitOp> {
if (!maybeExpandedTagMap)
return failure();
- // Build std.dma_wait operation with affine map results.
- rewriter.replaceOpWithNewOp<DmaWaitOp>(
+ // Build memref.dma_wait operation with affine map results.
+ rewriter.replaceOpWithNewOp<memref::DmaWaitOp>(
op, op.getTagMemRef(), *maybeExpandedTagMap, op.getNumElements());
return success();
}
@@ -777,8 +780,8 @@ class LowerAffinePass : public ConvertAffineToStandardBase<LowerAffinePass> {
populateAffineToStdConversionPatterns(patterns, &getContext());
populateAffineToVectorConversionPatterns(patterns, &getContext());
ConversionTarget target(getContext());
- target
- .addLegalDialect<scf::SCFDialect, StandardOpsDialect, VectorDialect>();
+ target.addLegalDialect<memref::MemRefDialect, scf::SCFDialect,
+ StandardOpsDialect, VectorDialect>();
if (failed(applyPartialConversion(getOperation(), target,
std::move(patterns))))
signalPassFailure();
diff --git a/mlir/lib/Conversion/AffineToStandard/CMakeLists.txt b/mlir/lib/Conversion/AffineToStandard/CMakeLists.txt
index 45c398195d15..a3412348a442 100644
--- a/mlir/lib/Conversion/AffineToStandard/CMakeLists.txt
+++ b/mlir/lib/Conversion/AffineToStandard/CMakeLists.txt
@@ -12,6 +12,7 @@ add_mlir_conversion_library(MLIRAffineToStandard
LINK_LIBS PUBLIC
MLIRAffine
+ MLIRMemRef
MLIRSCF
MLIRPass
MLIRStandard
diff --git a/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt b/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt
index 50f9e6a295e0..e2647103ba22 100644
--- a/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt
+++ b/mlir/lib/Conversion/GPUToNVVM/CMakeLists.txt
@@ -13,6 +13,7 @@ add_mlir_conversion_library(MLIRGPUToNVVMTransforms
MLIRGPU
MLIRGPUToGPURuntimeTransforms
MLIRLLVMIR
+ MLIRMemRef
MLIRNVVMIR
MLIRPass
MLIRStandardToLLVM
diff --git a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
index bc4f65182bdb..9e167125904a 100644
--- a/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
+++ b/mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
@@ -18,6 +18,7 @@
#include "mlir/Dialect/GPU/Passes.h"
#include "mlir/Dialect/LLVMIR/NVVMDialect.h"
#include "mlir/Dialect/Math/IR/Math.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/BlockAndValueMapping.h"
#include "mlir/Transforms/DialectConversion.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
diff --git a/mlir/lib/Conversion/LinalgToStandard/CMakeLists.txt b/mlir/lib/Conversion/LinalgToStandard/CMakeLists.txt
index b38a4b8e9f66..11639ce088ff 100644
--- a/mlir/lib/Conversion/LinalgToStandard/CMakeLists.txt
+++ b/mlir/lib/Conversion/LinalgToStandard/CMakeLists.txt
@@ -14,6 +14,7 @@ add_mlir_conversion_library(MLIRLinalgToStandard
MLIREDSC
MLIRIR
MLIRLinalg
+ MLIRMemRef
MLIRPass
MLIRSCF
MLIRTransforms
diff --git a/mlir/lib/Conversion/LinalgToStandard/LinalgToStandard.cpp b/mlir/lib/Conversion/LinalgToStandard/LinalgToStandard.cpp
index 276b124a9f10..bf947a4ed088 100644
--- a/mlir/lib/Conversion/LinalgToStandard/LinalgToStandard.cpp
+++ b/mlir/lib/Conversion/LinalgToStandard/LinalgToStandard.cpp
@@ -12,6 +12,7 @@
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
@@ -93,7 +94,7 @@ createTypeCanonicalizedMemRefOperands(OpBuilder &b, Location loc,
continue;
}
Value cast =
- b.create<MemRefCastOp>(loc, eraseStridedLayout(memrefType), op);
+ b.create<memref::CastOp>(loc, eraseStridedLayout(memrefType), op);
res.push_back(cast);
}
return res;
@@ -143,12 +144,12 @@ LogicalResult mlir::linalg::CopyTransposeRewrite::matchAndRewrite(
// If either inputPerm or outputPerm are non-identities, insert transposes.
auto inputPerm = op.inputPermutation();
if (inputPerm.hasValue() && !inputPerm->isIdentity())
- in = rewriter.create<TransposeOp>(op.getLoc(), in,
- AffineMapAttr::get(*inputPerm));
+ in = rewriter.create<memref::TransposeOp>(op.getLoc(), in,
+ AffineMapAttr::get(*inputPerm));
auto outputPerm = op.outputPermutation();
if (outputPerm.hasValue() && !outputPerm->isIdentity())
- out = rewriter.create<TransposeOp>(op.getLoc(), out,
- AffineMapAttr::get(*outputPerm));
+ out = rewriter.create<memref::TransposeOp>(op.getLoc(), out,
+ AffineMapAttr::get(*outputPerm));
// If nothing was transposed, fail and let the conversion kick in.
if (in == op.input() && out == op.output())
@@ -213,7 +214,8 @@ struct ConvertLinalgToStandardPass
void ConvertLinalgToStandardPass::runOnOperation() {
auto module = getOperation();
ConversionTarget target(getContext());
- target.addLegalDialect<AffineDialect, scf::SCFDialect, StandardOpsDialect>();
+ target.addLegalDialect<AffineDialect, memref::MemRefDialect, scf::SCFDialect,
+ StandardOpsDialect>();
target.addLegalOp<ModuleOp, FuncOp, ModuleTerminatorOp, ReturnOp>();
target.addLegalOp<linalg::ReshapeOp, linalg::RangeOp>();
OwningRewritePatternList patterns;
diff --git a/mlir/lib/Conversion/PassDetail.h b/mlir/lib/Conversion/PassDetail.h
index aa05bcd9bc43..4e9e0861d312 100644
--- a/mlir/lib/Conversion/PassDetail.h
+++ b/mlir/lib/Conversion/PassDetail.h
@@ -38,6 +38,10 @@ namespace NVVM {
class NVVMDialect;
} // end namespace NVVM
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
namespace omp {
class OpenMPDialect;
} // end namespace omp
diff --git a/mlir/lib/Conversion/SCFToGPU/CMakeLists.txt b/mlir/lib/Conversion/SCFToGPU/CMakeLists.txt
index f3de289c28cc..6329161eb458 100644
--- a/mlir/lib/Conversion/SCFToGPU/CMakeLists.txt
+++ b/mlir/lib/Conversion/SCFToGPU/CMakeLists.txt
@@ -15,6 +15,7 @@ add_mlir_conversion_library(MLIRSCFToGPU
MLIRGPU
MLIRIR
MLIRLinalg
+ MLIRMemRef
MLIRPass
MLIRStandard
MLIRSupport
diff --git a/mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp b/mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp
index e756a3744f64..9f5e4abc42ca 100644
--- a/mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp
+++ b/mlir/lib/Conversion/SCFToGPU/SCFToGPU.cpp
@@ -18,6 +18,7 @@
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/GPU/GPUDialect.h"
#include "mlir/Dialect/GPU/ParallelLoopMapper.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/AffineExpr.h"
@@ -647,6 +648,7 @@ void mlir::populateParallelLoopToGPUPatterns(OwningRewritePatternList &patterns,
}
void mlir::configureParallelLoopToGPULegality(ConversionTarget &target) {
+ target.addLegalDialect<memref::MemRefDialect>();
target.addDynamicallyLegalOp<scf::ParallelOp>([](scf::ParallelOp parallelOp) {
return !parallelOp->getAttr(gpu::getMappingAttrName());
});
diff --git a/mlir/lib/Conversion/ShapeToStandard/CMakeLists.txt b/mlir/lib/Conversion/ShapeToStandard/CMakeLists.txt
index f65e9aec3142..5700b5ad9dc4 100644
--- a/mlir/lib/Conversion/ShapeToStandard/CMakeLists.txt
+++ b/mlir/lib/Conversion/ShapeToStandard/CMakeLists.txt
@@ -19,6 +19,7 @@ add_mlir_conversion_library(MLIRShapeToStandard
LINK_LIBS PUBLIC
MLIREDSC
MLIRIR
+ MLIRMemRef
MLIRShape
MLIRTensor
MLIRPass
diff --git a/mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp b/mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp
index 49c44ad78e8f..2c06702e32c1 100644
--- a/mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp
+++ b/mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp
@@ -9,6 +9,7 @@
#include "mlir/Conversion/ShapeToStandard/ShapeToStandard.h"
#include "../PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/Shape/IR/Shape.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
@@ -139,7 +140,7 @@ LogicalResult BroadcastOpConverter::matchAndRewrite(
// dimension in the tensor.
SmallVector<Value> ranks, rankDiffs;
llvm::append_range(ranks, llvm::map_range(transformed.shapes(), [&](Value v) {
- return lb.create<DimOp>(v, zero);
+ return lb.create<memref::DimOp>(v, zero);
}));
// Find the maximum rank
@@ -252,7 +253,7 @@ LogicalResult IsBroadcastableOpConverter::matchAndRewrite(
// dimension in the tensor.
SmallVector<Value> ranks, rankDiffs;
llvm::append_range(ranks, llvm::map_range(transformed.shapes(), [&](Value v) {
- return lb.create<DimOp>(v, zero);
+ return lb.create<memref::DimOp>(v, zero);
}));
// Find the maximum rank
@@ -344,8 +345,8 @@ LogicalResult GetExtentOpConverter::matchAndRewrite(
// circumvents the necessity to materialize the shape in memory.
if (auto shapeOfOp = op.shape().getDefiningOp<ShapeOfOp>()) {
if (shapeOfOp.arg().getType().isa<ShapedType>()) {
- rewriter.replaceOpWithNewOp<DimOp>(op, shapeOfOp.arg(),
- transformed.dim());
+ rewriter.replaceOpWithNewOp<memref::DimOp>(op, shapeOfOp.arg(),
+ transformed.dim());
return success();
}
}
@@ -375,7 +376,7 @@ RankOpConverter::matchAndRewrite(shape::RankOp op, ArrayRef<Value> operands,
return failure();
shape::RankOp::Adaptor transformed(operands);
- rewriter.replaceOpWithNewOp<DimOp>(op, transformed.shape(), 0);
+ rewriter.replaceOpWithNewOp<memref::DimOp>(op, transformed.shape(), 0);
return success();
}
@@ -404,7 +405,8 @@ ReduceOpConverter::matchAndRewrite(shape::ReduceOp op, ArrayRef<Value> operands,
Value zero = rewriter.create<ConstantIndexOp>(loc, 0);
Value one = rewriter.create<ConstantIndexOp>(loc, 1);
Type indexTy = rewriter.getIndexType();
- Value rank = rewriter.create<DimOp>(loc, indexTy, transformed.shape(), zero);
+ Value rank =
+ rewriter.create<memref::DimOp>(loc, indexTy, transformed.shape(), zero);
auto loop = rewriter.create<scf::ForOp>(
loc, zero, rank, one, op.initVals(),
@@ -490,11 +492,12 @@ ShapeEqOpConverter::matchAndRewrite(ShapeEqOp op, ArrayRef<Value> operands,
Type indexTy = rewriter.getIndexType();
Value zero = rewriter.create<ConstantIndexOp>(loc, 0);
Value firstShape = transformed.shapes().front();
- Value firstRank = rewriter.create<DimOp>(loc, indexTy, firstShape, zero);
+ Value firstRank =
+ rewriter.create<memref::DimOp>(loc, indexTy, firstShape, zero);
Value result = nullptr;
// Generate a linear sequence of compares, all with firstShape as lhs.
for (Value shape : transformed.shapes().drop_front(1)) {
- Value rank = rewriter.create<DimOp>(loc, indexTy, shape, zero);
+ Value rank = rewriter.create<memref::DimOp>(loc, indexTy, shape, zero);
Value eqRank =
rewriter.create<CmpIOp>(loc, CmpIPredicate::eq, firstRank, rank);
auto same = rewriter.create<IfOp>(
@@ -559,7 +562,7 @@ LogicalResult ShapeOfOpConversion::matchAndRewrite(
int64_t rank = rankedTensorTy.getRank();
for (int64_t i = 0; i < rank; i++) {
if (rankedTensorTy.isDynamicDim(i)) {
- Value extent = rewriter.create<DimOp>(loc, tensor, i);
+ Value extent = rewriter.create<memref::DimOp>(loc, tensor, i);
extentValues.push_back(extent);
} else {
Value extent =
@@ -583,7 +586,7 @@ LogicalResult ShapeOfOpConversion::matchAndRewrite(
op, getExtentTensorType(ctx), ValueRange{rank},
[&](OpBuilder &b, Location loc, ValueRange args) {
Value dim = args.front();
- Value extent = b.create<DimOp>(loc, tensor, dim);
+ Value extent = b.create<memref::DimOp>(loc, tensor, dim);
b.create<tensor::YieldOp>(loc, extent);
});
@@ -613,7 +616,7 @@ LogicalResult SplitAtOpConversion::matchAndRewrite(
SplitAtOp::Adaptor transformed(op);
ImplicitLocOpBuilder b(op.getLoc(), rewriter);
Value zero = b.create<ConstantIndexOp>(0);
- Value rank = b.create<DimOp>(transformed.operand(), zero);
+ Value rank = b.create<memref::DimOp>(transformed.operand(), zero);
// index < 0 ? index + rank : index
Value originalIndex = transformed.index();
@@ -670,8 +673,8 @@ void ConvertShapeToStandardPass::runOnOperation() {
// Setup target legality.
MLIRContext &ctx = getContext();
ConversionTarget target(ctx);
- target
- .addLegalDialect<StandardOpsDialect, SCFDialect, tensor::TensorDialect>();
+ target.addLegalDialect<memref::MemRefDialect, StandardOpsDialect, SCFDialect,
+ tensor::TensorDialect>();
target.addLegalOp<CstrRequireOp, FuncOp, ModuleOp, ModuleTerminatorOp>();
// Setup conversion patterns.
diff --git a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt b/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt
index 6ef145ba8589..e1e13f0b1cc2 100644
--- a/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt
+++ b/mlir/lib/Conversion/StandardToLLVM/CMakeLists.txt
@@ -14,5 +14,6 @@ add_mlir_conversion_library(MLIRStandardToLLVM
LINK_LIBS PUBLIC
MLIRLLVMIR
MLIRMath
+ MLIRMemRef
MLIRTransforms
)
diff --git a/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp b/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp
index 3601af43c73d..b3a2bb634f39 100644
--- a/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp
+++ b/mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp
@@ -17,6 +17,7 @@
#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/IR/Attributes.h"
#include "mlir/IR/BlockAndValueMapping.h"
@@ -1864,13 +1865,13 @@ struct AllocLikeOpLowering : public ConvertToLLVMPattern {
struct AllocOpLowering : public AllocLikeOpLowering {
AllocOpLowering(LLVMTypeConverter &converter)
- : AllocLikeOpLowering(AllocOp::getOperationName(), converter) {}
+ : AllocLikeOpLowering(memref::AllocOp::getOperationName(), converter) {}
std::tuple<Value, Value> allocateBuffer(ConversionPatternRewriter &rewriter,
Location loc, Value sizeBytes,
Operation *op) const override {
// Heap allocations.
- AllocOp allocOp = cast<AllocOp>(op);
+ memref::AllocOp allocOp = cast<memref::AllocOp>(op);
MemRefType memRefType = allocOp.getType();
Value alignment;
@@ -1917,7 +1918,7 @@ struct AllocOpLowering : public AllocLikeOpLowering {
struct AlignedAllocOpLowering : public AllocLikeOpLowering {
AlignedAllocOpLowering(LLVMTypeConverter &converter)
- : AllocLikeOpLowering(AllocOp::getOperationName(), converter) {}
+ : AllocLikeOpLowering(memref::AllocOp::getOperationName(), converter) {}
/// Returns the memref's element size in bytes.
// TODO: there are other places where this is used. Expose publicly?
@@ -1950,7 +1951,7 @@ struct AlignedAllocOpLowering : public AllocLikeOpLowering {
/// 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(AllocOp allocOp) const {
+ int64_t getAllocationAlignment(memref::AllocOp allocOp) const {
if (Optional<uint64_t> alignment = allocOp.alignment())
return *alignment;
@@ -1966,7 +1967,7 @@ struct AlignedAllocOpLowering : public AllocLikeOpLowering {
Location loc, Value sizeBytes,
Operation *op) const override {
// Heap allocations.
- AllocOp allocOp = cast<AllocOp>(op);
+ memref::AllocOp allocOp = cast<memref::AllocOp>(op);
MemRefType memRefType = allocOp.getType();
int64_t alignment = getAllocationAlignment(allocOp);
Value allocAlignment = createIndexConstant(rewriter, loc, alignment);
@@ -1997,7 +1998,7 @@ constexpr uint64_t AlignedAllocOpLowering::kMinAlignedAllocAlignment;
struct AllocaOpLowering : public AllocLikeOpLowering {
AllocaOpLowering(LLVMTypeConverter &converter)
- : AllocLikeOpLowering(AllocaOp::getOperationName(), converter) {}
+ : AllocLikeOpLowering(memref::AllocaOp::getOperationName(), converter) {}
/// Allocates the underlying buffer using the right call. `allocatedBytePtr`
/// is set to null for stack allocations. `accessAlignment` is set if
@@ -2008,7 +2009,7 @@ struct AllocaOpLowering : public AllocLikeOpLowering {
// With alloca, one gets a pointer to the element type right away.
// For stack allocations.
- auto allocaOp = cast<AllocaOp>(op);
+ auto allocaOp = cast<memref::AllocaOp>(op);
auto elementPtrType = this->getElementPtrType(allocaOp.getType());
auto allocatedElementPtr = rewriter.create<LLVM::AllocaOp>(
@@ -2180,17 +2181,17 @@ struct CallIndirectOpLowering : public CallOpInterfaceLowering<CallIndirectOp> {
// 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<DeallocOp> {
- using ConvertOpToLLVMPattern<DeallocOp>::ConvertOpToLLVMPattern;
+struct DeallocOpLowering : public ConvertOpToLLVMPattern<memref::DeallocOp> {
+ using ConvertOpToLLVMPattern<memref::DeallocOp>::ConvertOpToLLVMPattern;
explicit DeallocOpLowering(LLVMTypeConverter &converter)
- : ConvertOpToLLVMPattern<DeallocOp>(converter) {}
+ : ConvertOpToLLVMPattern<memref::DeallocOp>(converter) {}
LogicalResult
- matchAndRewrite(DeallocOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::DeallocOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
assert(operands.size() == 1 && "dealloc takes one operand");
- DeallocOp::Adaptor transformed(operands);
+ memref::DeallocOp::Adaptor transformed(operands);
// Insert the `free` declaration if it is not already present.
auto freeFunc = LLVM::lookupOrCreateFreeFn(op->getParentOfType<ModuleOp>());
@@ -2209,7 +2210,7 @@ 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 global_memref's with an initial value,
+ // 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 = unwrap(typeConverter.convertType(type.getElementType()));
@@ -2221,11 +2222,12 @@ static Type convertGlobalMemrefTypeToLLVM(MemRefType type,
}
/// GlobalMemrefOp is lowered to a LLVM Global Variable.
-struct GlobalMemrefOpLowering : public ConvertOpToLLVMPattern<GlobalMemrefOp> {
- using ConvertOpToLLVMPattern<GlobalMemrefOp>::ConvertOpToLLVMPattern;
+struct GlobalMemrefOpLowering
+ : public ConvertOpToLLVMPattern<memref::GlobalOp> {
+ using ConvertOpToLLVMPattern<memref::GlobalOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(GlobalMemrefOp global, ArrayRef<Value> operands,
+ matchAndRewrite(memref::GlobalOp global, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
MemRefType type = global.type().cast<MemRefType>();
if (!isConvertibleAndHasIdentityMaps(type))
@@ -2259,14 +2261,15 @@ struct GlobalMemrefOpLowering : public ConvertOpToLLVMPattern<GlobalMemrefOp> {
/// `AllocLikeOpLowering` to reuse the Memref descriptor construction.
struct GetGlobalMemrefOpLowering : public AllocLikeOpLowering {
GetGlobalMemrefOpLowering(LLVMTypeConverter &converter)
- : AllocLikeOpLowering(GetGlobalMemrefOp::getOperationName(), converter) {}
+ : AllocLikeOpLowering(memref::GetGlobalOp::getOperationName(),
+ converter) {}
- /// Buffer "allocation" for get_global_memref op is getting the address of
+ /// Buffer "allocation" for memref.get_global op is getting the address of
/// the global variable referenced.
std::tuple<Value, Value> allocateBuffer(ConversionPatternRewriter &rewriter,
Location loc, Value sizeBytes,
Operation *op) const override {
- auto getGlobalOp = cast<GetGlobalMemrefOp>(op);
+ auto getGlobalOp = cast<memref::GetGlobalOp>(op);
MemRefType type = getGlobalOp.result().getType().cast<MemRefType>();
unsigned memSpace = type.getMemorySpaceAsInt();
@@ -2285,7 +2288,7 @@ struct GetGlobalMemrefOpLowering : public AllocLikeOpLowering {
createIndexConstant(rewriter, loc, 0));
auto gep = rewriter.create<LLVM::GEPOp>(loc, elementPtrType, operands);
- // We do not expect the memref obtained using `get_global_memref` to be
+ // 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);
@@ -2354,17 +2357,17 @@ struct RsqrtOpLowering : public ConvertOpToLLVMPattern<math::RsqrtOp> {
}
};
-struct MemRefCastOpLowering : public ConvertOpToLLVMPattern<MemRefCastOp> {
- using ConvertOpToLLVMPattern<MemRefCastOp>::ConvertOpToLLVMPattern;
+struct MemRefCastOpLowering : public ConvertOpToLLVMPattern<memref::CastOp> {
+ using ConvertOpToLLVMPattern<memref::CastOp>::ConvertOpToLLVMPattern;
- LogicalResult match(MemRefCastOp memRefCastOp) const override {
+ LogicalResult match(memref::CastOp memRefCastOp) const override {
Type srcType = memRefCastOp.getOperand().getType();
Type dstType = memRefCastOp.getType();
- // MemRefCastOp 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
+ // 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<MemRefType>() && dstType.isa<MemRefType>())
return success(typeConverter->convertType(srcType) ==
@@ -2381,9 +2384,9 @@ struct MemRefCastOpLowering : public ConvertOpToLLVMPattern<MemRefCastOp> {
: failure();
}
- void rewrite(MemRefCastOp memRefCastOp, ArrayRef<Value> operands,
+ void rewrite(memref::CastOp memRefCastOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- MemRefCastOp::Adaptor transformed(operands);
+ memref::CastOp::Adaptor transformed(operands);
auto srcType = memRefCastOp.getOperand().getType();
auto dstType = memRefCastOp.getType();
@@ -2486,14 +2489,15 @@ static void extractPointersAndOffset(Location loc,
}
struct MemRefReinterpretCastOpLowering
- : public ConvertOpToLLVMPattern<MemRefReinterpretCastOp> {
- using ConvertOpToLLVMPattern<MemRefReinterpretCastOp>::ConvertOpToLLVMPattern;
+ : public ConvertOpToLLVMPattern<memref::ReinterpretCastOp> {
+ using ConvertOpToLLVMPattern<
+ memref::ReinterpretCastOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(MemRefReinterpretCastOp castOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::ReinterpretCastOp castOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- MemRefReinterpretCastOp::Adaptor adaptor(operands,
- castOp->getAttrDictionary());
+ memref::ReinterpretCastOp::Adaptor adaptor(operands,
+ castOp->getAttrDictionary());
Type srcType = castOp.source().getType();
Value descriptor;
@@ -2505,11 +2509,10 @@ struct MemRefReinterpretCastOpLowering
}
private:
- LogicalResult
- convertSourceMemRefToDescriptor(ConversionPatternRewriter &rewriter,
- Type srcType, MemRefReinterpretCastOp castOp,
- MemRefReinterpretCastOp::Adaptor adaptor,
- Value *descriptor) const {
+ LogicalResult convertSourceMemRefToDescriptor(
+ ConversionPatternRewriter &rewriter, Type srcType,
+ memref::ReinterpretCastOp castOp,
+ memref::ReinterpretCastOp::Adaptor adaptor, Value *descriptor) const {
MemRefType targetMemRefType =
castOp.getResult().getType().cast<MemRefType>();
auto llvmTargetDescriptorTy = typeConverter->convertType(targetMemRefType)
@@ -2555,14 +2558,14 @@ struct MemRefReinterpretCastOpLowering
};
struct MemRefReshapeOpLowering
- : public ConvertOpToLLVMPattern<MemRefReshapeOp> {
- using ConvertOpToLLVMPattern<MemRefReshapeOp>::ConvertOpToLLVMPattern;
+ : public ConvertOpToLLVMPattern<memref::ReshapeOp> {
+ using ConvertOpToLLVMPattern<memref::ReshapeOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(MemRefReshapeOp reshapeOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::ReshapeOp reshapeOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto *op = reshapeOp.getOperation();
- MemRefReshapeOp::Adaptor adaptor(operands, op->getAttrDictionary());
+ memref::ReshapeOp::Adaptor adaptor(operands, op->getAttrDictionary());
Type srcType = reshapeOp.source().getType();
Value descriptor;
@@ -2576,8 +2579,8 @@ struct MemRefReshapeOpLowering
private:
LogicalResult
convertSourceMemRefToDescriptor(ConversionPatternRewriter &rewriter,
- Type srcType, MemRefReshapeOp reshapeOp,
- MemRefReshapeOp::Adaptor adaptor,
+ Type srcType, memref::ReshapeOp reshapeOp,
+ memref::ReshapeOp::Adaptor adaptor,
Value *descriptor) const {
// Conversion for statically-known shape args is performed via
// `memref_reinterpret_cast`.
@@ -2722,11 +2725,11 @@ struct DialectCastOpLowering
// 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<DimOp> {
- using ConvertOpToLLVMPattern<DimOp>::ConvertOpToLLVMPattern;
+struct DimOpLowering : public ConvertOpToLLVMPattern<memref::DimOp> {
+ using ConvertOpToLLVMPattern<memref::DimOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(DimOp dimOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::DimOp dimOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
Type operandType = dimOp.memrefOrTensor().getType();
if (operandType.isa<UnrankedMemRefType>()) {
@@ -2744,11 +2747,11 @@ struct DimOpLowering : public ConvertOpToLLVMPattern<DimOp> {
}
private:
- Value extractSizeOfUnrankedMemRef(Type operandType, DimOp dimOp,
+ Value extractSizeOfUnrankedMemRef(Type operandType, memref::DimOp dimOp,
ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
Location loc = dimOp.getLoc();
- DimOp::Adaptor transformed(operands);
+ memref::DimOp::Adaptor transformed(operands);
auto unrankedMemRefType = operandType.cast<UnrankedMemRefType>();
auto scalarMemRefType =
@@ -2785,11 +2788,11 @@ struct DimOpLowering : public ConvertOpToLLVMPattern<DimOp> {
return rewriter.create<LLVM::LoadOp>(loc, sizePtr);
}
- Value extractSizeOfRankedMemRef(Type operandType, DimOp dimOp,
+ Value extractSizeOfRankedMemRef(Type operandType, memref::DimOp dimOp,
ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
Location loc = dimOp.getLoc();
- DimOp::Adaptor transformed(operands);
+ memref::DimOp::Adaptor transformed(operands);
// Take advantage if index is constant.
MemRefType memRefType = operandType.cast<MemRefType>();
if (Optional<int64_t> index = dimOp.getConstantIndex()) {
@@ -2833,7 +2836,7 @@ struct RankOpLowering : public ConvertOpToLLVMPattern<RankOp> {
};
// Common base for load and store operations on MemRefs. Restricts the match
-// to supported MemRef types. Provides functionality to emit code accessing a
+// to supported MemRef types. Provides functionality to emit code accessing a
// specific element of the underlying data buffer.
template <typename Derived>
struct LoadStoreOpLowering : public ConvertOpToLLVMPattern<Derived> {
@@ -2849,13 +2852,13 @@ struct LoadStoreOpLowering : public ConvertOpToLLVMPattern<Derived> {
// Load operation is lowered to obtaining a pointer to the indexed element
// and loading it.
-struct LoadOpLowering : public LoadStoreOpLowering<LoadOp> {
+struct LoadOpLowering : public LoadStoreOpLowering<memref::LoadOp> {
using Base::Base;
LogicalResult
- matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::LoadOp loadOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- LoadOp::Adaptor transformed(operands);
+ memref::LoadOp::Adaptor transformed(operands);
auto type = loadOp.getMemRefType();
Value dataPtr =
@@ -2868,14 +2871,14 @@ struct LoadOpLowering : public LoadStoreOpLowering<LoadOp> {
// Store operation is lowered to obtaining a pointer to the indexed element,
// and storing the given value to it.
-struct StoreOpLowering : public LoadStoreOpLowering<StoreOp> {
+struct StoreOpLowering : public LoadStoreOpLowering<memref::StoreOp> {
using Base::Base;
LogicalResult
- matchAndRewrite(StoreOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::StoreOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto type = op.getMemRefType();
- StoreOp::Adaptor transformed(operands);
+ memref::StoreOp::Adaptor transformed(operands);
Value dataPtr =
getStridedElementPtr(op.getLoc(), type, transformed.memref(),
@@ -2888,13 +2891,13 @@ struct StoreOpLowering : public LoadStoreOpLowering<StoreOp> {
// 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<PrefetchOp> {
+struct PrefetchOpLowering : public LoadStoreOpLowering<memref::PrefetchOp> {
using Base::Base;
LogicalResult
- matchAndRewrite(PrefetchOp prefetchOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::PrefetchOp prefetchOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- PrefetchOp::Adaptor transformed(operands);
+ memref::PrefetchOp::Adaptor transformed(operands);
auto type = prefetchOp.getMemRefType();
auto loc = prefetchOp.getLoc();
@@ -3221,11 +3224,11 @@ static SmallVector<int64_t, 4> extractFromI64ArrayAttr(Attribute attr) {
/// 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<SubViewOp> {
- using ConvertOpToLLVMPattern<SubViewOp>::ConvertOpToLLVMPattern;
+struct SubViewOpLowering : public ConvertOpToLLVMPattern<memref::SubViewOp> {
+ using ConvertOpToLLVMPattern<memref::SubViewOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(SubViewOp subViewOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::SubViewOp subViewOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto loc = subViewOp.getLoc();
@@ -3234,7 +3237,7 @@ struct SubViewOpLowering : public ConvertOpToLLVMPattern<SubViewOp> {
typeConverter->convertType(sourceMemRefType.getElementType());
auto viewMemRefType = subViewOp.getType();
- auto inferredType = SubViewOp::inferResultType(
+ auto inferredType = memref::SubViewOp::inferResultType(
subViewOp.getSourceType(),
extractFromI64ArrayAttr(subViewOp.static_offsets()),
extractFromI64ArrayAttr(subViewOp.static_sizes()),
@@ -3335,7 +3338,7 @@ struct SubViewOpLowering : public ConvertOpToLLVMPattern<SubViewOp> {
if (static_cast<unsigned>(i) >= mixedSizes.size()) {
size = rewriter.create<LLVM::DialectCastOp>(
loc, llvmIndexType,
- rewriter.create<DimOp>(loc, subViewOp.source(), i));
+ rewriter.create<memref::DimOp>(loc, subViewOp.source(), i));
stride = rewriter.create<LLVM::ConstantOp>(
loc, llvmIndexType, rewriter.getI64IntegerAttr(1));
} else {
@@ -3376,15 +3379,15 @@ struct SubViewOpLowering : public ConvertOpToLLVMPattern<SubViewOp> {
/// 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<TransposeOp> {
+class TransposeOpLowering : public ConvertOpToLLVMPattern<memref::TransposeOp> {
public:
- using ConvertOpToLLVMPattern<TransposeOp>::ConvertOpToLLVMPattern;
+ using ConvertOpToLLVMPattern<memref::TransposeOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(TransposeOp transposeOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::TransposeOp transposeOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto loc = transposeOp.getLoc();
- TransposeOpAdaptor adaptor(operands);
+ memref::TransposeOpAdaptor adaptor(operands);
MemRefDescriptor viewMemRef(adaptor.in());
// No permutation, early exit.
@@ -3424,8 +3427,8 @@ class TransposeOpLowering : public ConvertOpToLLVMPattern<TransposeOp> {
/// 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<ViewOp> {
- using ConvertOpToLLVMPattern<ViewOp>::ConvertOpToLLVMPattern;
+struct ViewOpLowering : public ConvertOpToLLVMPattern<memref::ViewOp> {
+ using ConvertOpToLLVMPattern<memref::ViewOp>::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.
@@ -3461,10 +3464,10 @@ struct ViewOpLowering : public ConvertOpToLLVMPattern<ViewOp> {
}
LogicalResult
- matchAndRewrite(ViewOp viewOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::ViewOp viewOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto loc = viewOp.getLoc();
- ViewOpAdaptor adaptor(operands);
+ memref::ViewOpAdaptor adaptor(operands);
auto viewMemRefType = viewOp.getType();
auto targetElementTy =
@@ -3540,13 +3543,14 @@ struct ViewOpLowering : public ConvertOpToLLVMPattern<ViewOp> {
};
struct AssumeAlignmentOpLowering
- : public ConvertOpToLLVMPattern<AssumeAlignmentOp> {
- using ConvertOpToLLVMPattern<AssumeAlignmentOp>::ConvertOpToLLVMPattern;
+ : public ConvertOpToLLVMPattern<memref::AssumeAlignmentOp> {
+ using ConvertOpToLLVMPattern<
+ memref::AssumeAlignmentOp>::ConvertOpToLLVMPattern;
LogicalResult
- matchAndRewrite(AssumeAlignmentOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::AssumeAlignmentOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- AssumeAlignmentOp::Adaptor transformed(operands);
+ memref::AssumeAlignmentOp::Adaptor transformed(operands);
Value memref = transformed.memref();
unsigned alignment = op.alignment();
auto loc = op.getLoc();
diff --git a/mlir/lib/Conversion/StandardToSPIRV/CMakeLists.txt b/mlir/lib/Conversion/StandardToSPIRV/CMakeLists.txt
index 1d823bfce5c6..094934dd53ad 100644
--- a/mlir/lib/Conversion/StandardToSPIRV/CMakeLists.txt
+++ b/mlir/lib/Conversion/StandardToSPIRV/CMakeLists.txt
@@ -13,6 +13,7 @@ add_mlir_conversion_library(MLIRStandardToSPIRV
LINK_LIBS PUBLIC
MLIRIR
MLIRMath
+ MLIRMemRef
MLIRPass
MLIRSPIRV
MLIRSPIRVConversion
diff --git a/mlir/lib/Conversion/StandardToSPIRV/LegalizeStandardForSPIRV.cpp b/mlir/lib/Conversion/StandardToSPIRV/LegalizeStandardForSPIRV.cpp
index 4c4705da6a48..00bf6c0f157f 100644
--- a/mlir/lib/Conversion/StandardToSPIRV/LegalizeStandardForSPIRV.cpp
+++ b/mlir/lib/Conversion/StandardToSPIRV/LegalizeStandardForSPIRV.cpp
@@ -14,6 +14,7 @@
#include "../PassDetail.h"
#include "mlir/Conversion/StandardToSPIRV/StandardToSPIRV.h"
#include "mlir/Conversion/StandardToSPIRV/StandardToSPIRVPass.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Vector/VectorOps.h"
@@ -23,11 +24,11 @@
using namespace mlir;
/// Helpers to access the memref operand for each op.
-static Value getMemRefOperand(LoadOp op) { return op.memref(); }
+static Value getMemRefOperand(memref::LoadOp op) { return op.memref(); }
static Value getMemRefOperand(vector::TransferReadOp op) { return op.source(); }
-static Value getMemRefOperand(StoreOp op) { return op.memref(); }
+static Value getMemRefOperand(memref::StoreOp op) { return op.memref(); }
static Value getMemRefOperand(vector::TransferWriteOp op) {
return op.source();
@@ -44,7 +45,7 @@ class LoadOpOfSubViewFolder final : public OpRewritePattern<OpTy> {
PatternRewriter &rewriter) const override;
private:
- void replaceOp(OpTy loadOp, SubViewOp subViewOp,
+ void replaceOp(OpTy loadOp, memref::SubViewOp subViewOp,
ArrayRef<Value> sourceIndices,
PatternRewriter &rewriter) const;
};
@@ -59,23 +60,22 @@ class StoreOpOfSubViewFolder final : public OpRewritePattern<OpTy> {
PatternRewriter &rewriter) const override;
private:
- void replaceOp(OpTy StoreOp, SubViewOp subViewOp,
+ void replaceOp(OpTy storeOp, memref::SubViewOp subViewOp,
ArrayRef<Value> sourceIndices,
PatternRewriter &rewriter) const;
};
template <>
-void LoadOpOfSubViewFolder<LoadOp>::replaceOp(LoadOp loadOp,
- SubViewOp subViewOp,
- ArrayRef<Value> sourceIndices,
- PatternRewriter &rewriter) const {
- rewriter.replaceOpWithNewOp<LoadOp>(loadOp, subViewOp.source(),
- sourceIndices);
+void LoadOpOfSubViewFolder<memref::LoadOp>::replaceOp(
+ memref::LoadOp loadOp, memref::SubViewOp subViewOp,
+ ArrayRef<Value> sourceIndices, PatternRewriter &rewriter) const {
+ rewriter.replaceOpWithNewOp<memref::LoadOp>(loadOp, subViewOp.source(),
+ sourceIndices);
}
template <>
void LoadOpOfSubViewFolder<vector::TransferReadOp>::replaceOp(
- vector::TransferReadOp loadOp, SubViewOp subViewOp,
+ vector::TransferReadOp loadOp, memref::SubViewOp subViewOp,
ArrayRef<Value> sourceIndices, PatternRewriter &rewriter) const {
rewriter.replaceOpWithNewOp<vector::TransferReadOp>(
loadOp, loadOp.getVectorType(), subViewOp.source(), sourceIndices,
@@ -83,16 +83,16 @@ void LoadOpOfSubViewFolder<vector::TransferReadOp>::replaceOp(
}
template <>
-void StoreOpOfSubViewFolder<StoreOp>::replaceOp(
- StoreOp storeOp, SubViewOp subViewOp, ArrayRef<Value> sourceIndices,
- PatternRewriter &rewriter) const {
- rewriter.replaceOpWithNewOp<StoreOp>(storeOp, storeOp.value(),
- subViewOp.source(), sourceIndices);
+void StoreOpOfSubViewFolder<memref::StoreOp>::replaceOp(
+ memref::StoreOp storeOp, memref::SubViewOp subViewOp,
+ ArrayRef<Value> sourceIndices, PatternRewriter &rewriter) const {
+ rewriter.replaceOpWithNewOp<memref::StoreOp>(
+ storeOp, storeOp.value(), subViewOp.source(), sourceIndices);
}
template <>
void StoreOpOfSubViewFolder<vector::TransferWriteOp>::replaceOp(
- vector::TransferWriteOp tranferWriteOp, SubViewOp subViewOp,
+ vector::TransferWriteOp tranferWriteOp, memref::SubViewOp subViewOp,
ArrayRef<Value> sourceIndices, PatternRewriter &rewriter) const {
rewriter.replaceOpWithNewOp<vector::TransferWriteOp>(
tranferWriteOp, tranferWriteOp.vector(), subViewOp.source(),
@@ -120,7 +120,7 @@ void StoreOpOfSubViewFolder<vector::TransferWriteOp>::replaceOp(
/// memref<12x42xf32>
static LogicalResult
resolveSourceIndices(Location loc, PatternRewriter &rewriter,
- SubViewOp subViewOp, ValueRange indices,
+ memref::SubViewOp subViewOp, ValueRange indices,
SmallVectorImpl<Value> &sourceIndices) {
// TODO: Aborting when the offsets are static. There might be a way to fold
// the subview op with load even if the offsets have been canonicalized
@@ -152,7 +152,8 @@ template <typename OpTy>
LogicalResult
LoadOpOfSubViewFolder<OpTy>::matchAndRewrite(OpTy loadOp,
PatternRewriter &rewriter) const {
- auto subViewOp = getMemRefOperand(loadOp).template getDefiningOp<SubViewOp>();
+ auto subViewOp =
+ getMemRefOperand(loadOp).template getDefiningOp<memref::SubViewOp>();
if (!subViewOp) {
return failure();
}
@@ -174,7 +175,7 @@ LogicalResult
StoreOpOfSubViewFolder<OpTy>::matchAndRewrite(OpTy storeOp,
PatternRewriter &rewriter) const {
auto subViewOp =
- getMemRefOperand(storeOp).template getDefiningOp<SubViewOp>();
+ getMemRefOperand(storeOp).template getDefiningOp<memref::SubViewOp>();
if (!subViewOp) {
return failure();
}
@@ -193,9 +194,9 @@ StoreOpOfSubViewFolder<OpTy>::matchAndRewrite(OpTy storeOp,
void mlir::populateStdLegalizationPatternsForSPIRVLowering(
MLIRContext *context, OwningRewritePatternList &patterns) {
- patterns.insert<LoadOpOfSubViewFolder<LoadOp>,
+ patterns.insert<LoadOpOfSubViewFolder<memref::LoadOp>,
LoadOpOfSubViewFolder<vector::TransferReadOp>,
- StoreOpOfSubViewFolder<StoreOp>,
+ StoreOpOfSubViewFolder<memref::StoreOp>,
StoreOpOfSubViewFolder<vector::TransferWriteOp>>(context);
}
diff --git a/mlir/lib/Conversion/StandardToSPIRV/StandardToSPIRV.cpp b/mlir/lib/Conversion/StandardToSPIRV/StandardToSPIRV.cpp
index 96557dfa1ac7..ed1b72c60723 100644
--- a/mlir/lib/Conversion/StandardToSPIRV/StandardToSPIRV.cpp
+++ b/mlir/lib/Conversion/StandardToSPIRV/StandardToSPIRV.cpp
@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "mlir/Dialect/Math/IR/Math.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h"
#include "mlir/Dialect/SPIRV/IR/SPIRVOps.h"
#include "mlir/Dialect/SPIRV/Transforms/SPIRVConversion.h"
@@ -237,12 +238,12 @@ namespace {
/// to Workgroup memory when the size is constant. Note that this pattern needs
/// to be applied in a pass that runs at least at spv.module scope since it wil
/// ladd global variables into the spv.module.
-class AllocOpPattern final : public OpConversionPattern<AllocOp> {
+class AllocOpPattern final : public OpConversionPattern<memref::AllocOp> {
public:
- using OpConversionPattern<AllocOp>::OpConversionPattern;
+ using OpConversionPattern<memref::AllocOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(AllocOp operation, ArrayRef<Value> operands,
+ matchAndRewrite(memref::AllocOp operation, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
MemRefType allocType = operation.getType();
if (!isAllocationSupported(allocType))
@@ -278,12 +279,12 @@ class AllocOpPattern final : public OpConversionPattern<AllocOp> {
/// Removed a deallocation if it is a supported allocation. Currently only
/// removes deallocation if the memory space is workgroup memory.
-class DeallocOpPattern final : public OpConversionPattern<DeallocOp> {
+class DeallocOpPattern final : public OpConversionPattern<memref::DeallocOp> {
public:
- using OpConversionPattern<DeallocOp>::OpConversionPattern;
+ using OpConversionPattern<memref::DeallocOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(DeallocOp operation, ArrayRef<Value> operands,
+ matchAndRewrite(memref::DeallocOp operation, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
MemRefType deallocType = operation.memref().getType().cast<MemRefType>();
if (!isAllocationSupported(deallocType))
@@ -430,23 +431,23 @@ class CmpIOpPattern final : public OpConversionPattern<CmpIOp> {
ConversionPatternRewriter &rewriter) const override;
};
-/// Converts std.load to spv.Load.
-class IntLoadOpPattern final : public OpConversionPattern<LoadOp> {
+/// Converts memref.load to spv.Load.
+class IntLoadOpPattern final : public OpConversionPattern<memref::LoadOp> {
public:
- using OpConversionPattern<LoadOp>::OpConversionPattern;
+ using OpConversionPattern<memref::LoadOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::LoadOp loadOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override;
};
-/// Converts std.load to spv.Load.
-class LoadOpPattern final : public OpConversionPattern<LoadOp> {
+/// Converts memref.load to spv.Load.
+class LoadOpPattern final : public OpConversionPattern<memref::LoadOp> {
public:
- using OpConversionPattern<LoadOp>::OpConversionPattern;
+ using OpConversionPattern<memref::LoadOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::LoadOp loadOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override;
};
@@ -469,23 +470,23 @@ class SelectOpPattern final : public OpConversionPattern<SelectOp> {
ConversionPatternRewriter &rewriter) const override;
};
-/// Converts std.store to spv.Store on integers.
-class IntStoreOpPattern final : public OpConversionPattern<StoreOp> {
+/// Converts memref.store to spv.Store on integers.
+class IntStoreOpPattern final : public OpConversionPattern<memref::StoreOp> {
public:
- using OpConversionPattern<StoreOp>::OpConversionPattern;
+ using OpConversionPattern<memref::StoreOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(StoreOp storeOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::StoreOp storeOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override;
};
-/// Converts std.store to spv.Store.
-class StoreOpPattern final : public OpConversionPattern<StoreOp> {
+/// Converts memref.store to spv.Store.
+class StoreOpPattern final : public OpConversionPattern<memref::StoreOp> {
public:
- using OpConversionPattern<StoreOp>::OpConversionPattern;
+ using OpConversionPattern<memref::StoreOp>::OpConversionPattern;
LogicalResult
- matchAndRewrite(StoreOp storeOp, ArrayRef<Value> operands,
+ matchAndRewrite(memref::StoreOp storeOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override;
};
@@ -975,9 +976,10 @@ CmpIOpPattern::matchAndRewrite(CmpIOp cmpIOp, ArrayRef<Value> operands,
//===----------------------------------------------------------------------===//
LogicalResult
-IntLoadOpPattern::matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
+IntLoadOpPattern::matchAndRewrite(memref::LoadOp loadOp,
+ ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
- LoadOpAdaptor loadOperands(operands);
+ memref::LoadOpAdaptor loadOperands(operands);
auto loc = loadOp.getLoc();
auto memrefType = loadOp.memref().getType().cast<MemRefType>();
if (!memrefType.getElementType().isSignlessInteger())
@@ -1051,9 +1053,9 @@ IntLoadOpPattern::matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
}
LogicalResult
-LoadOpPattern::matchAndRewrite(LoadOp loadOp, ArrayRef<Value> operands,
+LoadOpPattern::matchAndRewrite(memref::LoadOp loadOp, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
- LoadOpAdaptor loadOperands(operands);
+ memref::LoadOpAdaptor loadOperands(operands);
auto memrefType = loadOp.memref().getType().cast<MemRefType>();
if (memrefType.getElementType().isSignlessInteger())
return failure();
@@ -1101,9 +1103,10 @@ SelectOpPattern::matchAndRewrite(SelectOp op, ArrayRef<Value> operands,
//===----------------------------------------------------------------------===//
LogicalResult
-IntStoreOpPattern::matchAndRewrite(StoreOp storeOp, ArrayRef<Value> operands,
+IntStoreOpPattern::matchAndRewrite(memref::StoreOp storeOp,
+ ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
- StoreOpAdaptor storeOperands(operands);
+ memref::StoreOpAdaptor storeOperands(operands);
auto memrefType = storeOp.memref().getType().cast<MemRefType>();
if (!memrefType.getElementType().isSignlessInteger())
return failure();
@@ -1180,9 +1183,10 @@ IntStoreOpPattern::matchAndRewrite(StoreOp storeOp, ArrayRef<Value> operands,
}
LogicalResult
-StoreOpPattern::matchAndRewrite(StoreOp storeOp, ArrayRef<Value> operands,
+StoreOpPattern::matchAndRewrite(memref::StoreOp storeOp,
+ ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const {
- StoreOpAdaptor storeOperands(operands);
+ memref::StoreOpAdaptor storeOperands(operands);
auto memrefType = storeOp.memref().getType().cast<MemRefType>();
if (memrefType.getElementType().isSignlessInteger())
return failure();
diff --git a/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt b/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt
index 806f4d497017..ace51f8b71d4 100644
--- a/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt
+++ b/mlir/lib/Conversion/VectorToLLVM/CMakeLists.txt
@@ -20,6 +20,7 @@ add_mlir_conversion_library(MLIRVectorToLLVM
MLIRArmSVEToLLVM
MLIRLLVMArmSVE
MLIRLLVMIR
+ MLIRMemRef
MLIRStandardToLLVM
MLIRTargetLLVMIRExport
MLIRTransforms
diff --git a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp
index ac0e3fc003d1..15553bbd9be5 100644
--- a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp
+++ b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVM.cpp
@@ -12,6 +12,7 @@
#include "mlir/Conversion/StandardToLLVM/ConvertStandardToLLVMPass.h"
#include "mlir/Dialect/LLVMIR/FunctionCallUtils.h"
#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Vector/VectorOps.h"
#include "mlir/IR/BuiltinTypes.h"
@@ -1262,7 +1263,7 @@ class VectorTransferConversion : public ConvertOpToLLVMPattern<ConcreteOp> {
unsigned vecWidth = LLVM::getVectorNumElements(vtp).getFixedValue();
unsigned lastIndex = llvm::size(xferOp.indices()) - 1;
Value off = xferOp.indices()[lastIndex];
- Value dim = rewriter.create<DimOp>(loc, xferOp.source(), lastIndex);
+ Value dim = rewriter.create<memref::DimOp>(loc, xferOp.source(), lastIndex);
Value mask = buildVectorComparison(
rewriter, xferOp, enableIndexOptimizations, vecWidth, dim, &off);
diff --git a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp
index 3be8cbfe770d..207a06e584c2 100644
--- a/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp
+++ b/mlir/lib/Conversion/VectorToLLVM/ConvertVectorToLLVMPass.cpp
@@ -19,6 +19,7 @@
#include "mlir/Dialect/ArmSVE/ArmSVEDialect.h"
#include "mlir/Dialect/LLVMIR/LLVMArmSVEDialect.h"
#include "mlir/Dialect/LLVMIR/LLVMDialect.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Vector/VectorOps.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
@@ -39,6 +40,7 @@ struct LowerVectorToLLVMPass
// Override explicitly to allow conditional dialect dependence.
void getDependentDialects(DialectRegistry ®istry) const override {
registry.insert<LLVM::LLVMDialect>();
+ registry.insert<memref::MemRefDialect>();
if (enableArmNeon)
registry.insert<arm_neon::ArmNeonDialect>();
if (enableArmSVE)
@@ -72,6 +74,7 @@ void LowerVectorToLLVMPass::runOnOperation() {
// Architecture specific augmentations.
LLVMConversionTarget target(getContext());
target.addLegalOp<LLVM::DialectCastOp>();
+ target.addLegalDialect<memref::MemRefDialect>();
target.addLegalDialect<StandardOpsDialect>();
target.addLegalOp<UnrealizedConversionCastOp>();
if (enableArmNeon) {
diff --git a/mlir/lib/Conversion/VectorToSCF/CMakeLists.txt b/mlir/lib/Conversion/VectorToSCF/CMakeLists.txt
index d6bc8dabf3a9..2a7ee5ea8a58 100644
--- a/mlir/lib/Conversion/VectorToSCF/CMakeLists.txt
+++ b/mlir/lib/Conversion/VectorToSCF/CMakeLists.txt
@@ -11,5 +11,6 @@ add_mlir_conversion_library(MLIRVectorToSCF
MLIREDSC
MLIRAffineEDSC
MLIRLLVMIR
+ MLIRMemRef
MLIRTransforms
)
diff --git a/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp b/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
index 724d0afdf67b..dce5b64284ec 100644
--- a/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
+++ b/mlir/lib/Conversion/VectorToSCF/VectorToSCF.cpp
@@ -16,6 +16,7 @@
#include "../PassDetail.h"
#include "mlir/Dialect/Affine/EDSC/Intrinsics.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/EDSC/Builders.h"
#include "mlir/Dialect/SCF/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
@@ -252,7 +253,7 @@ static Value setAllocAtFunctionEntry(MemRefType memRefMinorVectorType,
op->getParentWithTrait<OpTrait::AutomaticAllocationScope>();
assert(scope && "Expected op to be inside automatic allocation scope");
b.setInsertionPointToStart(&scope->getRegion(0).front());
- Value res = std_alloca(memRefMinorVectorType);
+ Value res = memref_alloca(memRefMinorVectorType);
return res;
}
@@ -314,7 +315,7 @@ LogicalResult NDTransferOpHelper<TransferReadOp>::doReplace() {
return {vector};
}
// 3.b. Otherwise, just go through the temporary `alloc`.
- std_store(vector, alloc, majorIvs);
+ memref_store(vector, alloc, majorIvs);
return {};
},
[&]() -> scf::ValueVector {
@@ -326,7 +327,7 @@ LogicalResult NDTransferOpHelper<TransferReadOp>::doReplace() {
return {vector};
}
// 3.d. Otherwise, just go through the temporary `alloc`.
- std_store(vector, alloc, majorIvs);
+ memref_store(vector, alloc, majorIvs);
return {};
});
@@ -341,14 +342,15 @@ LogicalResult NDTransferOpHelper<TransferReadOp>::doReplace() {
result = vector_insert(loaded1D, result, majorIvs);
// 5.b. Otherwise, just go through the temporary `alloc`.
else
- std_store(loaded1D, alloc, majorIvs);
+ memref_store(loaded1D, alloc, majorIvs);
}
});
assert((!options.unroll ^ (bool)result) &&
"Expected resulting Value iff unroll");
if (!result)
- result = std_load(vector_type_cast(MemRefType::get({}, vectorType), alloc));
+ result =
+ memref_load(vector_type_cast(MemRefType::get({}, vectorType), alloc));
rewriter.replaceOp(op, result);
return success();
@@ -359,8 +361,8 @@ LogicalResult NDTransferOpHelper<TransferWriteOp>::doReplace() {
Value alloc;
if (!options.unroll) {
alloc = setAllocAtFunctionEntry(memRefMinorVectorType, op);
- std_store(xferOp.vector(),
- vector_type_cast(MemRefType::get({}, vectorType), alloc));
+ memref_store(xferOp.vector(),
+ vector_type_cast(MemRefType::get({}, vectorType), alloc));
}
emitLoops([&](ValueRange majorIvs, ValueRange leadingOffsets,
@@ -379,7 +381,7 @@ LogicalResult NDTransferOpHelper<TransferWriteOp>::doReplace() {
if (options.unroll)
result = vector_extract(xferOp.vector(), majorIvs);
else
- result = std_load(alloc, majorIvs);
+ result = memref_load(alloc, majorIvs);
auto map =
getTransferMinorIdentityMap(xferOp.getShapedType(), minorVectorType);
ArrayAttr masked;
@@ -560,7 +562,7 @@ LogicalResult VectorTransferRewriter<TransferReadOp>::matchAndRewrite(
// Conservative lowering to scalar load / stores.
// 1. Setup all the captures.
ScopedContext scope(rewriter, transfer.getLoc());
- StdIndexedValue remote(transfer.source());
+ MemRefIndexedValue remote(transfer.source());
MemRefBoundsCapture memRefBoundsCapture(transfer.source());
VectorBoundsCapture vectorBoundsCapture(transfer.vector());
int coalescedIdx = computeCoalescedIndex(transfer);
@@ -579,7 +581,7 @@ LogicalResult VectorTransferRewriter<TransferReadOp>::matchAndRewrite(
// 2. Emit alloc-copy-load-dealloc.
MLIRContext *ctx = op->getContext();
Value tmp = setAllocAtFunctionEntry(tmpMemRefType(transfer), transfer);
- StdIndexedValue local(tmp);
+ MemRefIndexedValue local(tmp);
loopNestBuilder(lbs, ubs, steps, [&](ValueRange loopIvs) {
auto ivsStorage = llvm::to_vector<8>(loopIvs);
// Swap the ivs which will reorder memory accesses.
@@ -601,7 +603,7 @@ LogicalResult VectorTransferRewriter<TransferReadOp>::matchAndRewrite(
rewriter, cast<VectorTransferOpInterface>(transfer.getOperation()), ivs,
memRefBoundsCapture, loadValue, loadPadding);
});
- Value vectorValue = std_load(vector_type_cast(tmp));
+ Value vectorValue = memref_load(vector_type_cast(tmp));
// 3. Propagate.
rewriter.replaceOp(op, vectorValue);
@@ -646,7 +648,7 @@ LogicalResult VectorTransferRewriter<TransferWriteOp>::matchAndRewrite(
// 1. Setup all the captures.
ScopedContext scope(rewriter, transfer.getLoc());
- StdIndexedValue remote(transfer.source());
+ MemRefIndexedValue remote(transfer.source());
MemRefBoundsCapture memRefBoundsCapture(transfer.source());
Value vectorValue(transfer.vector());
VectorBoundsCapture vectorBoundsCapture(transfer.vector());
@@ -665,9 +667,9 @@ LogicalResult VectorTransferRewriter<TransferWriteOp>::matchAndRewrite(
// 2. Emit alloc-store-copy-dealloc.
Value tmp = setAllocAtFunctionEntry(tmpMemRefType(transfer), transfer);
- StdIndexedValue local(tmp);
+ MemRefIndexedValue local(tmp);
Value vec = vector_type_cast(tmp);
- std_store(vectorValue, vec);
+ memref_store(vectorValue, vec);
loopNestBuilder(lbs, ubs, steps, [&](ValueRange loopIvs) {
auto ivsStorage = llvm::to_vector<8>(loopIvs);
// Swap the ivsStorage which will reorder memory accesses.
diff --git a/mlir/lib/Dialect/Affine/IR/AffineOps.cpp b/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
index 7a97bb219ccd..7962ec21b5de 100644
--- a/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
+++ b/mlir/lib/Dialect/Affine/IR/AffineOps.cpp
@@ -8,6 +8,7 @@
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Affine/IR/AffineValueMap.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BlockAndValueMapping.h"
#include "mlir/IR/BuiltinOps.h"
@@ -64,7 +65,7 @@ remainsLegalAfterInline(Value value, Region *src, Region *dest,
// op won't be top-level anymore after inlining.
Attribute operandCst;
return matchPattern(value.getDefiningOp(), m_Constant(&operandCst)) ||
- value.getDefiningOp<DimOp>();
+ value.getDefiningOp<memref::DimOp>();
}
/// Checks if all values known to be legal affine dimensions or symbols in `src`
@@ -295,7 +296,7 @@ bool mlir::isValidDim(Value value, Region *region) {
return applyOp.isValidDim(region);
// The dim op is okay if its operand memref/tensor is defined at the top
// level.
- if (auto dimOp = dyn_cast<DimOp>(op))
+ if (auto dimOp = dyn_cast<memref::DimOp>(op))
return isTopLevelValue(dimOp.memrefOrTensor());
return false;
}
@@ -317,9 +318,8 @@ static bool isMemRefSizeValidSymbol(AnyMemRefDefOp memrefDefOp, unsigned index,
}
/// Returns true if the result of the dim op is a valid symbol for `region`.
-static bool isDimOpValidSymbol(DimOp dimOp, Region *region) {
- // The dim op is okay if its operand memref/tensor is defined at the top
- // level.
+static bool isDimOpValidSymbol(memref::DimOp dimOp, Region *region) {
+ // The dim op is okay if its operand memref is defined at the top level.
if (isTopLevelValue(dimOp.memrefOrTensor()))
return true;
@@ -328,14 +328,14 @@ static bool isDimOpValidSymbol(DimOp dimOp, Region *region) {
if (dimOp.memrefOrTensor().isa<BlockArgument>())
return false;
- // The dim op is also okay if its operand memref/tensor is a view/subview
- // whose corresponding size is a valid symbol.
+ // The dim op is also okay if its operand memref is a view/subview whose
+ // corresponding size is a valid symbol.
Optional<int64_t> index = dimOp.getConstantIndex();
assert(index.hasValue() &&
"expect only `dim` operations with a constant index");
int64_t i = index.getValue();
return TypeSwitch<Operation *, bool>(dimOp.memrefOrTensor().getDefiningOp())
- .Case<ViewOp, SubViewOp, AllocOp>(
+ .Case<memref::ViewOp, memref::SubViewOp, memref::AllocOp>(
[&](auto op) { return isMemRefSizeValidSymbol(op, i, region); })
.Default([](Operation *) { return false; });
}
@@ -404,7 +404,7 @@ bool mlir::isValidSymbol(Value value, Region *region) {
return applyOp.isValidSymbol(region);
// Dim op results could be valid symbols at any level.
- if (auto dimOp = dyn_cast<DimOp>(defOp))
+ if (auto dimOp = dyn_cast<memref::DimOp>(defOp))
return isDimOpValidSymbol(dimOp, region);
// Check for values dominating `region`'s parent op.
@@ -915,12 +915,12 @@ void AffineApplyOp::getCanonicalizationPatterns(
//===----------------------------------------------------------------------===//
/// This is a common class used for patterns of the form
-/// "someop(memrefcast) -> someop". It folds the source of any memref_cast
+/// "someop(memrefcast) -> someop". It folds the source of any memref.cast
/// into the root operation directly.
static LogicalResult foldMemRefCast(Operation *op) {
bool folded = false;
for (OpOperand &operand : op->getOpOperands()) {
- auto cast = operand.get().getDefiningOp<MemRefCastOp>();
+ auto cast = operand.get().getDefiningOp<memref::CastOp>();
if (cast && !cast.getOperand().getType().isa<UnrankedMemRefType>()) {
operand.set(cast.getOperand());
folded = true;
@@ -2254,7 +2254,8 @@ LogicalResult AffineStoreOp::fold(ArrayRef<Attribute> cstOperands,
// AffineMinMaxOpBase
//===----------------------------------------------------------------------===//
-template <typename T> static LogicalResult verifyAffineMinMaxOp(T op) {
+template <typename T>
+static LogicalResult verifyAffineMinMaxOp(T op) {
// Verify that operand count matches affine map dimension and symbol count.
if (op.getNumOperands() != op.map().getNumDims() + op.map().getNumSymbols())
return op.emitOpError(
@@ -2262,7 +2263,8 @@ template <typename T> static LogicalResult verifyAffineMinMaxOp(T op) {
return success();
}
-template <typename T> static void printAffineMinMaxOp(OpAsmPrinter &p, T op) {
+template <typename T>
+static void printAffineMinMaxOp(OpAsmPrinter &p, T op) {
p << op.getOperationName() << ' ' << op->getAttr(T::getMapAttrName());
auto operands = op.getOperands();
unsigned numDims = op.map().getNumDims();
diff --git a/mlir/lib/Dialect/Affine/IR/CMakeLists.txt b/mlir/lib/Dialect/Affine/IR/CMakeLists.txt
index 03153389a33f..f6da1863fb64 100644
--- a/mlir/lib/Dialect/Affine/IR/CMakeLists.txt
+++ b/mlir/lib/Dialect/Affine/IR/CMakeLists.txt
@@ -14,6 +14,7 @@ add_mlir_dialect_library(MLIRAffine
MLIREDSC
MLIRIR
MLIRLoopLikeInterface
+ MLIRMemRef
MLIRSideEffectInterfaces
MLIRStandard
)
diff --git a/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp b/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp
index c8cf3dd9475a..e3834ea62367 100644
--- a/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp
+++ b/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp
@@ -23,6 +23,7 @@
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Affine/Passes.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
#include "mlir/Transforms/LoopUtils.h"
diff --git a/mlir/lib/Dialect/Affine/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Affine/Transforms/CMakeLists.txt
index d866a8fadd46..433ffcd27098 100644
--- a/mlir/lib/Dialect/Affine/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Affine/Transforms/CMakeLists.txt
@@ -22,6 +22,7 @@ add_mlir_dialect_library(MLIRAffineTransforms
MLIRAffineUtils
MLIREDSC
MLIRIR
+ MLIRMemRef
MLIRPass
MLIRSideEffectInterfaces
MLIRStandard
diff --git a/mlir/lib/Dialect/Affine/Transforms/PassDetail.h b/mlir/lib/Dialect/Affine/Transforms/PassDetail.h
index da8f7ac3fc81..3ba6fa182686 100644
--- a/mlir/lib/Dialect/Affine/Transforms/PassDetail.h
+++ b/mlir/lib/Dialect/Affine/Transforms/PassDetail.h
@@ -19,6 +19,11 @@ void registerDialect(DialectRegistry ®istry);
namespace linalg {
class LinalgDialect;
} // end namespace linalg
+
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
namespace vector {
class VectorDialect;
} // end namespace vector
diff --git a/mlir/lib/Dialect/CMakeLists.txt b/mlir/lib/Dialect/CMakeLists.txt
index 894a5f594534..b21ef9308af2 100644
--- a/mlir/lib/Dialect/CMakeLists.txt
+++ b/mlir/lib/Dialect/CMakeLists.txt
@@ -9,6 +9,7 @@ add_subdirectory(GPU)
add_subdirectory(Linalg)
add_subdirectory(LLVMIR)
add_subdirectory(Math)
+add_subdirectory(MemRef)
add_subdirectory(OpenACC)
add_subdirectory(OpenMP)
add_subdirectory(PDL)
diff --git a/mlir/lib/Dialect/GPU/CMakeLists.txt b/mlir/lib/Dialect/GPU/CMakeLists.txt
index 4d26ba400f18..d7fbfe0b5b61 100644
--- a/mlir/lib/Dialect/GPU/CMakeLists.txt
+++ b/mlir/lib/Dialect/GPU/CMakeLists.txt
@@ -35,6 +35,7 @@ add_mlir_dialect_library(MLIRGPU
MLIRAsync
MLIREDSC
MLIRIR
+ MLIRMemRef
MLIRLLVMIR
MLIRLLVMToLLVMIRTranslation
MLIRSCF
diff --git a/mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp b/mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp
index 669c75524a51..8e9ec0b353d5 100644
--- a/mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp
+++ b/mlir/lib/Dialect/GPU/Transforms/AllReduceLowering.cpp
@@ -13,6 +13,7 @@
#include "mlir/Dialect/GPU/GPUDialect.h"
#include "mlir/Dialect/GPU/Passes.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BlockAndValueMapping.h"
#include "mlir/IR/Builders.h"
@@ -107,7 +108,7 @@ struct GpuAllReduceRewriter {
createPredicatedBlock(isFirstLane, [&] {
Value subgroupId = getDivideBySubgroupSize(invocationIdx);
Value index = create<IndexCastOp>(indexType, subgroupId);
- create<StoreOp>(subgroupReduce, buffer, index);
+ create<memref::StoreOp>(subgroupReduce, buffer, index);
});
create<gpu::BarrierOp>();
@@ -124,27 +125,29 @@ struct GpuAllReduceRewriter {
Value zero = create<ConstantIndexOp>(0);
createPredicatedBlock(isValidSubgroup, [&] {
Value index = create<IndexCastOp>(indexType, invocationIdx);
- Value value = create<LoadOp>(valueType, buffer, index);
+ Value value = create<memref::LoadOp>(valueType, buffer, index);
Value result =
createSubgroupReduce(numSubgroups, laneId, value, accumFactory);
- create<StoreOp>(result, buffer, zero);
+ create<memref::StoreOp>(result, buffer, zero);
});
// Synchronize workgroup and load result from workgroup memory.
create<gpu::BarrierOp>();
- Value result = create<LoadOp>(valueType, buffer, zero);
+ Value result = create<memref::LoadOp>(valueType, buffer, zero);
rewriter.replaceOp(reduceOp, result);
}
private:
// Shortcut to create an op from rewriter using loc as the first argument.
- template <typename T, typename... Args> T create(Args... args) {
+ template <typename T, typename... Args>
+ T create(Args... args) {
return rewriter.create<T>(loc, std::forward<Args>(args)...);
}
// Creates dimension op of type T, with the result casted to int32.
- template <typename T> Value getDimOp(StringRef dimension) {
+ template <typename T>
+ Value getDimOp(StringRef dimension) {
Value dim = create<T>(indexType, rewriter.getStringAttr(dimension));
return create<IndexCastOp>(int32Type, dim);
}
@@ -236,7 +239,8 @@ struct GpuAllReduceRewriter {
}
/// Returns an accumulator factory that creates an op of type T.
- template <typename T> AccumulatorFactory getFactory() {
+ template <typename T>
+ AccumulatorFactory getFactory() {
return [&](Value lhs, Value rhs) {
return create<T>(lhs.getType(), lhs, rhs);
};
diff --git a/mlir/lib/Dialect/GPU/Transforms/KernelOutlining.cpp b/mlir/lib/Dialect/GPU/Transforms/KernelOutlining.cpp
index cee931aefc7b..98b22a8a6d92 100644
--- a/mlir/lib/Dialect/GPU/Transforms/KernelOutlining.cpp
+++ b/mlir/lib/Dialect/GPU/Transforms/KernelOutlining.cpp
@@ -14,6 +14,7 @@
#include "mlir/Dialect/GPU/GPUDialect.h"
#include "mlir/Dialect/GPU/Passes.h"
#include "mlir/Dialect/GPU/Utils.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BlockAndValueMapping.h"
#include "mlir/IR/Builders.h"
@@ -58,7 +59,7 @@ static void injectGpuIndexOperations(Location loc, Region &launchFuncOpBody,
/// operations may not have side-effects, as otherwise sinking (and hence
/// duplicating them) is not legal.
static bool isSinkingBeneficiary(Operation *op) {
- return isa<ConstantOp, DimOp, SelectOp, CmpIOp>(op);
+ return isa<ConstantOp, memref::DimOp, SelectOp, CmpIOp>(op);
}
/// For a given operation `op`, computes whether it is beneficial to sink the
diff --git a/mlir/lib/Dialect/GPU/Transforms/MemoryPromotion.cpp b/mlir/lib/Dialect/GPU/Transforms/MemoryPromotion.cpp
index c47b87138e2d..cc50865d0c6b 100644
--- a/mlir/lib/Dialect/GPU/Transforms/MemoryPromotion.cpp
+++ b/mlir/lib/Dialect/GPU/Transforms/MemoryPromotion.cpp
@@ -13,6 +13,7 @@
#include "mlir/Dialect/GPU/MemoryPromotion.h"
#include "mlir/Dialect/GPU/GPUDialect.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/EDSC/Builders.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Pass/Pass.h"
@@ -82,7 +83,7 @@ static void insertCopyLoops(OpBuilder &builder, Location loc,
loopNestBuilder(lbs, ubs, steps, [&](ValueRange loopIvs) {
ivs.assign(loopIvs.begin(), loopIvs.end());
auto activeIvs = llvm::makeArrayRef(ivs).take_back(rank);
- StdIndexedValue fromHandle(from), toHandle(to);
+ MemRefIndexedValue fromHandle(from), toHandle(to);
toHandle(activeIvs) = fromHandle(activeIvs);
});
diff --git a/mlir/lib/Dialect/Linalg/Analysis/CMakeLists.txt b/mlir/lib/Dialect/Linalg/Analysis/CMakeLists.txt
index b7c7a67fef1b..14a161b13325 100644
--- a/mlir/lib/Dialect/Linalg/Analysis/CMakeLists.txt
+++ b/mlir/lib/Dialect/Linalg/Analysis/CMakeLists.txt
@@ -7,5 +7,6 @@ add_mlir_dialect_library(MLIRLinalgAnalysis
LINK_LIBS PUBLIC
MLIRIR
MLIRLinalg
+ MLIRMemRef
MLIRStandard
)
diff --git a/mlir/lib/Dialect/Linalg/Analysis/DependenceAnalysis.cpp b/mlir/lib/Dialect/Linalg/Analysis/DependenceAnalysis.cpp
index 59004867a333..4b92667b0cc0 100644
--- a/mlir/lib/Dialect/Linalg/Analysis/DependenceAnalysis.cpp
+++ b/mlir/lib/Dialect/Linalg/Analysis/DependenceAnalysis.cpp
@@ -12,6 +12,7 @@
#include "mlir/Dialect/Linalg/Analysis/DependenceAnalysis.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BuiltinOps.h"
@@ -48,7 +49,7 @@ Value Aliases::find(Value v) {
// the aliasing further.
if (isa<RegionBranchOpInterface>(defOp))
return v;
- if (isa<TensorToMemrefOp>(defOp))
+ if (isa<memref::BufferCastOp>(defOp))
return v;
if (auto memEffect = dyn_cast<MemoryEffectOpInterface>(defOp)) {
diff --git a/mlir/lib/Dialect/Linalg/EDSC/CMakeLists.txt b/mlir/lib/Dialect/Linalg/EDSC/CMakeLists.txt
index 6b79896fe10d..c93e7582a876 100644
--- a/mlir/lib/Dialect/Linalg/EDSC/CMakeLists.txt
+++ b/mlir/lib/Dialect/Linalg/EDSC/CMakeLists.txt
@@ -11,6 +11,7 @@ add_mlir_dialect_library(MLIRLinalgEDSC
MLIRAffineEDSC
MLIRLinalg
MLIRMath
+ MLIRMemRef
MLIRSCF
MLIRStandard
)
diff --git a/mlir/lib/Dialect/Linalg/IR/CMakeLists.txt b/mlir/lib/Dialect/Linalg/IR/CMakeLists.txt
index 613e8f496be9..ba0eb047a2c9 100644
--- a/mlir/lib/Dialect/Linalg/IR/CMakeLists.txt
+++ b/mlir/lib/Dialect/Linalg/IR/CMakeLists.txt
@@ -19,5 +19,6 @@ add_mlir_dialect_library(MLIRLinalg
MLIRSideEffectInterfaces
MLIRViewLikeInterface
MLIRStandard
+ MLIRMemRef
MLIRTensor
)
diff --git a/mlir/lib/Dialect/Linalg/IR/LinalgInterfaces.cpp b/mlir/lib/Dialect/Linalg/IR/LinalgInterfaces.cpp
index 1f511a4ffcc4..1ab9467967b0 100644
--- a/mlir/lib/Dialect/Linalg/IR/LinalgInterfaces.cpp
+++ b/mlir/lib/Dialect/Linalg/IR/LinalgInterfaces.cpp
@@ -9,6 +9,7 @@
#include "mlir/Dialect/Linalg/IR/LinalgInterfaces.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/AffineExprVisitor.h"
#include "mlir/IR/AffineMap.h"
#include "llvm/ADT/SmallSet.h"
@@ -187,7 +188,7 @@ SmallVector<Value, 4> LinalgOp::createFlatListOfOperandDims(OpBuilder &b,
for (Value v : getShapedOperands()) {
ShapedType t = v.getType().template cast<ShapedType>();
for (unsigned i = 0, e = t.getRank(); i < e; ++i)
- res.push_back(b.create<DimOp>(loc, v, i));
+ res.push_back(b.create<memref::DimOp>(loc, v, i));
}
return res;
}
diff --git a/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp b/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
index 1f686b940812..3b268d703a74 100644
--- a/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
+++ b/mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp
@@ -15,6 +15,7 @@
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Linalg/EDSC/Intrinsics.h"
#include "mlir/Dialect/Linalg/IR/LinalgTypes.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/AffineExprVisitor.h"
#include "mlir/IR/Matchers.h"
@@ -109,12 +110,12 @@ static void dispatchIndexOpFoldResult(OpFoldResult ofr,
/// ```
/// someop(memrefcast) -> someop
/// ```
-/// It folds the source of the memref_cast into the root operation directly.
+/// It folds the source of the memref.cast into the root operation directly.
static LogicalResult foldMemRefCast(Operation *op) {
bool folded = false;
for (OpOperand &operand : op->getOpOperands()) {
- auto castOp = operand.get().getDefiningOp<MemRefCastOp>();
- if (castOp && canFoldIntoConsumerOp(castOp)) {
+ auto castOp = operand.get().getDefiningOp<memref::CastOp>();
+ if (castOp && memref::CastOp::canFoldIntoConsumerOp(castOp)) {
operand.set(castOp.getOperand());
folded = true;
}
@@ -776,10 +777,10 @@ struct ReplaceStaticShapeDims : OpRewritePattern<InitTensorOp> {
/// - A constant value if the size is static along the dimension.
/// - The dynamic value that defines the size of the result of
/// `linalg.init_tensor` op.
-struct ReplaceDimOfInitTensorOp : public OpRewritePattern<DimOp> {
- using OpRewritePattern<DimOp>::OpRewritePattern;
+struct ReplaceDimOfInitTensorOp : public OpRewritePattern<memref::DimOp> {
+ using OpRewritePattern<memref::DimOp>::OpRewritePattern;
- LogicalResult matchAndRewrite(DimOp dimOp,
+ LogicalResult matchAndRewrite(memref::DimOp dimOp,
PatternRewriter &rewriter) const override {
auto initTensorOp = dimOp.memrefOrTensor().getDefiningOp<InitTensorOp>();
if (!initTensorOp)
@@ -986,7 +987,7 @@ PadTensorOp PadTensorOp::createPadHighOp(Type type, Value source, Value pad,
assert(rankedTensorType.hasStaticShape());
int rank = rankedTensorType.getRank();
for (int i = 0; i < rank; ++i) {
- auto dimOp = builder.createOrFold<DimOp>(loc, source, i);
+ auto dimOp = builder.createOrFold<memref::DimOp>(loc, source, i);
auto resultDimSize = builder.createOrFold<ConstantIndexOp>(
loc, rankedTensorType.getDimSize(i));
auto highValue = builder.createOrFold<SubIOp>(loc, resultDimSize, dimOp);
@@ -1292,7 +1293,7 @@ getCollapsedOutputDimFromInputShape(OpBuilder &builder, Location loc,
AffineExpr expr;
SmallVector<Value, 2> dynamicDims;
for (auto dim : llvm::seq(startPos, endPos + 1)) {
- dynamicDims.push_back(builder.create<DimOp>(loc, src, dim));
+ dynamicDims.push_back(builder.create<memref::DimOp>(loc, src, dim));
AffineExpr currExpr = builder.getAffineSymbolExpr(dim - startPos);
expr = (expr ? expr * currExpr : currExpr);
}
@@ -1361,7 +1362,7 @@ static Value getExpandedOutputDimFromInputShape(
"dimensions");
linearizedStaticDim *= d.value();
}
- Value sourceDim = builder.create<DimOp>(loc, src, sourceDimPos);
+ Value sourceDim = builder.create<memref::DimOp>(loc, src, sourceDimPos);
return applyMapToValues(
builder, loc,
AffineMap::get(
@@ -1637,9 +1638,9 @@ struct FoldReshapeWithConstant : OpRewritePattern<TensorReshapeOp> {
};
/// Canonicalize dim ops that use the output shape with dim of the input.
-struct ReplaceDimOfReshapeOpResult : OpRewritePattern<DimOp> {
- using OpRewritePattern<DimOp>::OpRewritePattern;
- LogicalResult matchAndRewrite(DimOp dimOp,
+struct ReplaceDimOfReshapeOpResult : OpRewritePattern<memref::DimOp> {
+ using OpRewritePattern<memref::DimOp>::OpRewritePattern;
+ LogicalResult matchAndRewrite(memref::DimOp dimOp,
PatternRewriter &rewriter) const override {
Value dimValue = dimOp.memrefOrTensor();
Optional<int64_t> dimIndex = dimOp.getConstantIndex();
@@ -2445,24 +2446,25 @@ struct FoldTensorCastOp : public RewritePattern {
}
};
-/// Replaces std.dim operations that use the result of a LinalgOp (on tensors)
-/// with std.dim operations that use one of the arguments. For example,
+/// Replaces memref.dim operations that use the result of a LinalgOp (on
+/// tensors) with memref.dim operations that use one of the arguments. For
+/// example,
///
/// %0 = linalg.matmul ins(%arg0, %arg1, ...)
-/// %1 = dim %0, %c0
+/// %1 = memref.dim %0, %c0
///
/// with
///
-/// %1 = dim %arg0, %c0
+/// %1 = memref.dim %arg0, %c0
///
/// where possible. With this the result of the `linalg.matmul` is not used in
/// dim operations. If the value produced is replaced with another value (say by
/// tiling `linalg.matmul`) will make the `linalg.matmul` truly dead instead of
/// used in a dim op that would prevent the DCE of this op.
-struct ReplaceDimOfLinalgOpResult : public OpRewritePattern<DimOp> {
- using OpRewritePattern<DimOp>::OpRewritePattern;
+struct ReplaceDimOfLinalgOpResult : public OpRewritePattern<memref::DimOp> {
+ using OpRewritePattern<memref::DimOp>::OpRewritePattern;
- LogicalResult matchAndRewrite(DimOp dimOp,
+ LogicalResult matchAndRewrite(memref::DimOp dimOp,
PatternRewriter &rewriter) const override {
Value dimValue = dimOp.memrefOrTensor();
Optional<int64_t> dimIndex = dimOp.getConstantIndex();
@@ -2479,7 +2481,7 @@ struct ReplaceDimOfLinalgOpResult : public OpRewritePattern<DimOp> {
if (!operandDimValue) {
// Its always possible to replace using the corresponding `outs`
// parameter.
- operandDimValue = rewriter.create<DimOp>(
+ operandDimValue = rewriter.create<memref::DimOp>(
dimOp.getLoc(), linalgOp.getOutput(resultIndex), *dimIndex);
}
rewriter.replaceOp(dimOp, *operandDimValue);
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
index 8e0d22d45c0e..32b2ee706d19 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
@@ -25,8 +25,8 @@ using namespace ::mlir::linalg;
static Value cloneMemref(Location loc, Value memref, OpBuilder &b) {
auto memrefType = memref.getType().cast<MemRefType>();
- auto alloc =
- b.create<AllocOp>(loc, memrefType, getDynOperands(loc, memref, b));
+ auto alloc = b.create<memref::AllocOp>(loc, memrefType,
+ getDynOperands(loc, memref, b));
b.create<linalg::CopyOp>(loc, memref, alloc);
return alloc;
}
@@ -60,17 +60,17 @@ allocateBuffersForResults(Location loc, LinalgOp linalgOp,
continue;
}
- if (auto alloc = resultTensor.getDefiningOp<AllocOp>()) {
+ if (auto alloc = resultTensor.getDefiningOp<memref::AllocOp>()) {
resultBuffers.push_back(resultTensor);
continue;
}
// Allocate buffers for statically-shaped results.
if (memrefType.hasStaticShape()) {
- resultBuffers.push_back(b.create<AllocOp>(loc, memrefType));
+ resultBuffers.push_back(b.create<memref::AllocOp>(loc, memrefType));
continue;
}
- resultBuffers.push_back(b.create<AllocOp>(
+ resultBuffers.push_back(b.create<memref::AllocOp>(
loc, memrefType, getDynOperands(loc, resultTensor, b)));
}
return success();
@@ -148,7 +148,7 @@ class BufferizeInitTensorOp : public OpConversionPattern<InitTensorOp> {
matchAndRewrite(InitTensorOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const final {
linalg::InitTensorOpAdaptor adaptor(operands, op->getAttrDictionary());
- rewriter.replaceOpWithNewOp<AllocOp>(
+ rewriter.replaceOpWithNewOp<memref::AllocOp>(
op, getTypeConverter()->convertType(op.getType()).cast<MemRefType>(),
adaptor.sizes());
return success();
@@ -231,9 +231,9 @@ class SubTensorOpConverter : public OpConversionPattern<SubTensorOp> {
// op.sizes() capture exactly the dynamic alloc operands matching the
// subviewMemRefType thanks to subview/subtensor canonicalization and
// verification.
- Value alloc =
- rewriter.create<AllocOp>(op.getLoc(), subviewMemRefType, op.sizes());
- Value subView = rewriter.create<SubViewOp>(
+ Value alloc = rewriter.create<memref::AllocOp>(
+ op.getLoc(), subviewMemRefType, op.sizes());
+ Value subView = rewriter.create<memref::SubViewOp>(
op.getLoc(), sourceMemref, op.getMixedOffsets(), op.getMixedSizes(),
op.getMixedStrides());
rewriter.create<linalg::CopyOp>(op.getLoc(), subView, alloc);
@@ -243,8 +243,8 @@ class SubTensorOpConverter : public OpConversionPattern<SubTensorOp> {
};
/// Convert `subtensor_insert %source into %dest [offsets][sizes][strides] ->
-/// %t` to an tensor_to_memref + subview + copy + tensor_load pattern.
-/// tensor_to_memref and tensor_load are inserted automatically by the
+/// %t` to an buffer_cast + subview + copy + tensor_load pattern.
+/// buffer_cast and tensor_load are inserted automatically by the
/// conversion infra:
/// ```
/// %sv = subview %dest [offsets][sizes][strides]
@@ -273,7 +273,7 @@ class SubTensorInsertOpConverter
assert(destMemRef.getType().isa<MemRefType>());
// Take a subview to copy the small memref.
- Value subview = rewriter.create<SubViewOp>(
+ Value subview = rewriter.create<memref::SubViewOp>(
op.getLoc(), destMemRef, op.getMixedOffsets(), op.getMixedSizes(),
op.getMixedStrides());
// Copy the small memref.
@@ -295,7 +295,7 @@ struct LinalgBufferizePass : public LinalgBufferizeBase<LinalgBufferizePass> {
// Mark all Standard operations legal.
target.addLegalDialect<AffineDialect, math::MathDialect,
- StandardOpsDialect>();
+ memref::MemRefDialect, StandardOpsDialect>();
target.addIllegalOp<InitTensorOp, SubTensorOp, SubTensorInsertOp>();
// Mark all Linalg operations illegal as long as they work on tensors.
diff --git a/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
index 1469371e1466..5bca63e459a4 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
@@ -29,6 +29,7 @@ add_mlir_dialect_library(MLIRLinalgTransforms
MLIRAnalysis
MLIREDSC
MLIRIR
+ MLIRMemRef
MLIRLinalgAnalysis
MLIRLinalgEDSC
MLIRLinalg
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Fusion.cpp b/mlir/lib/Dialect/Linalg/Transforms/Fusion.cpp
index 7e52daef6588..ffa2811c37d2 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Fusion.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Fusion.cpp
@@ -18,6 +18,8 @@
#include "mlir/Dialect/Linalg/Passes.h"
#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
#include "mlir/Dialect/Linalg/Utils/Utils.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
#include "mlir/IR/AffineExpr.h"
@@ -104,11 +106,12 @@ static LinalgOp cloneWithLoopRanges(OpBuilder &b, Location loc, LinalgOp op,
SmallVector<OpFoldResult, 4> offsets, sizes, strides;
inferShapeComponents(map, loopRanges, offsets, sizes, strides);
Value shape = en.value();
- Value sub = shape.getType().isa<MemRefType>()
- ? b.create<SubViewOp>(loc, shape, offsets, sizes, strides)
- .getResult()
- : b.create<SubTensorOp>(loc, shape, offsets, sizes, strides)
- .getResult();
+ Value sub =
+ shape.getType().isa<MemRefType>()
+ ? b.create<memref::SubViewOp>(loc, shape, offsets, sizes, strides)
+ .getResult()
+ : b.create<SubTensorOp>(loc, shape, offsets, sizes, strides)
+ .getResult();
clonedShapes.push_back(sub);
}
// Append the other operands.
@@ -177,8 +180,8 @@ getShapeDefiningLoopRange(LinalgOp op, unsigned loopDepth,
// `ViewInterface`. The interface needs a `getOrCreateRanges` method which
// currently returns a `linalg.range`. The fix here is to move this op to
// `std` dialect and add the method to `ViewInterface`.
- if (fromSubViewOpOnly &&
- !isa_and_nonnull<SubViewOp, SubTensorOp>(en.value().getDefiningOp()))
+ if (fromSubViewOpOnly && !isa_and_nonnull<memref::SubViewOp, SubTensorOp>(
+ en.value().getDefiningOp()))
continue;
unsigned idx = en.index();
@@ -227,9 +230,8 @@ static LinalgOp fuse(OpBuilder &b, LinalgOp producer,
<< "existing LoopRange: " << loopRanges[i] << "\n");
else {
auto shapeDim = getShapeDefiningLoopRange(producer, i);
- loopRanges[i] = Range{std_constant_index(0),
- std_dim(shapeDim.shape, shapeDim.dimension),
- std_constant_index(1)};
+ Value dim = memref_dim(shapeDim.shape, shapeDim.dimension);
+ loopRanges[i] = Range{std_constant_index(0), dim, std_constant_index(1)};
LLVM_DEBUG(llvm::dbgs() << "new LoopRange: " << loopRanges[i] << "\n");
}
}
@@ -242,7 +244,7 @@ static LinalgOp fuse(OpBuilder &b, LinalgOp producer,
static Range getRangeFromOperandShape(OpBuilder &b, Location loc,
Value shapedOperand, unsigned dim) {
Operation *shapeProducingOp = shapedOperand.getDefiningOp();
- if (auto subViewOp = dyn_cast<SubViewOp>(shapeProducingOp))
+ if (auto subViewOp = dyn_cast<memref::SubViewOp>(shapeProducingOp))
return subViewOp.getOrCreateRanges(b, loc)[dim];
if (auto subTensorOp = dyn_cast<SubTensorOp>(shapeProducingOp))
return subTensorOp.getOrCreateRanges(b, loc)[dim];
@@ -425,7 +427,7 @@ mlir::linalg::fuseProducerOfBuffer(OpBuilder &b, OpOperand &consumerOpOperand,
// Must be a subview or a slice to guarantee there are loops we can fuse
// into.
- auto subView = consumerOpOperand.get().getDefiningOp<SubViewOp>();
+ auto subView = consumerOpOperand.get().getDefiningOp<memref::SubViewOp>();
if (!subView) {
LLVM_DEBUG(llvm::dbgs() << "\nNot fusable (not a subview)");
return llvm::None;
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Loops.cpp b/mlir/lib/Dialect/Linalg/Transforms/Loops.cpp
index d09d3e0b5edd..d6423f4badd4 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Loops.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Loops.cpp
@@ -14,6 +14,7 @@
#include "mlir/Dialect/Linalg/Passes.h"
#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
#include "mlir/Dialect/Linalg/Utils/Utils.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/EDSC/Builders.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/IR/AffineExpr.h"
@@ -200,7 +201,7 @@ Value getPaddedInput(Value input, ArrayRef<Value> indices,
conds.push_back(leftOutOfBound);
else
conds.push_back(conds.back() || leftOutOfBound);
- Value rightBound = std_dim(input, idx);
+ Value rightBound = memref_dim(input, idx);
conds.push_back(conds.back() || (sge(dim, rightBound)));
// When padding is involved, the indices will only be shifted to negative,
@@ -307,12 +308,12 @@ static void emitScalarImplementation(ArrayRef<Value> allIvs, ConvOp convOp) {
IndexedValueType F(convOp.filter()), O(convOp.output());
// Emit scalar form. Padded conv involves an affine.max in the memory access
- // which is not allowed by affine.load. Override to use an StdIndexedValue
+ // which is not allowed by affine.load. Override to use an MemRefIndexedValue
// when there is non-zero padding.
if (hasPadding(convOp)) {
Type type = convOp.input().getType().cast<MemRefType>().getElementType();
Value padValue = std_constant(type, getPadValueAttr<ConvOp>(type));
- Value paddedInput = getPaddedInput<StdIndexedValue>(
+ Value paddedInput = getPaddedInput<MemRefIndexedValue>(
convOp.input(), imIdx,
/* Only need to pad the window dimensions */
{0, static_cast<int>(imIdx.size()) - 1}, padValue);
@@ -338,9 +339,9 @@ static Value getPoolingInput(PoolingOp op, ArrayRef<Value> inputIndices) {
Type type =
op.input().getType().template cast<MemRefType>().getElementType();
Value padValue = std_constant(type, getPadValueAttr<PoolingOp>(type));
- return getPaddedInput<StdIndexedValue>(op.input(), inputIndices,
- /*Pad every dimension*/ {},
- padValue);
+ return getPaddedInput<MemRefIndexedValue>(op.input(), inputIndices,
+ /*Pad every dimension*/ {},
+ padValue);
}
IndexedValueType input(op.input());
return input(inputIndices);
@@ -546,7 +547,7 @@ static void lowerLinalgToLoopsImpl(FuncOp funcOp,
MLIRContext *context = funcOp.getContext();
OwningRewritePatternList patterns;
patterns.insert<LinalgRewritePattern<LoopType>>(interchangeVector);
- DimOp::getCanonicalizationPatterns(patterns, context);
+ memref::DimOp::getCanonicalizationPatterns(patterns, context);
AffineApplyOp::getCanonicalizationPatterns(patterns, context);
patterns.insert<FoldAffineOp>(context);
// Just apply the patterns greedily.
@@ -593,12 +594,18 @@ struct FoldAffineOp : public RewritePattern {
struct LowerToAffineLoops
: public LinalgLowerToAffineLoopsBase<LowerToAffineLoops> {
+ void getDependentDialects(DialectRegistry ®istry) const override {
+ registry.insert<memref::MemRefDialect>();
+ }
void runOnFunction() override {
lowerLinalgToLoopsImpl<AffineForOp>(getFunction(), interchangeVector);
}
};
struct LowerToLoops : public LinalgLowerToLoopsBase<LowerToLoops> {
+ void getDependentDialects(DialectRegistry ®istry) const override {
+ registry.insert<memref::MemRefDialect, scf::SCFDialect>();
+ }
void runOnFunction() override {
lowerLinalgToLoopsImpl<scf::ForOp>(getFunction(), interchangeVector);
}
diff --git a/mlir/lib/Dialect/Linalg/Transforms/PassDetail.h b/mlir/lib/Dialect/Linalg/Transforms/PassDetail.h
index dc23a3bd8599..283412d6e4da 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/PassDetail.h
+++ b/mlir/lib/Dialect/Linalg/Transforms/PassDetail.h
@@ -26,6 +26,10 @@ namespace scf {
class SCFDialect;
} // end namespace scf
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
namespace vector {
class VectorDialect;
} // end namespace vector
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Promotion.cpp b/mlir/lib/Dialect/Linalg/Transforms/Promotion.cpp
index f99bf85d3cbe..7c8aeb04e3d0 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Promotion.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Promotion.cpp
@@ -18,6 +18,7 @@
#include "mlir/Dialect/Linalg/Passes.h"
#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
#include "mlir/Dialect/Linalg/Utils/Utils.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/IR/AffineExpr.h"
@@ -38,9 +39,9 @@ using llvm::MapVector;
using folded_affine_min = FoldedValueBuilder<AffineMinOp>;
using folded_linalg_range = FoldedValueBuilder<linalg::RangeOp>;
-using folded_std_dim = FoldedValueBuilder<DimOp>;
-using folded_std_subview = FoldedValueBuilder<SubViewOp>;
-using folded_std_view = FoldedValueBuilder<ViewOp>;
+using folded_memref_dim = FoldedValueBuilder<memref::DimOp>;
+using folded_memref_subview = FoldedValueBuilder<memref::SubViewOp>;
+using folded_memref_view = FoldedValueBuilder<memref::ViewOp>;
#define DEBUG_TYPE "linalg-promotion"
@@ -59,22 +60,22 @@ static Value allocBuffer(const LinalgPromotionOptions &options,
if (!dynamicBuffers)
if (auto cst = size.getDefiningOp<ConstantIndexOp>())
return options.useAlloca
- ? std_alloca(MemRefType::get(width * cst.getValue(),
- IntegerType::get(ctx, 8)),
- ValueRange{}, alignment_attr)
+ ? memref_alloca(MemRefType::get(width * cst.getValue(),
+ IntegerType::get(ctx, 8)),
+ ValueRange{}, alignment_attr)
.value
- : std_alloc(MemRefType::get(width * cst.getValue(),
- IntegerType::get(ctx, 8)),
- ValueRange{}, alignment_attr)
+ : memref_alloc(MemRefType::get(width * cst.getValue(),
+ IntegerType::get(ctx, 8)),
+ ValueRange{}, alignment_attr)
.value;
Value mul =
folded_std_muli(folder, folded_std_constant_index(folder, width), size);
return options.useAlloca
- ? std_alloca(MemRefType::get(-1, IntegerType::get(ctx, 8)), mul,
- alignment_attr)
+ ? memref_alloca(MemRefType::get(-1, IntegerType::get(ctx, 8)), mul,
+ alignment_attr)
.value
- : std_alloc(MemRefType::get(-1, IntegerType::get(ctx, 8)), mul,
- alignment_attr)
+ : memref_alloc(MemRefType::get(-1, IntegerType::get(ctx, 8)), mul,
+ alignment_attr)
.value;
}
@@ -82,10 +83,12 @@ static Value allocBuffer(const LinalgPromotionOptions &options,
/// no call back to do so is provided. The default is to allocate a
/// memref<..xi8> and return a view to get a memref type of shape
/// boundingSubViewSize.
-static Optional<Value> defaultAllocBufferCallBack(
- const LinalgPromotionOptions &options, OpBuilder &builder,
- SubViewOp subView, ArrayRef<Value> boundingSubViewSize, bool dynamicBuffers,
- Optional<unsigned> alignment, OperationFolder *folder) {
+static Optional<Value>
+defaultAllocBufferCallBack(const LinalgPromotionOptions &options,
+ OpBuilder &builder, memref::SubViewOp subView,
+ ArrayRef<Value> boundingSubViewSize,
+ bool dynamicBuffers, Optional<unsigned> alignment,
+ OperationFolder *folder) {
ShapedType viewType = subView.getType();
int64_t rank = viewType.getRank();
(void)rank;
@@ -100,7 +103,7 @@ static Optional<Value> defaultAllocBufferCallBack(
dynamicBuffers, folder, alignment);
SmallVector<int64_t, 4> dynSizes(boundingSubViewSize.size(),
ShapedType::kDynamicSize);
- Value view = folded_std_view(
+ Value view = folded_memref_view(
folder, MemRefType::get(dynSizes, viewType.getElementType()), buffer,
zero, boundingSubViewSize);
return view;
@@ -112,10 +115,10 @@ static Optional<Value> defaultAllocBufferCallBack(
static LogicalResult
defaultDeallocBufferCallBack(const LinalgPromotionOptions &options,
OpBuilder &b, Value fullLocalView) {
- auto viewOp = fullLocalView.getDefiningOp<ViewOp>();
+ auto viewOp = fullLocalView.getDefiningOp<memref::ViewOp>();
assert(viewOp && "expected full local view to be a ViewOp");
if (!options.useAlloca)
- std_dealloc(viewOp.source());
+ memref_dealloc(viewOp.source());
return success();
}
@@ -161,21 +164,21 @@ LinalgOpInstancePromotionOptions::LinalgOpInstancePromotionOptions(
if (options.operandsToPromote && !options.operandsToPromote->count(idx))
continue;
auto *op = linalgOp.getShapedOperand(idx).getDefiningOp();
- if (auto sv = dyn_cast_or_null<SubViewOp>(op)) {
+ if (auto sv = dyn_cast_or_null<memref::SubViewOp>(op)) {
subViews[idx] = sv;
useFullTileBuffers[sv] = vUseFullTileBuffers[idx];
}
}
- allocationFn =
- (options.allocationFn ? *(options.allocationFn)
- : [&](OpBuilder &builder, SubViewOp subViewOp,
- ArrayRef<Value> boundingSubViewSize,
- OperationFolder *folder) -> Optional<Value> {
- return defaultAllocBufferCallBack(options, builder, subViewOp,
- boundingSubViewSize, dynamicBuffers,
- alignment, folder);
- });
+ allocationFn = (options.allocationFn
+ ? *(options.allocationFn)
+ : [&](OpBuilder &builder, memref::SubViewOp subViewOp,
+ ArrayRef<Value> boundingSubViewSize,
+ OperationFolder *folder) -> Optional<Value> {
+ return defaultAllocBufferCallBack(options, builder, subViewOp,
+ boundingSubViewSize, dynamicBuffers,
+ alignment, folder);
+ });
deallocationFn =
(options.deallocationFn
? *(options.deallocationFn)
@@ -209,7 +212,7 @@ LinalgOpInstancePromotionOptions::LinalgOpInstancePromotionOptions(
// boundary tiles. For now this is done with an unconditional `fill` op followed
// by a partial `copy` op.
Optional<PromotionInfo> mlir::linalg::promoteSubviewAsNewBuffer(
- OpBuilder &b, Location loc, SubViewOp subView,
+ OpBuilder &b, Location loc, memref::SubViewOp subView,
AllocBufferCallbackFn allocationFn, OperationFolder *folder) {
ScopedContext scopedContext(b, loc);
auto viewType = subView.getType();
@@ -227,7 +230,8 @@ Optional<PromotionInfo> mlir::linalg::promoteSubviewAsNewBuffer(
(!sizeAttr) ? rangeValue.size : b.create<ConstantOp>(loc, sizeAttr);
LLVM_DEBUG(llvm::dbgs() << "Extracted tightest: " << size << "\n");
fullSizes.push_back(size);
- partialSizes.push_back(folded_std_dim(folder, subView, en.index()).value);
+ partialSizes.push_back(
+ folded_memref_dim(folder, subView, en.index()).value);
}
SmallVector<int64_t, 4> dynSizes(fullSizes.size(), -1);
// If a callback is not specified, then use the default implementation for
@@ -238,7 +242,7 @@ Optional<PromotionInfo> mlir::linalg::promoteSubviewAsNewBuffer(
SmallVector<OpFoldResult, 4> zeros(fullSizes.size(), b.getIndexAttr(0));
SmallVector<OpFoldResult, 4> ones(fullSizes.size(), b.getIndexAttr(1));
auto partialLocalView =
- folded_std_subview(folder, *fullLocalView, zeros, partialSizes, ones);
+ folded_memref_subview(folder, *fullLocalView, zeros, partialSizes, ones);
return PromotionInfo{*fullLocalView, partialLocalView};
}
@@ -253,7 +257,8 @@ promoteSubViews(OpBuilder &b, Location loc,
MapVector<unsigned, PromotionInfo> promotionInfoMap;
for (auto v : options.subViews) {
- SubViewOp subView = cast<SubViewOp>(v.second.getDefiningOp());
+ memref::SubViewOp subView =
+ cast<memref::SubViewOp>(v.second.getDefiningOp());
Optional<PromotionInfo> promotionInfo = promoteSubviewAsNewBuffer(
b, loc, subView, options.allocationFn, folder);
if (!promotionInfo)
@@ -277,8 +282,9 @@ promoteSubViews(OpBuilder &b, Location loc,
auto info = promotionInfoMap.find(v.first);
if (info == promotionInfoMap.end())
continue;
- if (failed(options.copyInFn(b, cast<SubViewOp>(v.second.getDefiningOp()),
- info->second.partialLocalView)))
+ if (failed(options.copyInFn(
+ b, cast<memref::SubViewOp>(v.second.getDefiningOp()),
+ info->second.partialLocalView)))
return {};
}
return promotionInfoMap;
@@ -353,7 +359,7 @@ mlir::linalg::promoteSubviewsPrecondition(Operation *op,
return failure();
// Check that at least one of the requested operands is indeed a subview.
for (auto en : llvm::enumerate(linOp.getShapedOperands())) {
- auto sv = isa_and_nonnull<SubViewOp>(en.value().getDefiningOp());
+ auto sv = isa_and_nonnull<memref::SubViewOp>(en.value().getDefiningOp());
if (sv) {
if (!options.operandsToPromote.hasValue() ||
options.operandsToPromote->count(en.index()))
diff --git a/mlir/lib/Dialect/Linalg/Transforms/SparseLowering.cpp b/mlir/lib/Dialect/Linalg/Transforms/SparseLowering.cpp
index 9eb420cde555..d9c258083a52 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/SparseLowering.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/SparseLowering.cpp
@@ -44,11 +44,11 @@ class TensorFromPointerConverter
};
/// Sparse conversion rule for dimension accesses.
-class TensorToDimSizeConverter : public OpConversionPattern<DimOp> {
+class TensorToDimSizeConverter : public OpConversionPattern<memref::DimOp> {
public:
using OpConversionPattern::OpConversionPattern;
LogicalResult
- matchAndRewrite(DimOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::DimOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
if (!operands[0].getType().isa<LLVM::LLVMPointerType>())
return failure();
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp b/mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp
index ed3b8b31f97d..a940bd697f53 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Sparsification.cpp
@@ -533,13 +533,13 @@ static Value genOutputBuffer(CodeGen &codegen, PatternRewriter &rewriter,
// positions for the output tensor. Currently this results in functional,
// but slightly imprecise IR, so it is put under an experimental option.
if (codegen.options.fastOutput)
- return rewriter.create<TensorToMemrefOp>(loc, denseTp, tensor);
+ return rewriter.create<memref::BufferCastOp>(loc, denseTp, tensor);
// By default, a new buffer is allocated which is initialized to the
// tensor defined in the outs() clause. This is always correct but
// introduces a dense initialization component that may negatively
// impact the running complexity of the sparse kernel.
- Value init = rewriter.create<TensorToMemrefOp>(loc, denseTp, tensor);
- Value alloc = rewriter.create<AllocOp>(loc, denseTp, args);
+ Value init = rewriter.create<memref::BufferCastOp>(loc, denseTp, tensor);
+ Value alloc = rewriter.create<memref::AllocOp>(loc, denseTp, args);
rewriter.create<linalg::CopyOp>(loc, init, alloc);
return alloc;
}
@@ -585,8 +585,8 @@ static void genBuffers(Merger &merger, CodeGen &codegen,
}
// Find lower and upper bound in current dimension.
Value up;
- if (shape[d] == TensorType::kDynamicSize) {
- up = rewriter.create<DimOp>(loc, tensor, d);
+ if (shape[d] == MemRefType::kDynamicSize) {
+ up = rewriter.create<memref::DimOp>(loc, tensor, d);
args.push_back(up);
} else {
up = rewriter.create<ConstantIndexOp>(loc, shape[d]);
@@ -600,7 +600,7 @@ static void genBuffers(Merger &merger, CodeGen &codegen,
auto denseTp = MemRefType::get(shape, tensorType.getElementType());
if (t < numInputs)
codegen.buffers[t] =
- rewriter.create<TensorToMemrefOp>(loc, denseTp, tensor);
+ rewriter.create<memref::BufferCastOp>(loc, denseTp, tensor);
else
codegen.buffers[t] =
genOutputBuffer(codegen, rewriter, op, denseTp, args);
@@ -716,7 +716,7 @@ static Value genTensorLoad(Merger &merger, CodeGen &codegen,
Value ptr = codegen.buffers[tensor];
if (codegen.curVecLength > 1)
return genVectorLoad(codegen, rewriter, ptr, args);
- return rewriter.create<LoadOp>(loc, ptr, args);
+ return rewriter.create<memref::LoadOp>(loc, ptr, args);
}
/// Generates a store on a dense tensor.
@@ -744,7 +744,7 @@ static void genTensorStore(Merger &merger, CodeGen &codegen,
if (codegen.curVecLength > 1)
genVectorStore(codegen, rewriter, rhs, ptr, args);
else
- rewriter.create<StoreOp>(loc, rhs, ptr, args);
+ rewriter.create<memref::StoreOp>(loc, rhs, ptr, args);
}
/// Generates a pointer/index load from the sparse storage scheme.
@@ -752,7 +752,7 @@ static Value genLoad(CodeGen &codegen, PatternRewriter &rewriter, Location loc,
Value ptr, Value s) {
if (codegen.curVecLength > 1)
return genVectorLoad(codegen, rewriter, ptr, {s});
- Value load = rewriter.create<LoadOp>(loc, ptr, s);
+ Value load = rewriter.create<memref::LoadOp>(loc, ptr, s);
return load.getType().isa<IndexType>()
? load
: rewriter.create<IndexCastOp>(loc, load, rewriter.getIndexType());
@@ -1345,8 +1345,8 @@ struct GenericOpSparsifier : public OpRewritePattern<linalg::GenericOp> {
CodeGen codegen(options, numTensors, numLoops);
genBuffers(merger, codegen, rewriter, op);
genStmt(merger, codegen, rewriter, op, topSort, exp.getValue(), 0);
- Value result =
- rewriter.create<TensorLoadOp>(op.getLoc(), codegen.buffers.back());
+ Value result = rewriter.create<memref::TensorLoadOp>(
+ op.getLoc(), codegen.buffers.back());
rewriter.replaceOp(op, result);
return success();
}
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Tiling.cpp b/mlir/lib/Dialect/Linalg/Transforms/Tiling.cpp
index 938cb86ee8c4..d638c609aa9e 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Tiling.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Tiling.cpp
@@ -17,6 +17,8 @@
#include "mlir/Dialect/Linalg/Passes.h"
#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
#include "mlir/Dialect/Linalg/Utils/Utils.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/EDSC/Builders.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
@@ -34,7 +36,6 @@ using namespace mlir::edsc::intrinsics;
using namespace mlir::linalg;
using namespace mlir::scf;
-
#define DEBUG_TYPE "linalg-tiling"
static bool isZero(Value v) {
@@ -144,9 +145,9 @@ struct TileCheck : public AffineExprVisitor<TileCheck> {
// operand_dim_1 = dim %operand, 1 : memref<50x100xf32>
// scf.for %k = %c0 to operand_dim_0 step %c10 {
// scf.for %l = %c0 to operand_dim_1 step %c25 {
-// %4 = std.subview %operand[%k, %l][%c10, %c25][%c1, %c1]
+// %4 = memref.subview %operand[%k, %l][%c10, %c25][%c1, %c1]
// : memref<50x100xf32> to memref<?x?xf32, #strided>
-// %5 = std.subview %result[%k, %l][%c10, %c25][%c1, %c1]
+// %5 = memref.subview %result[%k, %l][%c10, %c25][%c1, %c1]
// : memref<50x100xf32> to memref<?x?xf32, #strided>
// linalg.indexed_generic pointwise_2d_trait %4, %5 {
// ^bb0(%i: index, %j: index, %operand_in: f32, %result_in: f32):
@@ -262,7 +263,7 @@ makeTiledShapes(OpBuilder &b, Location loc, LinalgOp linalgOp,
for (unsigned r = 0; r < rank; ++r) {
if (!isTiled(map.getSubMap({r}), tileSizes)) {
offsets.push_back(b.getIndexAttr(0));
- sizes.push_back(std_dim(shapedOp, r).value);
+ sizes.push_back(memref_dim(shapedOp, r).value);
strides.push_back(b.getIndexAttr(1));
continue;
}
@@ -290,7 +291,7 @@ makeTiledShapes(OpBuilder &b, Location loc, LinalgOp linalgOp,
getAffineDimExpr(/*position=*/1, b.getContext()) -
getAffineDimExpr(/*position=*/2, b.getContext())},
b.getContext());
- auto d = std_dim(shapedOp, r);
+ Value d = memref_dim(shapedOp, r);
SmallVector<Value, 4> operands{size, d, offset};
fullyComposeAffineMapAndOperands(&minMap, &operands);
size = affine_min(b.getIndexType(), minMap, operands);
@@ -302,7 +303,7 @@ makeTiledShapes(OpBuilder &b, Location loc, LinalgOp linalgOp,
if (shapedType.isa<MemRefType>())
res.push_back(
- b.create<SubViewOp>(loc, shapedOp, offsets, sizes, strides));
+ b.create<memref::SubViewOp>(loc, shapedOp, offsets, sizes, strides));
else
res.push_back(
b.create<SubTensorOp>(loc, shapedOp, offsets, sizes, strides));
@@ -474,7 +475,7 @@ Optional<TiledLinalgOp> static tileLinalgOpImpl(
if (!options.tileSizeComputationFunction)
return llvm::None;
-
+
// Enforce the convention that "tiling by zero" skips tiling a particular
// dimension. This convention is significantly simpler to handle instead of
// adjusting affine maps to account for missing dimensions.
@@ -564,9 +565,9 @@ void mlir::linalg::populateLinalgTilingCanonicalizationPatterns(
scf::ParallelOp::getCanonicalizationPatterns(patterns, ctx);
ConstantIndexOp::getCanonicalizationPatterns(patterns, ctx);
SubTensorOp::getCanonicalizationPatterns(patterns, ctx);
- SubViewOp::getCanonicalizationPatterns(patterns, ctx);
+ memref::SubViewOp::getCanonicalizationPatterns(patterns, ctx);
tensor::CastOp::getCanonicalizationPatterns(patterns, ctx);
- ViewOp::getCanonicalizationPatterns(patterns, ctx);
+ memref::ViewOp::getCanonicalizationPatterns(patterns, ctx);
CanonicalizationPatternList<
#define GET_OP_LIST
#include "mlir/Dialect/Linalg/IR/LinalgStructuredOps.cpp.inc"
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp b/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp
index 7f604807030d..c2e52c63eabd 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Transforms.cpp
@@ -212,7 +212,7 @@ static LogicalResult rewriteAsPaddedOp(PatternRewriter &rewriter,
SmallVector<OpFoldResult> offsets(rank, rewriter.getIndexAttr(0));
auto sizes = llvm::to_vector<4>(llvm::map_range(
llvm::seq<unsigned>(0, rank), [&](unsigned d) -> OpFoldResult {
- auto dimOp = rewriter.create<DimOp>(loc, std::get<0>(it), d);
+ auto dimOp = rewriter.create<memref::DimOp>(loc, std::get<0>(it), d);
newUsersOfOpToPad.insert(dimOp);
return dimOp.getResult();
}));
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
index 48b6165d7b68..880e7f385724 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
+++ b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
@@ -85,7 +85,7 @@ static VectorType extractVectorTypeFromShapedValue(Value v) {
}
/// Build a vector.transfer_read from `source` at indices set to all `0`.
-/// If source has rank zero, build an std.load.
+/// If source has rank zero, build an memref.load.
/// Return the produced value.
static Value buildVectorRead(OpBuilder &builder, Value source) {
edsc::ScopedContext scope(builder);
@@ -94,11 +94,11 @@ static Value buildVectorRead(OpBuilder &builder, Value source) {
SmallVector<Value> indices(shapedType.getRank(), std_constant_index(0));
return vector_transfer_read(vectorType, source, indices);
}
- return std_load(source);
+ return memref_load(source);
}
/// Build a vector.transfer_write of `value` into `dest` at indices set to all
-/// `0`. If `dest` has null rank, build an std.store.
+/// `0`. If `dest` has null rank, build an memref.store.
/// Return the produced value or null if no value is produced.
static Value buildVectorWrite(OpBuilder &builder, Value value, Value dest) {
edsc::ScopedContext scope(builder);
@@ -110,7 +110,7 @@ static Value buildVectorWrite(OpBuilder &builder, Value value, Value dest) {
value = vector_broadcast(vectorType, value);
write = vector_transfer_write(value, dest, indices);
} else {
- write = std_store(value, dest);
+ write = memref_store(value, dest);
}
LLVM_DEBUG(dbgs() << "\n[" DEBUG_TYPE "]: vectorized op: " << *write);
if (!write->getResults().empty())
@@ -544,7 +544,7 @@ LogicalResult ConvOpVectorization<ConvOp, N>::matchAndRewrite(
rewriter.getAffineMapArrayAttr(indexingMaps),
rewriter.getStrArrayAttr(iteratorTypes));
- rewriter.create<StoreOp>(loc, result, output, ValueRange(zeros));
+ rewriter.create<memref::StoreOp>(loc, result, output, ValueRange(zeros));
rewriter.eraseOp(op);
return success();
}
@@ -667,12 +667,12 @@ static bool mayExistInterleavedUses(Operation *firstOp, Operation *secondOp,
}
/// Return the unique subview use of `v` if it is indeed unique, null otherwise.
-static SubViewOp getSubViewUseIfUnique(Value v) {
- SubViewOp subViewOp;
+static memref::SubViewOp getSubViewUseIfUnique(Value v) {
+ memref::SubViewOp subViewOp;
for (auto &u : v.getUses()) {
- if (auto newSubViewOp = dyn_cast<SubViewOp>(u.getOwner())) {
+ if (auto newSubViewOp = dyn_cast<memref::SubViewOp>(u.getOwner())) {
if (subViewOp)
- return SubViewOp();
+ return memref::SubViewOp();
subViewOp = newSubViewOp;
}
}
@@ -686,14 +686,14 @@ LogicalResult LinalgCopyVTRForwardingPattern::matchAndRewrite(
// Transfer into `view`.
Value viewOrAlloc = xferOp.source();
- if (!viewOrAlloc.getDefiningOp<ViewOp>() &&
- !viewOrAlloc.getDefiningOp<AllocOp>())
+ if (!viewOrAlloc.getDefiningOp<memref::ViewOp>() &&
+ !viewOrAlloc.getDefiningOp<memref::AllocOp>())
return failure();
LLVM_DEBUG(llvm::dbgs() << "\n[" DEBUG_TYPE "]: " << viewOrAlloc);
// Ensure there is exactly one subview of `viewOrAlloc` defining `subView`.
- SubViewOp subViewOp = getSubViewUseIfUnique(viewOrAlloc);
+ memref::SubViewOp subViewOp = getSubViewUseIfUnique(viewOrAlloc);
if (!subViewOp)
return failure();
Value subView = subViewOp.getResult();
@@ -765,12 +765,12 @@ LogicalResult LinalgCopyVTWForwardingPattern::matchAndRewrite(
vector::TransferWriteOp xferOp, PatternRewriter &rewriter) const {
// Transfer into `viewOrAlloc`.
Value viewOrAlloc = xferOp.source();
- if (!viewOrAlloc.getDefiningOp<ViewOp>() &&
- !viewOrAlloc.getDefiningOp<AllocOp>())
+ if (!viewOrAlloc.getDefiningOp<memref::ViewOp>() &&
+ !viewOrAlloc.getDefiningOp<memref::AllocOp>())
return failure();
// Ensure there is exactly one subview of `viewOrAlloc` defining `subView`.
- SubViewOp subViewOp = getSubViewUseIfUnique(viewOrAlloc);
+ memref::SubViewOp subViewOp = getSubViewUseIfUnique(viewOrAlloc);
if (!subViewOp)
return failure();
Value subView = subViewOp.getResult();
diff --git a/mlir/lib/Dialect/MemRef/CMakeLists.txt b/mlir/lib/Dialect/MemRef/CMakeLists.txt
new file mode 100644
index 000000000000..f33061b2d87c
--- /dev/null
+++ b/mlir/lib/Dialect/MemRef/CMakeLists.txt
@@ -0,0 +1 @@
+add_subdirectory(IR)
diff --git a/mlir/lib/Dialect/MemRef/IR/CMakeLists.txt b/mlir/lib/Dialect/MemRef/IR/CMakeLists.txt
new file mode 100644
index 000000000000..e155d5515b8b
--- /dev/null
+++ b/mlir/lib/Dialect/MemRef/IR/CMakeLists.txt
@@ -0,0 +1,17 @@
+add_mlir_dialect_library(MLIRMemRef
+ MemRefDialect.cpp
+ MemRefOps.cpp
+
+ ADDITIONAL_HEADER_DIRS
+ ${PROJECT_SOURCE_DIR}/inlude/mlir/Dialect/MemRefDialect
+
+ DEPENDS
+ MLIRMemRefOpsIncGen
+
+ LINK_COMPONENTS
+ Core
+
+ LINK_LIBS PUBLIC
+ MLIRDialect
+ MLIRIR
+)
diff --git a/mlir/lib/Dialect/MemRef/IR/MemRefDialect.cpp b/mlir/lib/Dialect/MemRef/IR/MemRefDialect.cpp
new file mode 100644
index 000000000000..47e9eba1d854
--- /dev/null
+++ b/mlir/lib/Dialect/MemRef/IR/MemRefDialect.cpp
@@ -0,0 +1,39 @@
+//===----------------------------------------------------------------------===//
+//
+// 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/Dialect/MemRef/IR/MemRef.h"
+#include "mlir/Transforms/InliningUtils.h"
+
+using namespace mlir;
+using namespace mlir::memref;
+
+//===----------------------------------------------------------------------===//
+// MemRefDialect Dialect Interfaces
+//===----------------------------------------------------------------------===//
+
+namespace {
+struct MemRefInlinerInterface : public DialectInlinerInterface {
+ using DialectInlinerInterface::DialectInlinerInterface;
+ bool isLegalToInline(Region *dest, Region *src, bool wouldBeCloned,
+ BlockAndValueMapping &valueMapping) const final {
+ return true;
+ }
+ bool isLegalToInline(Operation *, Region *, bool wouldBeCloned,
+ BlockAndValueMapping &) const final {
+ return true;
+ }
+};
+} // end anonymous namespace
+
+void mlir::memref::MemRefDialect::initialize() {
+ addOperations<DmaStartOp, DmaWaitOp,
+#define GET_OP_LIST
+#include "mlir/Dialect/MemRef/IR/MemRefOps.cpp.inc"
+ >();
+ addInterfaces<MemRefInlinerInterface>();
+}
diff --git a/mlir/lib/Dialect/MemRef/IR/MemRefOps.cpp b/mlir/lib/Dialect/MemRef/IR/MemRefOps.cpp
new file mode 100644
index 000000000000..10bfa8b40825
--- /dev/null
+++ b/mlir/lib/Dialect/MemRef/IR/MemRefOps.cpp
@@ -0,0 +1,2128 @@
+//===----------------------------------------------------------------------===//
+//
+// 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/Dialect/MemRef/IR/MemRef.h"
+#include "mlir/Dialect/StandardOps/IR/Ops.h"
+#include "mlir/Dialect/StandardOps/Utils/Utils.h"
+#include "mlir/Dialect/Tensor/IR/Tensor.h"
+#include "mlir/IR/AffineMap.h"
+#include "mlir/IR/Builders.h"
+#include "mlir/IR/BuiltinTypes.h"
+#include "mlir/IR/Matchers.h"
+#include "mlir/IR/PatternMatch.h"
+#include "mlir/IR/TypeUtilities.h"
+#include "llvm/ADT/STLExtras.h"
+
+using namespace mlir;
+using namespace mlir::memref;
+
+/// Materialize a single constant operation from a given attribute value with
+/// the desired resultant type.
+Operation *MemRefDialect::materializeConstant(OpBuilder &builder,
+ Attribute value, Type type,
+ Location loc) {
+ return builder.create<mlir::ConstantOp>(loc, type, value);
+}
+
+/// Extract int64_t values from the assumed ArrayAttr of IntegerAttr.
+static SmallVector<int64_t, 4> extractFromI64ArrayAttr(Attribute attr) {
+ return llvm::to_vector<4>(
+ llvm::map_range(attr.cast<ArrayAttr>(), [](Attribute a) -> int64_t {
+ return a.cast<IntegerAttr>().getInt();
+ }));
+}
+
+/// Helper function to dispatch an OpFoldResult into either the `dynamicVec` if
+/// it is a Value or into `staticVec` if it is an IntegerAttr.
+/// In the case of a Value, a copy of the `sentinel` value is also pushed to
+/// `staticVec`. This is useful to extract mixed static and dynamic entries that
+/// come from an AttrSizedOperandSegments trait.
+static void dispatchIndexOpFoldResult(OpFoldResult ofr,
+ SmallVectorImpl<Value> &dynamicVec,
+ SmallVectorImpl<int64_t> &staticVec,
+ int64_t sentinel) {
+ if (auto v = ofr.dyn_cast<Value>()) {
+ dynamicVec.push_back(v);
+ staticVec.push_back(sentinel);
+ return;
+ }
+ APInt apInt = ofr.dyn_cast<Attribute>().cast<IntegerAttr>().getValue();
+ staticVec.push_back(apInt.getSExtValue());
+}
+
+static void dispatchIndexOpFoldResults(ArrayRef<OpFoldResult> ofrs,
+ SmallVectorImpl<Value> &dynamicVec,
+ SmallVectorImpl<int64_t> &staticVec,
+ int64_t sentinel) {
+ for (auto ofr : ofrs)
+ dispatchIndexOpFoldResult(ofr, dynamicVec, staticVec, sentinel);
+}
+
+//===----------------------------------------------------------------------===//
+// Common canonicalization pattern support logic
+//===----------------------------------------------------------------------===//
+
+/// This is a common class used for patterns of the form
+/// "someop(memrefcast) -> someop". It folds the source of any memref.cast
+/// into the root operation directly.
+static LogicalResult foldMemRefCast(Operation *op) {
+ bool folded = false;
+ for (OpOperand &operand : op->getOpOperands()) {
+ auto cast = operand.get().getDefiningOp<CastOp>();
+ if (cast && !cast.getOperand().getType().isa<UnrankedMemRefType>()) {
+ operand.set(cast.getOperand());
+ folded = true;
+ }
+ }
+ return success(folded);
+}
+
+//===----------------------------------------------------------------------===//
+// Helpers for GlobalOp
+//===----------------------------------------------------------------------===//
+
+static Type getTensorTypeFromMemRefType(Type type) {
+ if (auto memref = type.dyn_cast<MemRefType>())
+ return RankedTensorType::get(memref.getShape(), memref.getElementType());
+ if (auto memref = type.dyn_cast<UnrankedMemRefType>())
+ return UnrankedTensorType::get(memref.getElementType());
+ return NoneType::get(type.getContext());
+}
+
+//===----------------------------------------------------------------------===//
+// AllocOp / AllocaOp
+//===----------------------------------------------------------------------===//
+
+template <typename AllocLikeOp>
+static LogicalResult verifyAllocLikeOp(AllocLikeOp op) {
+ static_assert(llvm::is_one_of<AllocLikeOp, AllocOp, AllocaOp>::value,
+ "applies to only alloc or alloca");
+ auto memRefType = op.getResult().getType().template dyn_cast<MemRefType>();
+ if (!memRefType)
+ return op.emitOpError("result must be a memref");
+
+ if (static_cast<int64_t>(op.dynamicSizes().size()) !=
+ memRefType.getNumDynamicDims())
+ return op.emitOpError("dimension operand count does not equal memref "
+ "dynamic dimension count");
+
+ unsigned numSymbols = 0;
+ if (!memRefType.getAffineMaps().empty())
+ numSymbols = memRefType.getAffineMaps().front().getNumSymbols();
+ if (op.symbolOperands().size() != numSymbols)
+ return op.emitOpError(
+ "symbol operand count does not equal memref symbol count");
+
+ return success();
+}
+
+static LogicalResult verify(AllocOp op) { return verifyAllocLikeOp(op); }
+
+static LogicalResult verify(AllocaOp op) {
+ // An alloca op needs to have an ancestor with an allocation scope trait.
+ if (!op->getParentWithTrait<OpTrait::AutomaticAllocationScope>())
+ return op.emitOpError(
+ "requires an ancestor op with AutomaticAllocationScope trait");
+
+ return verifyAllocLikeOp(op);
+}
+
+namespace {
+/// Fold constant dimensions into an alloc like operation.
+template <typename AllocLikeOp>
+struct SimplifyAllocConst : public OpRewritePattern<AllocLikeOp> {
+ using OpRewritePattern<AllocLikeOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(AllocLikeOp alloc,
+ PatternRewriter &rewriter) const override {
+ // Check to see if any dimensions operands are constants. If so, we can
+ // substitute and drop them.
+ if (llvm::none_of(alloc.getOperands(), [](Value operand) {
+ return matchPattern(operand, matchConstantIndex());
+ }))
+ return failure();
+
+ auto memrefType = alloc.getType();
+
+ // Ok, we have one or more constant operands. Collect the non-constant ones
+ // and keep track of the resultant memref type to build.
+ SmallVector<int64_t, 4> newShapeConstants;
+ newShapeConstants.reserve(memrefType.getRank());
+ SmallVector<Value, 4> newOperands;
+
+ unsigned dynamicDimPos = 0;
+ for (unsigned dim = 0, e = memrefType.getRank(); dim < e; ++dim) {
+ int64_t dimSize = memrefType.getDimSize(dim);
+ // If this is already static dimension, keep it.
+ if (dimSize != -1) {
+ newShapeConstants.push_back(dimSize);
+ continue;
+ }
+ auto *defOp = alloc.getOperand(dynamicDimPos).getDefiningOp();
+ if (auto constantIndexOp = dyn_cast_or_null<ConstantIndexOp>(defOp)) {
+ // Dynamic shape dimension will be folded.
+ newShapeConstants.push_back(constantIndexOp.getValue());
+ } else {
+ // Dynamic shape dimension not folded; copy operand from old memref.
+ newShapeConstants.push_back(-1);
+ newOperands.push_back(alloc.getOperand(dynamicDimPos));
+ }
+ dynamicDimPos++;
+ }
+
+ // Create new memref type (which will have fewer dynamic dimensions).
+ MemRefType newMemRefType =
+ MemRefType::Builder(memrefType).setShape(newShapeConstants);
+ assert(static_cast<int64_t>(newOperands.size()) ==
+ newMemRefType.getNumDynamicDims());
+
+ // Create and insert the alloc op for the new memref.
+ auto newAlloc = rewriter.create<AllocLikeOp>(alloc.getLoc(), newMemRefType,
+ newOperands, IntegerAttr());
+ // Insert a cast so we have the same type as the old alloc.
+ auto resultCast =
+ rewriter.create<CastOp>(alloc.getLoc(), newAlloc, alloc.getType());
+
+ rewriter.replaceOp(alloc, {resultCast});
+ return success();
+ }
+};
+
+/// Fold alloc operations with no uses. Alloc has side effects on the heap,
+/// but can still be deleted if it has zero uses.
+struct SimplifyDeadAlloc : public OpRewritePattern<AllocOp> {
+ using OpRewritePattern<AllocOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(AllocOp alloc,
+ PatternRewriter &rewriter) const override {
+ if (alloc.use_empty()) {
+ rewriter.eraseOp(alloc);
+ return success();
+ }
+ return failure();
+ }
+};
+} // end anonymous namespace.
+
+void AllocOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<SimplifyAllocConst<AllocOp>, SimplifyDeadAlloc>(context);
+}
+
+void AllocaOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<SimplifyAllocConst<AllocaOp>>(context);
+}
+
+//===----------------------------------------------------------------------===//
+// AssumeAlignmentOp
+//===----------------------------------------------------------------------===//
+
+static LogicalResult verify(AssumeAlignmentOp op) {
+ unsigned alignment = op.alignment();
+ if (!llvm::isPowerOf2_32(alignment))
+ return op.emitOpError("alignment must be power of 2");
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// BufferCastOp
+//===----------------------------------------------------------------------===//
+
+OpFoldResult BufferCastOp::fold(ArrayRef<Attribute>) {
+ if (auto tensorLoad = tensor().getDefiningOp<TensorLoadOp>())
+ if (tensorLoad.memref().getType() == getType())
+ return tensorLoad.memref();
+ return {};
+}
+
+namespace {
+/// Replace tensor_cast + buffer_cast by buffer_cast + memref_cast.
+struct BufferCast : public OpRewritePattern<BufferCastOp> {
+ using OpRewritePattern<BufferCastOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(BufferCastOp bufferCast,
+ PatternRewriter &rewriter) const final {
+ auto tensorCastOperand =
+ bufferCast.getOperand().getDefiningOp<tensor::CastOp>();
+ if (!tensorCastOperand)
+ return failure();
+ auto srcTensorType =
+ tensorCastOperand.getOperand().getType().dyn_cast<RankedTensorType>();
+ if (!srcTensorType)
+ return failure();
+ auto memrefType = MemRefType::get(srcTensorType.getShape(),
+ srcTensorType.getElementType());
+ Value memref = rewriter.create<BufferCastOp>(
+ bufferCast.getLoc(), memrefType, tensorCastOperand.getOperand());
+ rewriter.replaceOpWithNewOp<CastOp>(bufferCast, bufferCast.getType(),
+ memref);
+ return success();
+ }
+};
+
+/// Canonicalize memref.tensor_load + memref.buffer_cast to memref.cast when
+/// type mismatches prevent `BufferCastOp::fold` to kick in.
+struct TensorLoadToMemRef : public OpRewritePattern<BufferCastOp> {
+ using OpRewritePattern<BufferCastOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(BufferCastOp bufferCast,
+ PatternRewriter &rewriter) const final {
+ auto tensorLoad = bufferCast.tensor().getDefiningOp<TensorLoadOp>();
+ // Bail unless we have a tensor_load + memref.buffer_cast with
diff erent
+ // types. `BufferCastOp::fold` handles the same type case.
+ if (!tensorLoad || tensorLoad.memref().getType() == bufferCast.getType())
+ return failure();
+ // If types are not cast-compatible, bail.
+ if (!CastOp::areCastCompatible(tensorLoad.memref().getType(),
+ bufferCast.getType()))
+ return failure();
+ rewriter.replaceOpWithNewOp<CastOp>(bufferCast, bufferCast.getType(),
+ tensorLoad.memref());
+ return success();
+ }
+};
+
+} // namespace
+
+void BufferCastOp::getCanonicalizationPatterns(
+ OwningRewritePatternList &results, MLIRContext *context) {
+ results.insert<BufferCast, TensorLoadToMemRef>(context);
+}
+
+//===----------------------------------------------------------------------===//
+// CastOp
+//===----------------------------------------------------------------------===//
+
+/// Determines whether MemRef_CastOp casts to a more dynamic version of the
+/// source memref. This is useful to to fold a memref.cast into a consuming op
+/// and implement canonicalization patterns for ops in
diff erent dialects that
+/// may consume the results of memref.cast operations. Such foldable memref.cast
+/// operations are typically inserted as `view` and `subview` ops are
+/// canonicalized, to preserve the type compatibility of their uses.
+///
+/// Returns true when all conditions are met:
+/// 1. source and result are ranked memrefs with strided semantics and same
+/// element type and rank.
+/// 2. each of the source's size, offset or stride has more static information
+/// than the corresponding result's size, offset or stride.
+///
+/// Example 1:
+/// ```mlir
+/// %1 = memref.cast %0 : memref<8x16xf32> to memref<?x?xf32>
+/// %2 = consumer %1 ... : memref<?x?xf32> ...
+/// ```
+///
+/// may fold into:
+///
+/// ```mlir
+/// %2 = consumer %0 ... : memref<8x16xf32> ...
+/// ```
+///
+/// Example 2:
+/// ```
+/// %1 = memref.cast %0 : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
+/// to memref<?x?xf32>
+/// consumer %1 : memref<?x?xf32> ...
+/// ```
+///
+/// may fold into:
+///
+/// ```
+/// consumer %0 ... : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
+/// ```
+bool CastOp::canFoldIntoConsumerOp(CastOp castOp) {
+ MemRefType sourceType = castOp.source().getType().dyn_cast<MemRefType>();
+ MemRefType resultType = castOp.getType().dyn_cast<MemRefType>();
+
+ // Requires ranked MemRefType.
+ if (!sourceType || !resultType)
+ return false;
+
+ // Requires same elemental type.
+ if (sourceType.getElementType() != resultType.getElementType())
+ return false;
+
+ // Requires same rank.
+ if (sourceType.getRank() != resultType.getRank())
+ return false;
+
+ // Only fold casts between strided memref forms.
+ int64_t sourceOffset, resultOffset;
+ SmallVector<int64_t, 4> sourceStrides, resultStrides;
+ if (failed(getStridesAndOffset(sourceType, sourceStrides, sourceOffset)) ||
+ failed(getStridesAndOffset(resultType, resultStrides, resultOffset)))
+ return false;
+
+ // If cast is towards more static sizes along any dimension, don't fold.
+ for (auto it : llvm::zip(sourceType.getShape(), resultType.getShape())) {
+ auto ss = std::get<0>(it), st = std::get<1>(it);
+ if (ss != st)
+ if (MemRefType::isDynamic(ss) && !MemRefType::isDynamic(st))
+ return false;
+ }
+
+ // If cast is towards more static offset along any dimension, don't fold.
+ if (sourceOffset != resultOffset)
+ if (MemRefType::isDynamicStrideOrOffset(sourceOffset) &&
+ !MemRefType::isDynamicStrideOrOffset(resultOffset))
+ return false;
+
+ // If cast is towards more static strides along any dimension, don't fold.
+ for (auto it : llvm::zip(sourceStrides, resultStrides)) {
+ auto ss = std::get<0>(it), st = std::get<1>(it);
+ if (ss != st)
+ if (MemRefType::isDynamicStrideOrOffset(ss) &&
+ !MemRefType::isDynamicStrideOrOffset(st))
+ return false;
+ }
+
+ return true;
+}
+
+bool CastOp::areCastCompatible(TypeRange inputs, TypeRange outputs) {
+ if (inputs.size() != 1 || outputs.size() != 1)
+ return false;
+ Type a = inputs.front(), b = outputs.front();
+ auto aT = a.dyn_cast<MemRefType>();
+ auto bT = b.dyn_cast<MemRefType>();
+
+ auto uaT = a.dyn_cast<UnrankedMemRefType>();
+ auto ubT = b.dyn_cast<UnrankedMemRefType>();
+
+ if (aT && bT) {
+ if (aT.getElementType() != bT.getElementType())
+ return false;
+ if (aT.getAffineMaps() != bT.getAffineMaps()) {
+ int64_t aOffset, bOffset;
+ SmallVector<int64_t, 4> aStrides, bStrides;
+ if (failed(getStridesAndOffset(aT, aStrides, aOffset)) ||
+ failed(getStridesAndOffset(bT, bStrides, bOffset)) ||
+ aStrides.size() != bStrides.size())
+ return false;
+
+ // Strides along a dimension/offset are compatible if the value in the
+ // source memref is static and the value in the target memref is the
+ // same. They are also compatible if either one is dynamic (see
+ // description of MemRefCastOp for details).
+ auto checkCompatible = [](int64_t a, int64_t b) {
+ return (a == MemRefType::getDynamicStrideOrOffset() ||
+ b == MemRefType::getDynamicStrideOrOffset() || a == b);
+ };
+ if (!checkCompatible(aOffset, bOffset))
+ return false;
+ for (auto aStride : enumerate(aStrides))
+ if (!checkCompatible(aStride.value(), bStrides[aStride.index()]))
+ return false;
+ }
+ if (aT.getMemorySpaceAsInt() != bT.getMemorySpaceAsInt())
+ return false;
+
+ // They must have the same rank, and any specified dimensions must match.
+ if (aT.getRank() != bT.getRank())
+ return false;
+
+ for (unsigned i = 0, e = aT.getRank(); i != e; ++i) {
+ int64_t aDim = aT.getDimSize(i), bDim = bT.getDimSize(i);
+ if (aDim != -1 && bDim != -1 && aDim != bDim)
+ return false;
+ }
+ return true;
+ } else {
+ if (!aT && !uaT)
+ return false;
+ if (!bT && !ubT)
+ return false;
+ // Unranked to unranked casting is unsupported
+ if (uaT && ubT)
+ return false;
+
+ auto aEltType = (aT) ? aT.getElementType() : uaT.getElementType();
+ auto bEltType = (bT) ? bT.getElementType() : ubT.getElementType();
+ if (aEltType != bEltType)
+ return false;
+
+ auto aMemSpace =
+ (aT) ? aT.getMemorySpaceAsInt() : uaT.getMemorySpaceAsInt();
+ auto bMemSpace =
+ (bT) ? bT.getMemorySpaceAsInt() : ubT.getMemorySpaceAsInt();
+ if (aMemSpace != bMemSpace)
+ return false;
+
+ return true;
+ }
+
+ return false;
+}
+
+OpFoldResult CastOp::fold(ArrayRef<Attribute> operands) {
+ return succeeded(foldMemRefCast(*this)) ? getResult() : Value();
+}
+
+//===----------------------------------------------------------------------===//
+// DeallocOp
+//===----------------------------------------------------------------------===//
+namespace {
+/// Fold Dealloc operations that are deallocating an AllocOp that is only used
+/// by other Dealloc operations.
+struct SimplifyDeadDealloc : public OpRewritePattern<DeallocOp> {
+ using OpRewritePattern<DeallocOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(DeallocOp dealloc,
+ PatternRewriter &rewriter) const override {
+ // Check that the memref operand's defining operation is an AllocOp.
+ Value memref = dealloc.memref();
+ if (!isa_and_nonnull<AllocOp>(memref.getDefiningOp()))
+ return failure();
+
+ // Check that all of the uses of the AllocOp are other DeallocOps.
+ for (auto *user : memref.getUsers())
+ if (!isa<DeallocOp>(user))
+ return failure();
+
+ // Erase the dealloc operation.
+ rewriter.eraseOp(dealloc);
+ return success();
+ }
+};
+} // end anonymous namespace.
+
+static LogicalResult verify(DeallocOp op) {
+ if (!op.memref().getType().isa<MemRefType>())
+ return op.emitOpError("operand must be a memref");
+ return success();
+}
+
+void DeallocOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<SimplifyDeadDealloc>(context);
+}
+
+LogicalResult DeallocOp::fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results) {
+ /// dealloc(memrefcast) -> dealloc
+ return foldMemRefCast(*this);
+}
+
+//===----------------------------------------------------------------------===//
+// DimOp
+//===----------------------------------------------------------------------===//
+
+void DimOp::build(OpBuilder &builder, OperationState &result, Value memref,
+ int64_t index) {
+ auto loc = result.location;
+ Value indexValue = builder.create<ConstantIndexOp>(loc, index);
+ build(builder, result, memref, indexValue);
+}
+
+void DimOp::build(OpBuilder &builder, OperationState &result, Value memref,
+ Value index) {
+ auto indexTy = builder.getIndexType();
+ build(builder, result, indexTy, memref, index);
+}
+
+Optional<int64_t> DimOp::getConstantIndex() {
+ if (auto constantOp = index().getDefiningOp<ConstantOp>())
+ return constantOp.getValue().cast<IntegerAttr>().getInt();
+ return {};
+}
+
+static LogicalResult verify(DimOp op) {
+ // Assume unknown index to be in range.
+ Optional<int64_t> index = op.getConstantIndex();
+ if (!index.hasValue())
+ return success();
+
+ // Check that constant index is not knowingly out of range.
+ auto type = op.memrefOrTensor().getType();
+ if (auto memrefType = type.dyn_cast<MemRefType>()) {
+ if (index.getValue() >= memrefType.getRank())
+ return op.emitOpError("index is out of range");
+ } else if (auto tensorType = type.dyn_cast<RankedTensorType>()) {
+ if (index.getValue() >= tensorType.getRank())
+ return op.emitOpError("index is out of range");
+ } else if (type.isa<UnrankedMemRefType>() || type.isa<UnrankedTensorType>()) {
+ // Assume index to be in range.
+ } else {
+ llvm_unreachable("expected operand with memref type");
+ }
+ return success();
+}
+
+OpFoldResult DimOp::fold(ArrayRef<Attribute> operands) {
+ auto index = operands[1].dyn_cast_or_null<IntegerAttr>();
+
+ // All forms of folding require a known index.
+ if (!index)
+ return {};
+
+ auto argTy = memrefOrTensor().getType();
+ // Fold if the shape extent along the given index is known.
+ if (auto shapedTy = argTy.dyn_cast<ShapedType>()) {
+ // Folding for unranked types (UnrankedMemRefType) is not supported.
+ if (!shapedTy.hasRank())
+ return {};
+ if (!shapedTy.isDynamicDim(index.getInt())) {
+ Builder builder(getContext());
+ return builder.getIndexAttr(shapedTy.getShape()[index.getInt()]);
+ }
+ }
+
+ Operation *definingOp = memrefOrTensor().getDefiningOp();
+
+ // dim(memref.tensor_load(memref)) -> dim(memref)
+ if (auto tensorLoadOp = dyn_cast_or_null<TensorLoadOp>(definingOp)) {
+ setOperand(0, tensorLoadOp.memref());
+ return getResult();
+ }
+
+ // Fold dim to the operand of tensor.generate.
+ if (auto fromElements = dyn_cast_or_null<tensor::GenerateOp>(definingOp)) {
+ auto resultType =
+ fromElements.getResult().getType().cast<RankedTensorType>();
+ // The case where the type encodes the size of the dimension is handled
+ // above.
+ assert(resultType.getShape()[index.getInt()] ==
+ RankedTensorType::kDynamicSize);
+
+ // Find the operand of the fromElements that corresponds to this index.
+ auto dynExtents = fromElements.dynamicExtents().begin();
+ for (auto dim : resultType.getShape().take_front(index.getInt()))
+ if (dim == RankedTensorType::kDynamicSize)
+ dynExtents++;
+
+ return Value{*dynExtents};
+ }
+
+ // The size at the given index is now known to be a dynamic size.
+ unsigned unsignedIndex = index.getValue().getZExtValue();
+
+ if (auto subtensor = dyn_cast_or_null<mlir::SubTensorOp>(definingOp)) {
+ assert(subtensor.isDynamicSize(unsignedIndex) &&
+ "Expected dynamic subtensor size");
+ return subtensor.getDynamicSize(unsignedIndex);
+ }
+
+ // Fold dim to the size argument for an `AllocOp`, `ViewOp`, or `SubViewOp`.
+ auto memrefType = argTy.dyn_cast<MemRefType>();
+ if (!memrefType)
+ return {};
+
+ if (auto alloc = dyn_cast_or_null<AllocOp>(definingOp))
+ return *(alloc.getDynamicSizes().begin() +
+ memrefType.getDynamicDimIndex(unsignedIndex));
+
+ if (auto view = dyn_cast_or_null<ViewOp>(definingOp))
+ return *(view.getDynamicSizes().begin() +
+ memrefType.getDynamicDimIndex(unsignedIndex));
+
+ if (auto subview = dyn_cast_or_null<SubViewOp>(definingOp)) {
+ assert(subview.isDynamicSize(unsignedIndex) &&
+ "Expected dynamic subview size");
+ return subview.getDynamicSize(unsignedIndex);
+ }
+
+ // dim(memrefcast) -> dim
+ if (succeeded(foldMemRefCast(*this)))
+ return getResult();
+
+ return {};
+}
+
+namespace {
+/// Fold dim of a memref reshape operation to a load into the reshape's shape
+/// operand.
+struct DimOfMemRefReshape : public OpRewritePattern<DimOp> {
+ using OpRewritePattern<DimOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(DimOp dim,
+ PatternRewriter &rewriter) const override {
+ auto reshape = dim.memrefOrTensor().getDefiningOp<ReshapeOp>();
+
+ if (!reshape)
+ return failure();
+
+ // Place the load directly after the reshape to ensure that the shape memref
+ // was not mutated.
+ rewriter.setInsertionPointAfter(reshape);
+ rewriter.replaceOpWithNewOp<LoadOp>(dim, reshape.shape(),
+ llvm::makeArrayRef({dim.index()}));
+ return success();
+ }
+};
+
+/// Fold dim of a dim of a cast into the dim of the source of the tensor cast.
+template <typename CastOpTy>
+struct DimOfCastOp : public OpRewritePattern<DimOp> {
+ using OpRewritePattern<DimOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(DimOp dimOp,
+ PatternRewriter &rewriter) const override {
+ auto castOp = dimOp.memrefOrTensor().getDefiningOp<CastOpTy>();
+ if (!castOp)
+ return failure();
+ Value newSource = castOp.getOperand();
+ rewriter.replaceOpWithNewOp<DimOp>(dimOp, newSource, dimOp.index());
+ return success();
+ }
+};
+} // end anonymous namespace.
+
+void DimOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<DimOfMemRefReshape, DimOfCastOp<BufferCastOp>,
+ DimOfCastOp<tensor::CastOp>>(context);
+}
+
+// ---------------------------------------------------------------------------
+// DmaStartOp
+// ---------------------------------------------------------------------------
+
+void DmaStartOp::build(OpBuilder &builder, OperationState &result,
+ Value srcMemRef, ValueRange srcIndices, Value destMemRef,
+ ValueRange destIndices, Value numElements,
+ Value tagMemRef, ValueRange tagIndices, Value stride,
+ Value elementsPerStride) {
+ result.addOperands(srcMemRef);
+ result.addOperands(srcIndices);
+ result.addOperands(destMemRef);
+ result.addOperands(destIndices);
+ result.addOperands({numElements, tagMemRef});
+ result.addOperands(tagIndices);
+ if (stride)
+ result.addOperands({stride, elementsPerStride});
+}
+
+void DmaStartOp::print(OpAsmPrinter &p) {
+ p << getOperationName() << " " << getSrcMemRef() << '[' << getSrcIndices()
+ << "], " << getDstMemRef() << '[' << getDstIndices() << "], "
+ << getNumElements() << ", " << getTagMemRef() << '[' << getTagIndices()
+ << ']';
+ if (isStrided())
+ p << ", " << getStride() << ", " << getNumElementsPerStride();
+
+ p.printOptionalAttrDict((*this)->getAttrs());
+ p << " : " << getSrcMemRef().getType() << ", " << getDstMemRef().getType()
+ << ", " << getTagMemRef().getType();
+}
+
+// Parse DmaStartOp.
+// Ex:
+// %dma_id = dma_start %src[%i, %j], %dst[%k, %l], %size,
+// %tag[%index], %stride, %num_elt_per_stride :
+// : memref<3076 x f32, 0>,
+// memref<1024 x f32, 2>,
+// memref<1 x i32>
+//
+ParseResult DmaStartOp::parse(OpAsmParser &parser, OperationState &result) {
+ OpAsmParser::OperandType srcMemRefInfo;
+ SmallVector<OpAsmParser::OperandType, 4> srcIndexInfos;
+ OpAsmParser::OperandType dstMemRefInfo;
+ SmallVector<OpAsmParser::OperandType, 4> dstIndexInfos;
+ OpAsmParser::OperandType numElementsInfo;
+ OpAsmParser::OperandType tagMemrefInfo;
+ SmallVector<OpAsmParser::OperandType, 4> tagIndexInfos;
+ SmallVector<OpAsmParser::OperandType, 2> strideInfo;
+
+ SmallVector<Type, 3> types;
+ auto indexType = parser.getBuilder().getIndexType();
+
+ // Parse and resolve the following list of operands:
+ // *) source memref followed by its indices (in square brackets).
+ // *) destination memref followed by its indices (in square brackets).
+ // *) dma size in KiB.
+ if (parser.parseOperand(srcMemRefInfo) ||
+ parser.parseOperandList(srcIndexInfos, OpAsmParser::Delimiter::Square) ||
+ parser.parseComma() || parser.parseOperand(dstMemRefInfo) ||
+ parser.parseOperandList(dstIndexInfos, OpAsmParser::Delimiter::Square) ||
+ parser.parseComma() || parser.parseOperand(numElementsInfo) ||
+ parser.parseComma() || parser.parseOperand(tagMemrefInfo) ||
+ parser.parseOperandList(tagIndexInfos, OpAsmParser::Delimiter::Square))
+ return failure();
+
+ // Parse optional stride and elements per stride.
+ if (parser.parseTrailingOperandList(strideInfo))
+ return failure();
+
+ bool isStrided = strideInfo.size() == 2;
+ if (!strideInfo.empty() && !isStrided) {
+ return parser.emitError(parser.getNameLoc(),
+ "expected two stride related operands");
+ }
+
+ if (parser.parseColonTypeList(types))
+ return failure();
+ if (types.size() != 3)
+ return parser.emitError(parser.getNameLoc(), "fewer/more types expected");
+
+ if (parser.resolveOperand(srcMemRefInfo, types[0], result.operands) ||
+ parser.resolveOperands(srcIndexInfos, indexType, result.operands) ||
+ parser.resolveOperand(dstMemRefInfo, types[1], result.operands) ||
+ parser.resolveOperands(dstIndexInfos, indexType, result.operands) ||
+ // size should be an index.
+ parser.resolveOperand(numElementsInfo, indexType, result.operands) ||
+ parser.resolveOperand(tagMemrefInfo, types[2], result.operands) ||
+ // tag indices should be index.
+ parser.resolveOperands(tagIndexInfos, indexType, result.operands))
+ return failure();
+
+ if (isStrided) {
+ if (parser.resolveOperands(strideInfo, indexType, result.operands))
+ return failure();
+ }
+
+ return success();
+}
+
+LogicalResult DmaStartOp::verify() {
+ unsigned numOperands = getNumOperands();
+
+ // Mandatory non-variadic operands are: src memref, dst memref, tag memref and
+ // the number of elements.
+ if (numOperands < 4)
+ return emitOpError("expected at least 4 operands");
+
+ // Check types of operands. The order of these calls is important: the later
+ // calls rely on some type properties to compute the operand position.
+ // 1. Source memref.
+ if (!getSrcMemRef().getType().isa<MemRefType>())
+ return emitOpError("expected source to be of memref type");
+ if (numOperands < getSrcMemRefRank() + 4)
+ return emitOpError() << "expected at least " << getSrcMemRefRank() + 4
+ << " operands";
+ if (!getSrcIndices().empty() &&
+ !llvm::all_of(getSrcIndices().getTypes(),
+ [](Type t) { return t.isIndex(); }))
+ return emitOpError("expected source indices to be of index type");
+
+ // 2. Destination memref.
+ if (!getDstMemRef().getType().isa<MemRefType>())
+ return emitOpError("expected destination to be of memref type");
+ unsigned numExpectedOperands = getSrcMemRefRank() + getDstMemRefRank() + 4;
+ if (numOperands < numExpectedOperands)
+ return emitOpError() << "expected at least " << numExpectedOperands
+ << " operands";
+ if (!getDstIndices().empty() &&
+ !llvm::all_of(getDstIndices().getTypes(),
+ [](Type t) { return t.isIndex(); }))
+ return emitOpError("expected destination indices to be of index type");
+
+ // 3. Number of elements.
+ if (!getNumElements().getType().isIndex())
+ return emitOpError("expected num elements to be of index type");
+
+ // 4. Tag memref.
+ if (!getTagMemRef().getType().isa<MemRefType>())
+ return emitOpError("expected tag to be of memref type");
+ numExpectedOperands += getTagMemRefRank();
+ if (numOperands < numExpectedOperands)
+ return emitOpError() << "expected at least " << numExpectedOperands
+ << " operands";
+ if (!getTagIndices().empty() &&
+ !llvm::all_of(getTagIndices().getTypes(),
+ [](Type t) { return t.isIndex(); }))
+ return emitOpError("expected tag indices to be of index type");
+
+ // DMAs from
diff erent memory spaces supported.
+ if (getSrcMemorySpace() == getDstMemorySpace())
+ return emitOpError("DMA should be between
diff erent memory spaces");
+
+ // Optional stride-related operands must be either both present or both
+ // absent.
+ if (numOperands != numExpectedOperands &&
+ numOperands != numExpectedOperands + 2)
+ return emitOpError("incorrect number of operands");
+
+ // 5. Strides.
+ if (isStrided()) {
+ if (!getStride().getType().isIndex() ||
+ !getNumElementsPerStride().getType().isIndex())
+ return emitOpError(
+ "expected stride and num elements per stride to be of type index");
+ }
+
+ return success();
+}
+
+LogicalResult DmaStartOp::fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results) {
+ /// dma_start(memrefcast) -> dma_start
+ return foldMemRefCast(*this);
+}
+
+// ---------------------------------------------------------------------------
+// DmaWaitOp
+// ---------------------------------------------------------------------------
+
+void DmaWaitOp::build(OpBuilder &builder, OperationState &result,
+ Value tagMemRef, ValueRange tagIndices,
+ Value numElements) {
+ result.addOperands(tagMemRef);
+ result.addOperands(tagIndices);
+ result.addOperands(numElements);
+}
+
+void DmaWaitOp::print(OpAsmPrinter &p) {
+ p << getOperationName() << " " << getTagMemRef() << '[' << getTagIndices()
+ << "], " << getNumElements();
+ p.printOptionalAttrDict((*this)->getAttrs());
+ p << " : " << getTagMemRef().getType();
+}
+
+// Parse DmaWaitOp.
+// Eg:
+// dma_wait %tag[%index], %num_elements : memref<1 x i32, (d0) -> (d0), 4>
+//
+ParseResult DmaWaitOp::parse(OpAsmParser &parser, OperationState &result) {
+ OpAsmParser::OperandType tagMemrefInfo;
+ SmallVector<OpAsmParser::OperandType, 2> tagIndexInfos;
+ Type type;
+ auto indexType = parser.getBuilder().getIndexType();
+ OpAsmParser::OperandType numElementsInfo;
+
+ // Parse tag memref, its indices, and dma size.
+ if (parser.parseOperand(tagMemrefInfo) ||
+ parser.parseOperandList(tagIndexInfos, OpAsmParser::Delimiter::Square) ||
+ parser.parseComma() || parser.parseOperand(numElementsInfo) ||
+ parser.parseColonType(type) ||
+ parser.resolveOperand(tagMemrefInfo, type, result.operands) ||
+ parser.resolveOperands(tagIndexInfos, indexType, result.operands) ||
+ parser.resolveOperand(numElementsInfo, indexType, result.operands))
+ return failure();
+
+ return success();
+}
+
+LogicalResult DmaWaitOp::fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results) {
+ /// dma_wait(memrefcast) -> dma_wait
+ return foldMemRefCast(*this);
+}
+
+LogicalResult DmaWaitOp::verify() {
+ // Mandatory non-variadic operands are tag and the number of elements.
+ if (getNumOperands() < 2)
+ return emitOpError() << "expected at least 2 operands";
+
+ // Check types of operands. The order of these calls is important: the later
+ // calls rely on some type properties to compute the operand position.
+ if (!getTagMemRef().getType().isa<MemRefType>())
+ return emitOpError() << "expected tag to be of memref type";
+
+ if (getNumOperands() != 2 + getTagMemRefRank())
+ return emitOpError() << "expected " << 2 + getTagMemRefRank()
+ << " operands";
+
+ if (!getTagIndices().empty() &&
+ !llvm::all_of(getTagIndices().getTypes(),
+ [](Type t) { return t.isIndex(); }))
+ return emitOpError() << "expected tag indices to be of index type";
+
+ if (!getNumElements().getType().isIndex())
+ return emitOpError()
+ << "expected the number of elements to be of index type";
+
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// GlobalOp
+//===----------------------------------------------------------------------===//
+
+static void printGlobalMemrefOpTypeAndInitialValue(OpAsmPrinter &p, GlobalOp op,
+ TypeAttr type,
+ Attribute initialValue) {
+ p << type;
+ if (!op.isExternal()) {
+ p << " = ";
+ if (op.isUninitialized())
+ p << "uninitialized";
+ else
+ p.printAttributeWithoutType(initialValue);
+ }
+}
+
+static ParseResult
+parseGlobalMemrefOpTypeAndInitialValue(OpAsmParser &parser, TypeAttr &typeAttr,
+ Attribute &initialValue) {
+ Type type;
+ if (parser.parseType(type))
+ return failure();
+
+ auto memrefType = type.dyn_cast<MemRefType>();
+ if (!memrefType || !memrefType.hasStaticShape())
+ return parser.emitError(parser.getNameLoc())
+ << "type should be static shaped memref, but got " << type;
+ typeAttr = TypeAttr::get(type);
+
+ if (parser.parseOptionalEqual())
+ return success();
+
+ if (succeeded(parser.parseOptionalKeyword("uninitialized"))) {
+ initialValue = UnitAttr::get(parser.getBuilder().getContext());
+ return success();
+ }
+
+ Type tensorType = getTensorTypeFromMemRefType(memrefType);
+ if (parser.parseAttribute(initialValue, tensorType))
+ return failure();
+ if (!initialValue.isa<ElementsAttr>())
+ return parser.emitError(parser.getNameLoc())
+ << "initial value should be a unit or elements attribute";
+ return success();
+}
+
+static LogicalResult verify(GlobalOp op) {
+ auto memrefType = op.type().dyn_cast<MemRefType>();
+ if (!memrefType || !memrefType.hasStaticShape())
+ return op.emitOpError("type should be static shaped memref, but got ")
+ << op.type();
+
+ // Verify that the initial value, if present, is either a unit attribute or
+ // an elements attribute.
+ if (op.initial_value().hasValue()) {
+ Attribute initValue = op.initial_value().getValue();
+ if (!initValue.isa<UnitAttr>() && !initValue.isa<ElementsAttr>())
+ return op.emitOpError("initial value should be a unit or elements "
+ "attribute, but got ")
+ << initValue;
+
+ // Check that the type of the initial value is compatible with the type of
+ // the global variable.
+ if (initValue.isa<ElementsAttr>()) {
+ Type initType = initValue.getType();
+ Type tensorType = getTensorTypeFromMemRefType(memrefType);
+ if (initType != tensorType)
+ return op.emitOpError("initial value expected to be of type ")
+ << tensorType << ", but was of type " << initType;
+ }
+ }
+
+ // TODO: verify visibility for declarations.
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// GetGlobalOp
+//===----------------------------------------------------------------------===//
+
+LogicalResult
+GetGlobalOp::verifySymbolUses(SymbolTableCollection &symbolTable) {
+ // Verify that the result type is same as the type of the referenced
+ // memref.global op.
+ auto global =
+ symbolTable.lookupNearestSymbolFrom<GlobalOp>(*this, nameAttr());
+ if (!global)
+ return emitOpError("'")
+ << name() << "' does not reference a valid global memref";
+
+ Type resultType = result().getType();
+ if (global.type() != resultType)
+ return emitOpError("result type ")
+ << resultType << " does not match type " << global.type()
+ << " of the global memref @" << name();
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// LoadOp
+//===----------------------------------------------------------------------===//
+
+static LogicalResult verify(LoadOp op) {
+ if (op.getNumOperands() != 1 + op.getMemRefType().getRank())
+ return op.emitOpError("incorrect number of indices for load");
+ return success();
+}
+
+OpFoldResult LoadOp::fold(ArrayRef<Attribute> cstOperands) {
+ /// load(memrefcast) -> load
+ if (succeeded(foldMemRefCast(*this)))
+ return getResult();
+ return OpFoldResult();
+}
+
+namespace {
+/// Fold a load on a buffer_cast operation into an tensor.extract on the
+/// corresponding tensor.
+struct LoadOfBufferCast : public OpRewritePattern<LoadOp> {
+ using OpRewritePattern<LoadOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(LoadOp load,
+ PatternRewriter &rewriter) const override {
+ auto buffercast = load.memref().getDefiningOp<BufferCastOp>();
+ if (!buffercast)
+ return failure();
+
+ rewriter.replaceOpWithNewOp<tensor::ExtractOp>(load, buffercast.tensor(),
+ load.indices());
+ return success();
+ }
+};
+} // end anonymous namespace.
+
+void LoadOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<LoadOfBufferCast>(context);
+}
+
+//===----------------------------------------------------------------------===//
+// PrefetchOp
+//===----------------------------------------------------------------------===//
+
+static void print(OpAsmPrinter &p, PrefetchOp op) {
+ p << PrefetchOp::getOperationName() << " " << op.memref() << '[';
+ p.printOperands(op.indices());
+ p << ']' << ", " << (op.isWrite() ? "write" : "read");
+ p << ", locality<" << op.localityHint();
+ p << ">, " << (op.isDataCache() ? "data" : "instr");
+ p.printOptionalAttrDict(
+ op->getAttrs(),
+ /*elidedAttrs=*/{"localityHint", "isWrite", "isDataCache"});
+ p << " : " << op.getMemRefType();
+}
+
+static ParseResult parsePrefetchOp(OpAsmParser &parser,
+ OperationState &result) {
+ OpAsmParser::OperandType memrefInfo;
+ SmallVector<OpAsmParser::OperandType, 4> indexInfo;
+ IntegerAttr localityHint;
+ MemRefType type;
+ StringRef readOrWrite, cacheType;
+
+ auto indexTy = parser.getBuilder().getIndexType();
+ auto i32Type = parser.getBuilder().getIntegerType(32);
+ if (parser.parseOperand(memrefInfo) ||
+ parser.parseOperandList(indexInfo, OpAsmParser::Delimiter::Square) ||
+ parser.parseComma() || parser.parseKeyword(&readOrWrite) ||
+ parser.parseComma() || parser.parseKeyword("locality") ||
+ parser.parseLess() ||
+ parser.parseAttribute(localityHint, i32Type, "localityHint",
+ result.attributes) ||
+ parser.parseGreater() || parser.parseComma() ||
+ parser.parseKeyword(&cacheType) || parser.parseColonType(type) ||
+ parser.resolveOperand(memrefInfo, type, result.operands) ||
+ parser.resolveOperands(indexInfo, indexTy, result.operands))
+ return failure();
+
+ if (!readOrWrite.equals("read") && !readOrWrite.equals("write"))
+ return parser.emitError(parser.getNameLoc(),
+ "rw specifier has to be 'read' or 'write'");
+ result.addAttribute(
+ PrefetchOp::getIsWriteAttrName(),
+ parser.getBuilder().getBoolAttr(readOrWrite.equals("write")));
+
+ if (!cacheType.equals("data") && !cacheType.equals("instr"))
+ return parser.emitError(parser.getNameLoc(),
+ "cache type has to be 'data' or 'instr'");
+
+ result.addAttribute(
+ PrefetchOp::getIsDataCacheAttrName(),
+ parser.getBuilder().getBoolAttr(cacheType.equals("data")));
+
+ return success();
+}
+
+static LogicalResult verify(PrefetchOp op) {
+ if (op.getNumOperands() != 1 + op.getMemRefType().getRank())
+ return op.emitOpError("too few indices");
+
+ return success();
+}
+
+LogicalResult PrefetchOp::fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results) {
+ // prefetch(memrefcast) -> prefetch
+ return foldMemRefCast(*this);
+}
+
+//===----------------------------------------------------------------------===//
+// ReinterpretCastOp
+//===----------------------------------------------------------------------===//
+
+/// Build a ReinterpretCastOp with all dynamic entries: `staticOffsets`,
+/// `staticSizes` and `staticStrides` are automatically filled with
+/// source-memref-rank sentinel values that encode dynamic entries.
+void ReinterpretCastOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source,
+ OpFoldResult offset, ArrayRef<OpFoldResult> sizes,
+ ArrayRef<OpFoldResult> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
+ SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
+ dispatchIndexOpFoldResults(offset, dynamicOffsets, staticOffsets,
+ ShapedType::kDynamicStrideOrOffset);
+ dispatchIndexOpFoldResults(sizes, dynamicSizes, staticSizes,
+ ShapedType::kDynamicSize);
+ dispatchIndexOpFoldResults(strides, dynamicStrides, staticStrides,
+ ShapedType::kDynamicStrideOrOffset);
+ build(b, result, resultType, source, dynamicOffsets, dynamicSizes,
+ dynamicStrides, b.getI64ArrayAttr(staticOffsets),
+ b.getI64ArrayAttr(staticSizes), b.getI64ArrayAttr(staticStrides));
+ result.addAttributes(attrs);
+}
+
+void ReinterpretCastOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source,
+ int64_t offset, ArrayRef<int64_t> sizes,
+ ArrayRef<int64_t> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<OpFoldResult> sizeValues =
+ llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
+ llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ build(b, result, resultType, source, b.getI64IntegerAttr(offset), sizeValues,
+ strideValues, attrs);
+}
+
+void ReinterpretCastOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source, Value offset,
+ ValueRange sizes, ValueRange strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<OpFoldResult> sizeValues = llvm::to_vector<4>(
+ llvm::map_range(sizes, [](Value v) -> OpFoldResult { return v; }));
+ SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
+ llvm::map_range(strides, [](Value v) -> OpFoldResult { return v; }));
+ build(b, result, resultType, source, offset, sizeValues, strideValues, attrs);
+}
+
+// TODO: ponder whether we want to allow missing trailing sizes/strides that are
+// completed automatically, like we have for subview and subtensor.
+static LogicalResult verify(ReinterpretCastOp op) {
+ // The source and result memrefs should be in the same memory space.
+ auto srcType = op.source().getType().cast<BaseMemRefType>();
+ auto resultType = op.getType().cast<MemRefType>();
+ if (srcType.getMemorySpaceAsInt() != resultType.getMemorySpaceAsInt())
+ return op.emitError("
diff erent memory spaces specified for source type ")
+ << srcType << " and result memref type " << resultType;
+ if (srcType.getElementType() != resultType.getElementType())
+ return op.emitError("
diff erent element types specified for source type ")
+ << srcType << " and result memref type " << resultType;
+
+ // Match sizes in result memref type and in static_sizes attribute.
+ for (auto &en :
+ llvm::enumerate(llvm::zip(resultType.getShape(),
+ extractFromI64ArrayAttr(op.static_sizes())))) {
+ int64_t resultSize = std::get<0>(en.value());
+ int64_t expectedSize = std::get<1>(en.value());
+ if (resultSize != expectedSize)
+ return op.emitError("expected result type with size = ")
+ << expectedSize << " instead of " << resultSize
+ << " in dim = " << en.index();
+ }
+
+ // Match offset and strides in static_offset and static_strides attributes if
+ // result memref type has an affine map specified.
+ if (!resultType.getAffineMaps().empty()) {
+ int64_t resultOffset;
+ SmallVector<int64_t, 4> resultStrides;
+ if (failed(getStridesAndOffset(resultType, resultStrides, resultOffset)))
+ return failure();
+
+ // Match offset in result memref type and in static_offsets attribute.
+ int64_t expectedOffset =
+ extractFromI64ArrayAttr(op.static_offsets()).front();
+ if (resultOffset != expectedOffset)
+ return op.emitError("expected result type with offset = ")
+ << resultOffset << " instead of " << expectedOffset;
+
+ // Match strides in result memref type and in static_strides attribute.
+ for (auto &en : llvm::enumerate(llvm::zip(
+ resultStrides, extractFromI64ArrayAttr(op.static_strides())))) {
+ int64_t resultStride = std::get<0>(en.value());
+ int64_t expectedStride = std::get<1>(en.value());
+ if (resultStride != expectedStride)
+ return op.emitError("expected result type with stride = ")
+ << expectedStride << " instead of " << resultStride
+ << " in dim = " << en.index();
+ }
+ }
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// ReshapeOp
+//===----------------------------------------------------------------------===//
+
+static LogicalResult verify(ReshapeOp op) {
+ Type operandType = op.source().getType();
+ Type resultType = op.result().getType();
+
+ Type operandElementType = operandType.cast<ShapedType>().getElementType();
+ Type resultElementType = resultType.cast<ShapedType>().getElementType();
+ if (operandElementType != resultElementType)
+ return op.emitOpError("element types of source and destination memref "
+ "types should be the same");
+
+ if (auto operandMemRefType = operandType.dyn_cast<MemRefType>())
+ if (!operandMemRefType.getAffineMaps().empty())
+ return op.emitOpError(
+ "source memref type should have identity affine map");
+
+ int64_t shapeSize = op.shape().getType().cast<MemRefType>().getDimSize(0);
+ auto resultMemRefType = resultType.dyn_cast<MemRefType>();
+ if (resultMemRefType) {
+ if (!resultMemRefType.getAffineMaps().empty())
+ return op.emitOpError(
+ "result memref type should have identity affine map");
+ if (shapeSize == ShapedType::kDynamicSize)
+ return op.emitOpError("cannot use shape operand with dynamic length to "
+ "reshape to statically-ranked memref type");
+ if (shapeSize != resultMemRefType.getRank())
+ return op.emitOpError(
+ "length of shape operand
diff ers from the result's memref rank");
+ }
+ return success();
+}
+
+//===----------------------------------------------------------------------===//
+// StoreOp
+//===----------------------------------------------------------------------===//
+
+static LogicalResult verify(StoreOp op) {
+ if (op.getNumOperands() != 2 + op.getMemRefType().getRank())
+ return op.emitOpError("store index operand count not equal to memref rank");
+
+ return success();
+}
+
+LogicalResult StoreOp::fold(ArrayRef<Attribute> cstOperands,
+ SmallVectorImpl<OpFoldResult> &results) {
+ /// store(memrefcast) -> store
+ return foldMemRefCast(*this);
+}
+
+//===----------------------------------------------------------------------===//
+// SubViewOp
+//===----------------------------------------------------------------------===//
+
+namespace {
+/// Helpers to write more idiomatic operations.
+namespace saturated_arith {
+struct Wrapper {
+ explicit Wrapper(int64_t v) : v(v) {}
+ operator int64_t() { return v; }
+ int64_t v;
+};
+Wrapper operator+(Wrapper a, int64_t b) {
+ if (ShapedType::isDynamicStrideOrOffset(a) ||
+ ShapedType::isDynamicStrideOrOffset(b))
+ return Wrapper(ShapedType::kDynamicStrideOrOffset);
+ return Wrapper(a.v + b);
+}
+Wrapper operator*(Wrapper a, int64_t b) {
+ if (ShapedType::isDynamicStrideOrOffset(a) ||
+ ShapedType::isDynamicStrideOrOffset(b))
+ return Wrapper(ShapedType::kDynamicStrideOrOffset);
+ return Wrapper(a.v * b);
+}
+} // end namespace saturated_arith
+} // end namespace
+
+/// A subview result type can be fully inferred from the source type and the
+/// static representation of offsets, sizes and strides. Special sentinels
+/// encode the dynamic case.
+Type SubViewOp::inferResultType(MemRefType sourceMemRefType,
+ ArrayRef<int64_t> leadingStaticOffsets,
+ ArrayRef<int64_t> leadingStaticSizes,
+ ArrayRef<int64_t> leadingStaticStrides) {
+ // A subview may specify only a leading subset of offset/sizes/strides in
+ // which case we complete with offset=0, sizes from memref type and strides=1.
+ unsigned rank = sourceMemRefType.getRank();
+ assert(leadingStaticOffsets.size() <= rank &&
+ "unexpected leadingStaticOffsets overflow");
+ assert(leadingStaticSizes.size() <= rank &&
+ "unexpected leadingStaticSizes overflow");
+ assert(leadingStaticStrides.size() <= rank &&
+ "unexpected leadingStaticStrides overflow");
+ auto staticOffsets = llvm::to_vector<4>(leadingStaticOffsets);
+ auto staticSizes = llvm::to_vector<4>(leadingStaticSizes);
+ auto staticStrides = llvm::to_vector<4>(leadingStaticStrides);
+ unsigned numTrailingOffsets = rank - staticOffsets.size();
+ unsigned numTrailingSizes = rank - staticSizes.size();
+ unsigned numTrailingStrides = rank - staticStrides.size();
+ staticOffsets.append(numTrailingOffsets, 0);
+ llvm::append_range(staticSizes,
+ sourceMemRefType.getShape().take_back(numTrailingSizes));
+ staticStrides.append(numTrailingStrides, 1);
+
+ // Extract source offset and strides.
+ int64_t sourceOffset;
+ SmallVector<int64_t, 4> sourceStrides;
+ auto res = getStridesAndOffset(sourceMemRefType, sourceStrides, sourceOffset);
+ assert(succeeded(res) && "SubViewOp expected strided memref type");
+ (void)res;
+
+ // Compute target offset whose value is:
+ // `sourceOffset + sum_i(staticOffset_i * sourceStrides_i)`.
+ int64_t targetOffset = sourceOffset;
+ for (auto it : llvm::zip(staticOffsets, sourceStrides)) {
+ auto staticOffset = std::get<0>(it), targetStride = std::get<1>(it);
+ using namespace saturated_arith;
+ targetOffset = Wrapper(targetOffset) + Wrapper(staticOffset) * targetStride;
+ }
+
+ // Compute target stride whose value is:
+ // `sourceStrides_i * staticStrides_i`.
+ SmallVector<int64_t, 4> targetStrides;
+ targetStrides.reserve(staticOffsets.size());
+ for (auto it : llvm::zip(sourceStrides, staticStrides)) {
+ auto sourceStride = std::get<0>(it), staticStride = std::get<1>(it);
+ using namespace saturated_arith;
+ targetStrides.push_back(Wrapper(sourceStride) * staticStride);
+ }
+
+ // The type is now known.
+ return MemRefType::get(
+ staticSizes, sourceMemRefType.getElementType(),
+ makeStridedLinearLayoutMap(targetStrides, targetOffset,
+ sourceMemRefType.getContext()),
+ sourceMemRefType.getMemorySpaceAsInt());
+}
+
+Type SubViewOp::inferResultType(MemRefType sourceMemRefType,
+ ArrayRef<OpFoldResult> leadingStaticOffsets,
+ ArrayRef<OpFoldResult> leadingStaticSizes,
+ ArrayRef<OpFoldResult> leadingStaticStrides) {
+ SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
+ SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
+ dispatchIndexOpFoldResults(leadingStaticOffsets, dynamicOffsets,
+ staticOffsets, ShapedType::kDynamicStrideOrOffset);
+ dispatchIndexOpFoldResults(leadingStaticSizes, dynamicSizes, staticSizes,
+ ShapedType::kDynamicSize);
+ dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
+ staticStrides, ShapedType::kDynamicStrideOrOffset);
+ return SubViewOp::inferResultType(sourceMemRefType, staticOffsets,
+ staticSizes, staticStrides)
+ .cast<MemRefType>();
+}
+
+Type SubViewOp::inferRankReducedResultType(
+ unsigned resultRank, MemRefType sourceRankedTensorType,
+ ArrayRef<int64_t> leadingStaticOffsets,
+ ArrayRef<int64_t> leadingStaticSizes,
+ ArrayRef<int64_t> leadingStaticStrides) {
+ auto inferredType =
+ inferResultType(sourceRankedTensorType, leadingStaticOffsets,
+ leadingStaticSizes, leadingStaticStrides)
+ .cast<MemRefType>();
+ assert(inferredType.getRank() >= resultRank && "expected ");
+ int rankDiff = inferredType.getRank() - resultRank;
+ if (rankDiff > 0) {
+ auto shape = inferredType.getShape();
+ llvm::SmallDenseSet<unsigned> dimsToProject;
+ mlir::getPositionsOfShapeOne(rankDiff, shape, dimsToProject);
+ SmallVector<int64_t> projectedShape;
+ for (unsigned pos = 0, e = shape.size(); pos < e; ++pos)
+ if (!dimsToProject.contains(pos))
+ projectedShape.push_back(shape[pos]);
+
+ AffineMap map;
+ auto maps = inferredType.getAffineMaps();
+ if (!maps.empty() && maps.front())
+ map = getProjectedMap(maps.front(), dimsToProject);
+ inferredType =
+ MemRefType::get(projectedShape, inferredType.getElementType(), map,
+ inferredType.getMemorySpaceAsInt());
+ }
+ return inferredType;
+}
+
+Type SubViewOp::inferRankReducedResultType(
+ unsigned resultRank, MemRefType sourceRankedTensorType,
+ ArrayRef<OpFoldResult> leadingStaticOffsets,
+ ArrayRef<OpFoldResult> leadingStaticSizes,
+ ArrayRef<OpFoldResult> leadingStaticStrides) {
+ SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
+ SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
+ dispatchIndexOpFoldResults(leadingStaticOffsets, dynamicOffsets,
+ staticOffsets, ShapedType::kDynamicStrideOrOffset);
+ dispatchIndexOpFoldResults(leadingStaticSizes, dynamicSizes, staticSizes,
+ ShapedType::kDynamicSize);
+ dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
+ staticStrides, ShapedType::kDynamicStrideOrOffset);
+ return SubViewOp::inferRankReducedResultType(
+ resultRank, sourceRankedTensorType, staticOffsets, staticSizes,
+ staticStrides);
+}
+// Build a SubViewOp with mixed static and dynamic entries and custom result
+// type. If the type passed is nullptr, it is inferred.
+void SubViewOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source,
+ ArrayRef<OpFoldResult> offsets,
+ ArrayRef<OpFoldResult> sizes,
+ ArrayRef<OpFoldResult> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
+ SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
+ dispatchIndexOpFoldResults(offsets, dynamicOffsets, staticOffsets,
+ ShapedType::kDynamicStrideOrOffset);
+ dispatchIndexOpFoldResults(sizes, dynamicSizes, staticSizes,
+ ShapedType::kDynamicSize);
+ dispatchIndexOpFoldResults(strides, dynamicStrides, staticStrides,
+ ShapedType::kDynamicStrideOrOffset);
+ auto sourceMemRefType = source.getType().cast<MemRefType>();
+ // Structuring implementation this way avoids duplication between builders.
+ if (!resultType) {
+ resultType = SubViewOp::inferResultType(sourceMemRefType, staticOffsets,
+ staticSizes, staticStrides)
+ .cast<MemRefType>();
+ }
+ build(b, result, resultType, source, dynamicOffsets, dynamicSizes,
+ dynamicStrides, b.getI64ArrayAttr(staticOffsets),
+ b.getI64ArrayAttr(staticSizes), b.getI64ArrayAttr(staticStrides));
+ result.addAttributes(attrs);
+}
+
+// Build a SubViewOp with mixed static and dynamic entries and inferred result
+// type.
+void SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
+ ArrayRef<OpFoldResult> offsets,
+ ArrayRef<OpFoldResult> sizes,
+ ArrayRef<OpFoldResult> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ build(b, result, MemRefType(), source, offsets, sizes, strides, attrs);
+}
+
+// Build a SubViewOp with static entries and inferred result type.
+void SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
+ ArrayRef<int64_t> offsets, ArrayRef<int64_t> sizes,
+ ArrayRef<int64_t> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
+ llvm::map_range(offsets, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ SmallVector<OpFoldResult> sizeValues =
+ llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
+ llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ build(b, result, source, offsetValues, sizeValues, strideValues, attrs);
+}
+
+// Build a SubViewOp with dynamic entries and custom result type. If the
+// type passed is nullptr, it is inferred.
+void SubViewOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source,
+ ArrayRef<int64_t> offsets, ArrayRef<int64_t> sizes,
+ ArrayRef<int64_t> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
+ llvm::map_range(offsets, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ SmallVector<OpFoldResult> sizeValues =
+ llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
+ llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
+ return b.getI64IntegerAttr(v);
+ }));
+ build(b, result, resultType, source, offsetValues, sizeValues, strideValues,
+ attrs);
+}
+
+// Build a SubViewOp with dynamic entries and custom result type. If the type
+// passed is nullptr, it is inferred.
+void SubViewOp::build(OpBuilder &b, OperationState &result,
+ MemRefType resultType, Value source, ValueRange offsets,
+ ValueRange sizes, ValueRange strides,
+ ArrayRef<NamedAttribute> attrs) {
+ SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
+ llvm::map_range(offsets, [](Value v) -> OpFoldResult { return v; }));
+ SmallVector<OpFoldResult> sizeValues = llvm::to_vector<4>(
+ llvm::map_range(sizes, [](Value v) -> OpFoldResult { return v; }));
+ SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
+ llvm::map_range(strides, [](Value v) -> OpFoldResult { return v; }));
+ build(b, result, resultType, source, offsetValues, sizeValues, strideValues);
+}
+
+// Build a SubViewOp with dynamic entries and inferred result type.
+void SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
+ ValueRange offsets, ValueRange sizes, ValueRange strides,
+ ArrayRef<NamedAttribute> attrs) {
+ build(b, result, MemRefType(), source, offsets, sizes, strides, attrs);
+}
+
+/// For ViewLikeOpInterface.
+Value SubViewOp::getViewSource() { return source(); }
+
+/// Given an `originalShape` and a `reducedShape` assumed to be a subset of
+/// `originalShape` with some `1` entries erased, return the set of indices
+/// that specifies which of the entries of `originalShape` are dropped to obtain
+/// `reducedShape`. The returned mask can be applied as a projection to
+/// `originalShape` to obtain the `reducedShape`. This mask is useful to track
+/// which dimensions must be kept when e.g. compute MemRef strides under
+/// rank-reducing operations. Return None if reducedShape cannot be obtained
+/// by dropping only `1` entries in `originalShape`.
+llvm::Optional<llvm::SmallDenseSet<unsigned>>
+mlir::computeRankReductionMask(ArrayRef<int64_t> originalShape,
+ ArrayRef<int64_t> reducedShape) {
+ size_t originalRank = originalShape.size(), reducedRank = reducedShape.size();
+ llvm::SmallDenseSet<unsigned> unusedDims;
+ unsigned reducedIdx = 0;
+ for (unsigned originalIdx = 0; originalIdx < originalRank; ++originalIdx) {
+ // Greedily insert `originalIdx` if no match.
+ if (reducedIdx < reducedRank &&
+ originalShape[originalIdx] == reducedShape[reducedIdx]) {
+ reducedIdx++;
+ continue;
+ }
+
+ unusedDims.insert(originalIdx);
+ // If no match on `originalIdx`, the `originalShape` at this dimension
+ // must be 1, otherwise we bail.
+ if (originalShape[originalIdx] != 1)
+ return llvm::None;
+ }
+ // The whole reducedShape must be scanned, otherwise we bail.
+ if (reducedIdx != reducedRank)
+ return llvm::None;
+ return unusedDims;
+}
+
+enum SubViewVerificationResult {
+ Success,
+ RankTooLarge,
+ SizeMismatch,
+ ElemTypeMismatch,
+ MemSpaceMismatch,
+ AffineMapMismatch
+};
+
+/// Checks if `original` Type type can be rank reduced to `reduced` type.
+/// This function is slight variant of `is subsequence` algorithm where
+/// not matching dimension must be 1.
+static SubViewVerificationResult
+isRankReducedType(Type originalType, Type candidateReducedType,
+ std::string *errMsg = nullptr) {
+ if (originalType == candidateReducedType)
+ return SubViewVerificationResult::Success;
+ if (!originalType.isa<MemRefType>())
+ return SubViewVerificationResult::Success;
+ if (originalType.isa<MemRefType>() && !candidateReducedType.isa<MemRefType>())
+ return SubViewVerificationResult::Success;
+
+ ShapedType originalShapedType = originalType.cast<ShapedType>();
+ ShapedType candidateReducedShapedType =
+ candidateReducedType.cast<ShapedType>();
+
+ // Rank and size logic is valid for all ShapedTypes.
+ ArrayRef<int64_t> originalShape = originalShapedType.getShape();
+ ArrayRef<int64_t> candidateReducedShape =
+ candidateReducedShapedType.getShape();
+ unsigned originalRank = originalShape.size(),
+ candidateReducedRank = candidateReducedShape.size();
+ if (candidateReducedRank > originalRank)
+ return SubViewVerificationResult::RankTooLarge;
+
+ auto optionalUnusedDimsMask =
+ computeRankReductionMask(originalShape, candidateReducedShape);
+
+ // Sizes cannot be matched in case empty vector is returned.
+ if (!optionalUnusedDimsMask.hasValue())
+ return SubViewVerificationResult::SizeMismatch;
+
+ if (originalShapedType.getElementType() !=
+ candidateReducedShapedType.getElementType())
+ return SubViewVerificationResult::ElemTypeMismatch;
+
+ // Strided layout logic is relevant for MemRefType only.
+ MemRefType original = originalType.cast<MemRefType>();
+ MemRefType candidateReduced = candidateReducedType.cast<MemRefType>();
+ if (original.getMemorySpaceAsInt() != candidateReduced.getMemorySpaceAsInt())
+ return SubViewVerificationResult::MemSpaceMismatch;
+
+ llvm::SmallDenseSet<unsigned> unusedDims = optionalUnusedDimsMask.getValue();
+ auto inferredType =
+ getProjectedMap(getStridedLinearLayoutMap(original), unusedDims);
+ AffineMap candidateLayout;
+ if (candidateReduced.getAffineMaps().empty())
+ candidateLayout = getStridedLinearLayoutMap(candidateReduced);
+ else
+ candidateLayout = candidateReduced.getAffineMaps().front();
+ assert(inferredType.getNumResults() == 1 &&
+ candidateLayout.getNumResults() == 1);
+ if (inferredType.getNumSymbols() != candidateLayout.getNumSymbols() ||
+ inferredType.getNumDims() != candidateLayout.getNumDims()) {
+ if (errMsg) {
+ llvm::raw_string_ostream os(*errMsg);
+ os << "inferred type: " << inferredType;
+ }
+ return SubViewVerificationResult::AffineMapMismatch;
+ }
+ // Check that the
diff erence of the affine maps simplifies to 0.
+ AffineExpr
diff Expr =
+ inferredType.getResult(0) - candidateLayout.getResult(0);
+
diff Expr = simplifyAffineExpr(
diff Expr, inferredType.getNumDims(),
+ inferredType.getNumSymbols());
+ auto cst =
diff Expr.dyn_cast<AffineConstantExpr>();
+ if (!(cst && cst.getValue() == 0)) {
+ if (errMsg) {
+ llvm::raw_string_ostream os(*errMsg);
+ os << "inferred type: " << inferredType;
+ }
+ return SubViewVerificationResult::AffineMapMismatch;
+ }
+ return SubViewVerificationResult::Success;
+}
+
+template <typename OpTy>
+static LogicalResult produceSubViewErrorMsg(SubViewVerificationResult result,
+ OpTy op, Type expectedType,
+ StringRef errMsg = "") {
+ auto memrefType = expectedType.cast<ShapedType>();
+ switch (result) {
+ case SubViewVerificationResult::Success:
+ return success();
+ case SubViewVerificationResult::RankTooLarge:
+ return op.emitError("expected result rank to be smaller or equal to ")
+ << "the source rank. " << errMsg;
+ case SubViewVerificationResult::SizeMismatch:
+ return op.emitError("expected result type to be ")
+ << expectedType
+ << " or a rank-reduced version. (mismatch of result sizes) "
+ << errMsg;
+ case SubViewVerificationResult::ElemTypeMismatch:
+ return op.emitError("expected result element type to be ")
+ << memrefType.getElementType() << errMsg;
+ case SubViewVerificationResult::MemSpaceMismatch:
+ return op.emitError("expected result and source memory spaces to match.")
+ << errMsg;
+ case SubViewVerificationResult::AffineMapMismatch:
+ return op.emitError("expected result type to be ")
+ << expectedType
+ << " or a rank-reduced version. (mismatch of result affine map) "
+ << errMsg;
+ }
+ llvm_unreachable("unexpected subview verification result");
+}
+
+/// Verifier for SubViewOp.
+static LogicalResult verify(SubViewOp op) {
+ MemRefType baseType = op.getSourceType();
+ MemRefType subViewType = op.getType();
+
+ // The base memref and the view memref should be in the same memory space.
+ if (baseType.getMemorySpaceAsInt() != subViewType.getMemorySpaceAsInt())
+ return op.emitError("
diff erent memory spaces specified for base memref "
+ "type ")
+ << baseType << " and subview memref type " << subViewType;
+
+ // Verify that the base memref type has a strided layout map.
+ if (!isStrided(baseType))
+ return op.emitError("base type ") << baseType << " is not strided";
+
+ // Verify result type against inferred type.
+ auto expectedType = SubViewOp::inferResultType(
+ baseType, extractFromI64ArrayAttr(op.static_offsets()),
+ extractFromI64ArrayAttr(op.static_sizes()),
+ extractFromI64ArrayAttr(op.static_strides()));
+
+ std::string errMsg;
+ auto result = isRankReducedType(expectedType, subViewType, &errMsg);
+ return produceSubViewErrorMsg(result, op, expectedType, errMsg);
+}
+
+raw_ostream &mlir::operator<<(raw_ostream &os, Range &range) {
+ return os << "range " << range.offset << ":" << range.size << ":"
+ << range.stride;
+}
+
+/// Return the list of Range (i.e. offset, size, stride). Each Range
+/// entry contains either the dynamic value or a ConstantIndexOp constructed
+/// with `b` at location `loc`.
+SmallVector<Range, 8> mlir::getOrCreateRanges(OffsetSizeAndStrideOpInterface op,
+ OpBuilder &b, Location loc) {
+ std::array<unsigned, 3> ranks = op.getArrayAttrMaxRanks();
+ assert(ranks[0] == ranks[1] && "expected offset and sizes of equal ranks");
+ assert(ranks[1] == ranks[2] && "expected sizes and strides of equal ranks");
+ SmallVector<Range, 8> res;
+ unsigned rank = ranks[0];
+ res.reserve(rank);
+ for (unsigned idx = 0; idx < rank; ++idx) {
+ Value offset =
+ op.isDynamicOffset(idx)
+ ? op.getDynamicOffset(idx)
+ : b.create<ConstantIndexOp>(loc, op.getStaticOffset(idx));
+ Value size = op.isDynamicSize(idx)
+ ? op.getDynamicSize(idx)
+ : b.create<ConstantIndexOp>(loc, op.getStaticSize(idx));
+ Value stride =
+ op.isDynamicStride(idx)
+ ? op.getDynamicStride(idx)
+ : b.create<ConstantIndexOp>(loc, op.getStaticStride(idx));
+ res.emplace_back(Range{offset, size, stride});
+ }
+ return res;
+}
+
+namespace {
+/// Pattern to rewrite a subview op with MemRefCast arguments.
+/// This essentially pushes memref.cast past its consuming subview when
+/// `canFoldIntoConsumerOp` is true.
+///
+/// Example:
+/// ```
+/// %0 = memref.cast %V : memref<16x16xf32> to memref<?x?xf32>
+/// %1 = memref.subview %0[0, 0][3, 4][1, 1] :
+/// memref<?x?xf32> to memref<3x4xf32, offset:?, strides:[?, 1]>
+/// ```
+/// is rewritten into:
+/// ```
+/// %0 = memref.subview %V: memref<16x16xf32> to memref<3x4xf32, #[[map0]]>
+/// %1 = memref.cast %0: memref<3x4xf32, offset:0, strides:[16, 1]> to
+/// memref<3x4xf32, offset:?, strides:[?, 1]>
+/// ```
+class SubViewOpMemRefCastFolder final : public OpRewritePattern<SubViewOp> {
+public:
+ using OpRewritePattern<SubViewOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(SubViewOp subViewOp,
+ PatternRewriter &rewriter) const override {
+ // Any constant operand, just return to let SubViewOpConstantFolder kick in.
+ if (llvm::any_of(subViewOp.getOperands(), [](Value operand) {
+ return matchPattern(operand, matchConstantIndex());
+ }))
+ return failure();
+
+ auto castOp = subViewOp.source().getDefiningOp<CastOp>();
+ if (!castOp)
+ return failure();
+
+ if (!CastOp::canFoldIntoConsumerOp(castOp))
+ return failure();
+
+ /// Deduce the resultType of the SubViewOp using `inferSubViewResultType` on
+ /// the cast source operand type and the SubViewOp static information. This
+ /// is the resulting type if the MemRefCastOp were folded.
+ auto resultType = SubViewOp::inferRankReducedResultType(
+ subViewOp.getType().getRank(),
+ castOp.source().getType().cast<MemRefType>(),
+ subViewOp.getMixedOffsets(), subViewOp.getMixedSizes(),
+ subViewOp.getMixedStrides());
+ Value newSubView = rewriter.create<SubViewOp>(
+ subViewOp.getLoc(), resultType, castOp.source(), subViewOp.offsets(),
+ subViewOp.sizes(), subViewOp.strides(), subViewOp.static_offsets(),
+ subViewOp.static_sizes(), subViewOp.static_strides());
+ rewriter.replaceOpWithNewOp<CastOp>(subViewOp, subViewOp.getType(),
+ newSubView);
+ return success();
+ }
+};
+} // namespace
+
+/// A canonicalizer wrapper to replace SubViewOps.
+struct SubViewCanonicalizer {
+ void operator()(PatternRewriter &rewriter, SubViewOp op, SubViewOp newOp) {
+ rewriter.replaceOpWithNewOp<CastOp>(op, newOp, op.getType());
+ }
+};
+
+void SubViewOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<OpWithOffsetSizesAndStridesConstantArgumentFolder<
+ SubViewOp, SubViewCanonicalizer>,
+ SubViewOpMemRefCastFolder>(context);
+}
+
+OpFoldResult SubViewOp::fold(ArrayRef<Attribute> operands) {
+ auto resultShapedType = getResult().getType().cast<ShapedType>();
+ auto sourceShapedType = source().getType().cast<ShapedType>();
+
+ if (resultShapedType.hasStaticShape() &&
+ resultShapedType == sourceShapedType) {
+ return getViewSource();
+ }
+
+ return {};
+}
+
+//===----------------------------------------------------------------------===//
+// TensorLoadOp
+//===----------------------------------------------------------------------===//
+
+OpFoldResult TensorLoadOp::fold(ArrayRef<Attribute>) {
+ if (auto bufferCast = memref().getDefiningOp<BufferCastOp>())
+ // Approximate alias analysis by conservatively folding only when no there
+ // is no interleaved operation.
+ if (bufferCast->getBlock() == this->getOperation()->getBlock() &&
+ bufferCast->getNextNode() == this->getOperation())
+ return bufferCast.tensor();
+ return {};
+}
+
+//===----------------------------------------------------------------------===//
+// TransposeOp
+//===----------------------------------------------------------------------===//
+
+/// Build a strided memref type by applying `permutationMap` tp `memRefType`.
+static MemRefType inferTransposeResultType(MemRefType memRefType,
+ AffineMap permutationMap) {
+ auto rank = memRefType.getRank();
+ auto originalSizes = memRefType.getShape();
+ // Compute permuted sizes.
+ SmallVector<int64_t, 4> sizes(rank, 0);
+ for (auto en : llvm::enumerate(permutationMap.getResults()))
+ sizes[en.index()] =
+ originalSizes[en.value().cast<AffineDimExpr>().getPosition()];
+
+ // Compute permuted strides.
+ int64_t offset;
+ SmallVector<int64_t, 4> strides;
+ auto res = getStridesAndOffset(memRefType, strides, offset);
+ assert(succeeded(res) && strides.size() == static_cast<unsigned>(rank));
+ (void)res;
+ auto map =
+ makeStridedLinearLayoutMap(strides, offset, memRefType.getContext());
+ map = permutationMap ? map.compose(permutationMap) : map;
+ return MemRefType::Builder(memRefType).setShape(sizes).setAffineMaps(map);
+}
+
+void TransposeOp::build(OpBuilder &b, OperationState &result, Value in,
+ AffineMapAttr permutation,
+ ArrayRef<NamedAttribute> attrs) {
+ auto permutationMap = permutation.getValue();
+ assert(permutationMap);
+
+ auto memRefType = in.getType().cast<MemRefType>();
+ // Compute result type.
+ MemRefType resultType = inferTransposeResultType(memRefType, permutationMap);
+
+ build(b, result, resultType, in, attrs);
+ result.addAttribute(TransposeOp::getPermutationAttrName(), permutation);
+}
+
+// transpose $in $permutation attr-dict : type($in) `to` type(results)
+static void print(OpAsmPrinter &p, TransposeOp op) {
+ p << "memref.transpose " << op.in() << " " << op.permutation();
+ p.printOptionalAttrDict(op->getAttrs(),
+ {TransposeOp::getPermutationAttrName()});
+ p << " : " << op.in().getType() << " to " << op.getType();
+}
+
+static ParseResult parseTransposeOp(OpAsmParser &parser,
+ OperationState &result) {
+ OpAsmParser::OperandType in;
+ AffineMap permutation;
+ MemRefType srcType, dstType;
+ if (parser.parseOperand(in) || parser.parseAffineMap(permutation) ||
+ parser.parseOptionalAttrDict(result.attributes) ||
+ parser.parseColonType(srcType) ||
+ parser.resolveOperand(in, srcType, result.operands) ||
+ parser.parseKeywordType("to", dstType) ||
+ parser.addTypeToList(dstType, result.types))
+ return failure();
+
+ result.addAttribute(TransposeOp::getPermutationAttrName(),
+ AffineMapAttr::get(permutation));
+ return success();
+}
+
+static LogicalResult verify(TransposeOp op) {
+ if (!op.permutation().isPermutation())
+ return op.emitOpError("expected a permutation map");
+ if (op.permutation().getNumDims() != op.getShapedType().getRank())
+ return op.emitOpError(
+ "expected a permutation map of same rank as the input");
+
+ auto srcType = op.in().getType().cast<MemRefType>();
+ auto dstType = op.getType().cast<MemRefType>();
+ auto transposedType = inferTransposeResultType(srcType, op.permutation());
+ if (dstType != transposedType)
+ return op.emitOpError("output type ")
+ << dstType << " does not match transposed input type " << srcType
+ << ", " << transposedType;
+ return success();
+}
+
+OpFoldResult TransposeOp::fold(ArrayRef<Attribute>) {
+ if (succeeded(foldMemRefCast(*this)))
+ return getResult();
+ return {};
+}
+
+//===----------------------------------------------------------------------===//
+// ViewOp
+//===----------------------------------------------------------------------===//
+
+static ParseResult parseViewOp(OpAsmParser &parser, OperationState &result) {
+ OpAsmParser::OperandType srcInfo;
+ SmallVector<OpAsmParser::OperandType, 1> offsetInfo;
+ SmallVector<OpAsmParser::OperandType, 4> sizesInfo;
+ auto indexType = parser.getBuilder().getIndexType();
+ Type srcType, dstType;
+ llvm::SMLoc offsetLoc;
+ if (parser.parseOperand(srcInfo) || parser.getCurrentLocation(&offsetLoc) ||
+ parser.parseOperandList(offsetInfo, OpAsmParser::Delimiter::Square))
+ return failure();
+
+ if (offsetInfo.size() != 1)
+ return parser.emitError(offsetLoc) << "expects 1 offset operand";
+
+ return failure(
+ parser.parseOperandList(sizesInfo, OpAsmParser::Delimiter::Square) ||
+ parser.parseOptionalAttrDict(result.attributes) ||
+ parser.parseColonType(srcType) ||
+ parser.resolveOperand(srcInfo, srcType, result.operands) ||
+ parser.resolveOperands(offsetInfo, indexType, result.operands) ||
+ parser.resolveOperands(sizesInfo, indexType, result.operands) ||
+ parser.parseKeywordType("to", dstType) ||
+ parser.addTypeToList(dstType, result.types));
+}
+
+static void print(OpAsmPrinter &p, ViewOp op) {
+ p << op.getOperationName() << ' ' << op.getOperand(0) << '[';
+ p.printOperand(op.byte_shift());
+ p << "][" << op.sizes() << ']';
+ p.printOptionalAttrDict(op->getAttrs());
+ p << " : " << op.getOperand(0).getType() << " to " << op.getType();
+}
+
+static LogicalResult verify(ViewOp op) {
+ auto baseType = op.getOperand(0).getType().cast<MemRefType>();
+ auto viewType = op.getType();
+
+ // The base memref should have identity layout map (or none).
+ if (baseType.getAffineMaps().size() > 1 ||
+ (baseType.getAffineMaps().size() == 1 &&
+ !baseType.getAffineMaps()[0].isIdentity()))
+ return op.emitError("unsupported map for base memref type ") << baseType;
+
+ // The result memref should have identity layout map (or none).
+ if (viewType.getAffineMaps().size() > 1 ||
+ (viewType.getAffineMaps().size() == 1 &&
+ !viewType.getAffineMaps()[0].isIdentity()))
+ return op.emitError("unsupported map for result memref type ") << viewType;
+
+ // The base memref and the view memref should be in the same memory space.
+ if (baseType.getMemorySpaceAsInt() != viewType.getMemorySpaceAsInt())
+ return op.emitError("
diff erent memory spaces specified for base memref "
+ "type ")
+ << baseType << " and view memref type " << viewType;
+
+ // Verify that we have the correct number of sizes for the result type.
+ unsigned numDynamicDims = viewType.getNumDynamicDims();
+ if (op.sizes().size() != numDynamicDims)
+ return op.emitError("incorrect number of size operands for type ")
+ << viewType;
+
+ return success();
+}
+
+Value ViewOp::getViewSource() { return source(); }
+
+namespace {
+
+struct ViewOpShapeFolder : public OpRewritePattern<ViewOp> {
+ using OpRewritePattern<ViewOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(ViewOp viewOp,
+ PatternRewriter &rewriter) const override {
+ // Return if none of the operands are constants.
+ if (llvm::none_of(viewOp.getOperands(), [](Value operand) {
+ return matchPattern(operand, matchConstantIndex());
+ }))
+ return failure();
+
+ // Get result memref type.
+ auto memrefType = viewOp.getType();
+
+ // Get offset from old memref view type 'memRefType'.
+ int64_t oldOffset;
+ SmallVector<int64_t, 4> oldStrides;
+ if (failed(getStridesAndOffset(memrefType, oldStrides, oldOffset)))
+ return failure();
+ assert(oldOffset == 0 && "Expected 0 offset");
+
+ SmallVector<Value, 4> newOperands;
+
+ // Offset cannot be folded into result type.
+
+ // Fold any dynamic dim operands which are produced by a constant.
+ SmallVector<int64_t, 4> newShapeConstants;
+ newShapeConstants.reserve(memrefType.getRank());
+
+ unsigned dynamicDimPos = 0;
+ unsigned rank = memrefType.getRank();
+ for (unsigned dim = 0, e = rank; dim < e; ++dim) {
+ int64_t dimSize = memrefType.getDimSize(dim);
+ // If this is already static dimension, keep it.
+ if (!ShapedType::isDynamic(dimSize)) {
+ newShapeConstants.push_back(dimSize);
+ continue;
+ }
+ auto *defOp = viewOp.sizes()[dynamicDimPos].getDefiningOp();
+ if (auto constantIndexOp = dyn_cast_or_null<ConstantIndexOp>(defOp)) {
+ // Dynamic shape dimension will be folded.
+ newShapeConstants.push_back(constantIndexOp.getValue());
+ } else {
+ // Dynamic shape dimension not folded; copy operand from old memref.
+ newShapeConstants.push_back(dimSize);
+ newOperands.push_back(viewOp.sizes()[dynamicDimPos]);
+ }
+ dynamicDimPos++;
+ }
+
+ // Create new memref type with constant folded dims.
+ MemRefType newMemRefType =
+ MemRefType::Builder(memrefType).setShape(newShapeConstants);
+ // Nothing new, don't fold.
+ if (newMemRefType == memrefType)
+ return failure();
+
+ // Create new ViewOp.
+ auto newViewOp = rewriter.create<ViewOp>(viewOp.getLoc(), newMemRefType,
+ viewOp.getOperand(0),
+ viewOp.byte_shift(), newOperands);
+ // Insert a cast so we have the same type as the old memref type.
+ rewriter.replaceOpWithNewOp<CastOp>(viewOp, newViewOp, viewOp.getType());
+ return success();
+ }
+};
+
+struct ViewOpMemrefCastFolder : public OpRewritePattern<ViewOp> {
+ using OpRewritePattern<ViewOp>::OpRewritePattern;
+
+ LogicalResult matchAndRewrite(ViewOp viewOp,
+ PatternRewriter &rewriter) const override {
+ Value memrefOperand = viewOp.getOperand(0);
+ CastOp memrefCastOp = memrefOperand.getDefiningOp<CastOp>();
+ if (!memrefCastOp)
+ return failure();
+ Value allocOperand = memrefCastOp.getOperand();
+ AllocOp allocOp = allocOperand.getDefiningOp<AllocOp>();
+ if (!allocOp)
+ return failure();
+ rewriter.replaceOpWithNewOp<ViewOp>(viewOp, viewOp.getType(), allocOperand,
+ viewOp.byte_shift(), viewOp.sizes());
+ return success();
+ }
+};
+
+} // end anonymous namespace
+
+void ViewOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
+ MLIRContext *context) {
+ results.insert<ViewOpShapeFolder, ViewOpMemrefCastFolder>(context);
+}
+
+//===----------------------------------------------------------------------===//
+// TableGen'd op method definitions
+//===----------------------------------------------------------------------===//
+
+#define GET_OP_CLASSES
+#include "mlir/Dialect/MemRef/IR/MemRefOps.cpp.inc"
diff --git a/mlir/lib/Dialect/SCF/CMakeLists.txt b/mlir/lib/Dialect/SCF/CMakeLists.txt
index 297e918cb6ab..e3e8f96a62e6 100644
--- a/mlir/lib/Dialect/SCF/CMakeLists.txt
+++ b/mlir/lib/Dialect/SCF/CMakeLists.txt
@@ -12,6 +12,7 @@ add_mlir_dialect_library(MLIRSCF
MLIREDSC
MLIRIR
MLIRLoopLikeInterface
+ MLIRMemRef
MLIRSideEffectInterfaces
MLIRStandard
)
diff --git a/mlir/lib/Dialect/SCF/SCF.cpp b/mlir/lib/Dialect/SCF/SCF.cpp
index 57315754a910..9c0df1b47c35 100644
--- a/mlir/lib/Dialect/SCF/SCF.cpp
+++ b/mlir/lib/Dialect/SCF/SCF.cpp
@@ -7,6 +7,7 @@
//===----------------------------------------------------------------------===//
#include "mlir/Dialect/SCF/SCF.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BlockAndValueMapping.h"
#include "mlir/IR/PatternMatch.h"
@@ -568,7 +569,7 @@ struct SimplifyTrivialLoops : public OpRewritePattern<ForOp> {
/// %t0 = ... : tensor_type
/// %0 = scf.for ... iter_args(%bb0 : %t0) -> (tensor_type) {
/// ...
-/// // %m is either tensor_to_memref(%bb00) or defined above the loop
+/// // %m is either buffer_cast(%bb00) or defined above the loop
/// %m... : memref_type
/// ... // uses of %m with potential inplace updates
/// %new_tensor = tensor_load %m : memref_type
@@ -578,7 +579,7 @@ struct SimplifyTrivialLoops : public OpRewritePattern<ForOp> {
/// ```
///
/// `%bb0` may have either 0 or 1 use. If it has 1 use it must be exactly a
-/// `%m = tensor_to_memref %bb0` op that feeds into the yielded `tensor_load`
+/// `%m = buffer_cast %bb0` op that feeds into the yielded `tensor_load`
/// op.
///
/// If no aliasing write to the memref `%m`, from which `%new_tensor`is loaded,
@@ -590,7 +591,7 @@ struct SimplifyTrivialLoops : public OpRewritePattern<ForOp> {
///
/// The canonicalization rewrites the pattern as:
/// ```
-/// // %m is either a tensor_to_memref or defined above
+/// // %m is either a buffer_cast or defined above
/// %m... : memref_type
/// scf.for ... iter_args(%bb0 : %t0) -> (tensor_type) {
/// ... // uses of %m with potential inplace updates
@@ -601,7 +602,7 @@ struct SimplifyTrivialLoops : public OpRewritePattern<ForOp> {
///
/// A later bbArg canonicalization will further rewrite as:
/// ```
-/// // %m is either a tensor_to_memref or defined above
+/// // %m is either a buffer_cast or defined above
/// %m... : memref_type
/// scf.for ... { // no iter_args
/// ... // uses of %m with potential inplace updates
@@ -622,19 +623,18 @@ struct LastTensorLoadCanonicalization : public OpRewritePattern<ForOp> {
unsigned idx = bbArg.getArgNumber() - /*numIv=*/1;
auto yieldOp = cast<scf::YieldOp>(forOp.region().front().getTerminator());
Value yieldVal = yieldOp->getOperand(idx);
- auto tensorLoadOp = yieldVal.getDefiningOp<TensorLoadOp>();
+ auto tensorLoadOp = yieldVal.getDefiningOp<memref::TensorLoadOp>();
bool isTensor = bbArg.getType().isa<TensorType>();
- TensorToMemrefOp tensorToMemRefOp;
- // Either bbArg has no use or it has a single tensor_to_memref use.
+ memref::BufferCastOp bufferCastOp;
+ // Either bbArg has no use or it has a single buffer_cast use.
if (bbArg.hasOneUse())
- tensorToMemRefOp =
- dyn_cast<TensorToMemrefOp>(*bbArg.getUsers().begin());
- if (!isTensor || !tensorLoadOp ||
- (!bbArg.use_empty() && !tensorToMemRefOp))
+ bufferCastOp =
+ dyn_cast<memref::BufferCastOp>(*bbArg.getUsers().begin());
+ if (!isTensor || !tensorLoadOp || (!bbArg.use_empty() && !bufferCastOp))
continue;
- // If tensorToMemRefOp is present, it must feed into the `tensorLoadOp`.
- if (tensorToMemRefOp && tensorLoadOp.memref() != tensorToMemRefOp)
+ // If bufferCastOp is present, it must feed into the `tensorLoadOp`.
+ if (bufferCastOp && tensorLoadOp.memref() != bufferCastOp)
continue;
// TODO: Any aliasing write of tensorLoadOp.memref() nested under `forOp`
// must be before `tensorLoadOp` in the block so that the lastWrite
@@ -644,18 +644,18 @@ struct LastTensorLoadCanonicalization : public OpRewritePattern<ForOp> {
if (tensorLoadOp->getNextNode() != yieldOp)
continue;
- // Clone the optional tensorToMemRefOp before forOp.
- if (tensorToMemRefOp) {
+ // Clone the optional bufferCastOp before forOp.
+ if (bufferCastOp) {
rewriter.setInsertionPoint(forOp);
- rewriter.replaceOpWithNewOp<TensorToMemrefOp>(
- tensorToMemRefOp, tensorToMemRefOp.memref().getType(),
- tensorToMemRefOp.tensor());
+ rewriter.replaceOpWithNewOp<memref::BufferCastOp>(
+ bufferCastOp, bufferCastOp.memref().getType(),
+ bufferCastOp.tensor());
}
// Clone the tensorLoad after forOp.
rewriter.setInsertionPointAfter(forOp);
Value newTensorLoad =
- rewriter.create<TensorLoadOp>(loc, tensorLoadOp.memref());
+ rewriter.create<memref::TensorLoadOp>(loc, tensorLoadOp.memref());
Value forOpResult = forOp.getResult(bbArg.getArgNumber() - /*iv=*/1);
replacements.insert(std::make_pair(forOpResult, newTensorLoad));
diff --git a/mlir/lib/Dialect/SCF/Transforms/Bufferize.cpp b/mlir/lib/Dialect/SCF/Transforms/Bufferize.cpp
index 7cf0dfabd917..aa25f47e4801 100644
--- a/mlir/lib/Dialect/SCF/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/SCF/Transforms/Bufferize.cpp
@@ -8,6 +8,7 @@
#include "mlir/Transforms/Bufferize.h"
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/Passes.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/SCF/Transforms.h"
diff --git a/mlir/lib/Dialect/SCF/Transforms/CMakeLists.txt b/mlir/lib/Dialect/SCF/Transforms/CMakeLists.txt
index bd36cf758650..3c44e1d0472d 100644
--- a/mlir/lib/Dialect/SCF/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/SCF/Transforms/CMakeLists.txt
@@ -15,6 +15,7 @@ add_mlir_dialect_library(MLIRSCFTransforms
LINK_LIBS PUBLIC
MLIRAffine
MLIRIR
+ MLIRMemRef
MLIRPass
MLIRSCF
MLIRStandard
diff --git a/mlir/lib/Dialect/SCF/Transforms/ParallelLoopFusion.cpp b/mlir/lib/Dialect/SCF/Transforms/ParallelLoopFusion.cpp
index f2fc78e4290e..32c5abc2decf 100644
--- a/mlir/lib/Dialect/SCF/Transforms/ParallelLoopFusion.cpp
+++ b/mlir/lib/Dialect/SCF/Transforms/ParallelLoopFusion.cpp
@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/Passes.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/SCF/Transforms.h"
@@ -52,10 +53,10 @@ static bool haveNoReadsAfterWriteExceptSameIndex(
ParallelOp firstPloop, ParallelOp secondPloop,
const BlockAndValueMapping &firstToSecondPloopIndices) {
DenseMap<Value, SmallVector<ValueRange, 1>> bufferStores;
- firstPloop.getBody()->walk([&](StoreOp store) {
+ firstPloop.getBody()->walk([&](memref::StoreOp store) {
bufferStores[store.getMemRef()].push_back(store.indices());
});
- auto walkResult = secondPloop.getBody()->walk([&](LoadOp load) {
+ auto walkResult = secondPloop.getBody()->walk([&](memref::LoadOp load) {
// Stop if the memref is defined in secondPloop body. Careful alias analysis
// is needed.
auto *memrefDef = load.getMemRef().getDefiningOp();
diff --git a/mlir/lib/Dialect/SCF/Transforms/PassDetail.h b/mlir/lib/Dialect/SCF/Transforms/PassDetail.h
index 6fa7f227d3da..2e27ed2c229c 100644
--- a/mlir/lib/Dialect/SCF/Transforms/PassDetail.h
+++ b/mlir/lib/Dialect/SCF/Transforms/PassDetail.h
@@ -18,6 +18,10 @@ void registerDialect(DialectRegistry ®istry);
class AffineDialect;
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
#define GEN_PASS_CLASSES
#include "mlir/Dialect/SCF/Passes.h.inc"
diff --git a/mlir/lib/Dialect/Shape/IR/Shape.cpp b/mlir/lib/Dialect/Shape/IR/Shape.cpp
index 472197c52f4e..ed8dcfc13549 100644
--- a/mlir/lib/Dialect/Shape/IR/Shape.cpp
+++ b/mlir/lib/Dialect/Shape/IR/Shape.cpp
@@ -407,7 +407,7 @@ struct RemoveDuplicateOperandsPattern : public OpRewritePattern<OpTy> {
// Reduce op to equivalent with unique operands.
if (unique.size() < op.getNumOperands()) {
rewriter.replaceOpWithNewOp<OpTy>(op, op->getResultTypes(), unique,
- op.getAttrs());
+ op->getAttrs());
return success();
}
diff --git a/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
index 1b51f330be66..36b5eac501c3 100644
--- a/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
@@ -8,6 +8,7 @@
#include "mlir/Transforms/Bufferize.h"
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/Shape/Transforms/Passes.h"
#include "mlir/Pass/Pass.h"
diff --git a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
index 123a3664df89..646e5b98c90b 100644
--- a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
@@ -14,6 +14,7 @@ add_mlir_dialect_library(MLIRShapeOpsTransforms
target_link_libraries(MLIRShapeOpsTransforms
PUBLIC
MLIRIR
+ MLIRMemRef
MLIRPass
MLIRShape
MLIRSupport
diff --git a/mlir/lib/Dialect/Shape/Transforms/PassDetail.h b/mlir/lib/Dialect/Shape/Transforms/PassDetail.h
index abb5c21d66bb..c7796d4ba82c 100644
--- a/mlir/lib/Dialect/Shape/Transforms/PassDetail.h
+++ b/mlir/lib/Dialect/Shape/Transforms/PassDetail.h
@@ -13,6 +13,10 @@
namespace mlir {
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
#define GEN_PASS_CLASSES
#include "mlir/Dialect/Shape/Transforms/Passes.h.inc"
diff --git a/mlir/lib/Dialect/StandardOps/CMakeLists.txt b/mlir/lib/Dialect/StandardOps/CMakeLists.txt
index 058e680ef677..aa06559bf80f 100644
--- a/mlir/lib/Dialect/StandardOps/CMakeLists.txt
+++ b/mlir/lib/Dialect/StandardOps/CMakeLists.txt
@@ -16,6 +16,7 @@ add_mlir_dialect_library(MLIRStandard
MLIRControlFlowInterfaces
MLIREDSC
MLIRIR
+ MLIRMemRef
MLIRSideEffectInterfaces
MLIRTensor
MLIRVectorInterfaces
diff --git a/mlir/lib/Dialect/StandardOps/EDSC/Builders.cpp b/mlir/lib/Dialect/StandardOps/EDSC/Builders.cpp
index 30a1a8ca30ab..b6e6c9960338 100644
--- a/mlir/lib/Dialect/StandardOps/EDSC/Builders.cpp
+++ b/mlir/lib/Dialect/StandardOps/EDSC/Builders.cpp
@@ -6,6 +6,7 @@
//
//===----------------------------------------------------------------------===//
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/IR/AffineExpr.h"
#include "mlir/IR/AffineMap.h"
@@ -23,7 +24,7 @@ static SmallVector<Value, 8> getMemRefSizes(Value memRef) {
const auto &shape = memRefType.getShape();
for (unsigned idx = 0, n = shape.size(); idx < n; ++idx) {
if (shape[idx] == -1)
- res.push_back(std_dim(memRef, idx));
+ res.push_back(memref_dim(memRef, idx));
else
res.push_back(std_constant_index(shape[idx]));
}
diff --git a/mlir/lib/Dialect/StandardOps/IR/Ops.cpp b/mlir/lib/Dialect/StandardOps/IR/Ops.cpp
index 2302457bae57..bd38e154bcf6 100644
--- a/mlir/lib/Dialect/StandardOps/IR/Ops.cpp
+++ b/mlir/lib/Dialect/StandardOps/IR/Ops.cpp
@@ -9,6 +9,7 @@
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/CommonFolders.h"
+#include "mlir/Dialect/StandardOps/Utils/Utils.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
#include "mlir/IR/AffineExpr.h"
#include "mlir/IR/AffineMap.h"
@@ -216,10 +217,10 @@ static void printStandardCastOp(Operation *op, OpAsmPrinter &p) {
void StandardOpsDialect::initialize() {
getContext()->loadDialect<tensor::TensorDialect>();
- addOperations<DmaStartOp, DmaWaitOp,
+ addOperations<
#define GET_OP_LIST
#include "mlir/Dialect/StandardOps/IR/Ops.cpp.inc"
- >();
+ >();
addInterfaces<StdInlinerInterface>();
}
@@ -231,32 +232,6 @@ Operation *StandardOpsDialect::materializeConstant(OpBuilder &builder,
return builder.create<ConstantOp>(loc, type, value);
}
-/// Matches a ConstantIndexOp.
-/// TODO: This should probably just be a general matcher that uses m_Constant
-/// and checks the operation for an index type.
-static detail::op_matcher<ConstantIndexOp> m_ConstantIndex() {
- return detail::op_matcher<ConstantIndexOp>();
-}
-
-//===----------------------------------------------------------------------===//
-// Common canonicalization pattern support logic
-//===----------------------------------------------------------------------===//
-
-/// This is a common class used for patterns of the form
-/// "someop(memrefcast) -> someop". It folds the source of any memref_cast
-/// into the root operation directly.
-static LogicalResult foldMemRefCast(Operation *op) {
- bool folded = false;
- for (OpOperand &operand : op->getOpOperands()) {
- auto cast = operand.get().getDefiningOp<MemRefCastOp>();
- if (cast && !cast.getOperand().getType().isa<UnrankedMemRefType>()) {
- operand.set(cast.getOperand());
- folded = true;
- }
- }
- return success(folded);
-}
-
//===----------------------------------------------------------------------===//
// Common cast compatibility check for vector types.
//===----------------------------------------------------------------------===//
@@ -278,18 +253,6 @@ static bool areVectorCastSimpleCompatible(
return false;
}
-//===----------------------------------------------------------------------===//
-// Helpers for Tensor[Load|Store]Op, TensorToMemrefOp, and GlobalMemrefOp
-//===----------------------------------------------------------------------===//
-
-static Type getTensorTypeFromMemRefType(Type type) {
- if (auto memref = type.dyn_cast<MemRefType>())
- return RankedTensorType::get(memref.getShape(), memref.getElementType());
- if (auto memref = type.dyn_cast<UnrankedMemRefType>())
- return UnrankedTensorType::get(memref.getElementType());
- return NoneType::get(type.getContext());
-}
-
//===----------------------------------------------------------------------===//
// AddFOp
//===----------------------------------------------------------------------===//
@@ -320,131 +283,6 @@ static SmallVector<int64_t, 4> extractFromI64ArrayAttr(Attribute attr) {
}));
}
-//===----------------------------------------------------------------------===//
-// AllocOp / AllocaOp
-//===----------------------------------------------------------------------===//
-
-template <typename AllocLikeOp>
-static LogicalResult verifyAllocLikeOp(AllocLikeOp op) {
- static_assert(llvm::is_one_of<AllocLikeOp, AllocOp, AllocaOp>::value,
- "applies to only alloc or alloca");
- auto memRefType = op.getResult().getType().template dyn_cast<MemRefType>();
- if (!memRefType)
- return op.emitOpError("result must be a memref");
-
- if (static_cast<int64_t>(op.dynamicSizes().size()) !=
- memRefType.getNumDynamicDims())
- return op.emitOpError("dimension operand count does not equal memref "
- "dynamic dimension count");
-
- unsigned numSymbols = 0;
- if (!memRefType.getAffineMaps().empty())
- numSymbols = memRefType.getAffineMaps().front().getNumSymbols();
- if (op.symbolOperands().size() != numSymbols)
- return op.emitOpError(
- "symbol operand count does not equal memref symbol count");
-
- return success();
-}
-
-static LogicalResult verify(AllocOp op) { return verifyAllocLikeOp(op); }
-
-static LogicalResult verify(AllocaOp op) {
- // An alloca op needs to have an ancestor with an allocation scope trait.
- if (!op->getParentWithTrait<OpTrait::AutomaticAllocationScope>())
- return op.emitOpError(
- "requires an ancestor op with AutomaticAllocationScope trait");
-
- return verifyAllocLikeOp(op);
-}
-
-namespace {
-/// Fold constant dimensions into an alloc like operation.
-template <typename AllocLikeOp>
-struct SimplifyAllocConst : public OpRewritePattern<AllocLikeOp> {
- using OpRewritePattern<AllocLikeOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(AllocLikeOp alloc,
- PatternRewriter &rewriter) const override {
- // Check to see if any dimensions operands are constants. If so, we can
- // substitute and drop them.
- if (llvm::none_of(alloc.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
- }))
- return failure();
-
- auto memrefType = alloc.getType();
-
- // Ok, we have one or more constant operands. Collect the non-constant ones
- // and keep track of the resultant memref type to build.
- SmallVector<int64_t, 4> newShapeConstants;
- newShapeConstants.reserve(memrefType.getRank());
- SmallVector<Value, 4> newOperands;
-
- unsigned dynamicDimPos = 0;
- for (unsigned dim = 0, e = memrefType.getRank(); dim < e; ++dim) {
- int64_t dimSize = memrefType.getDimSize(dim);
- // If this is already static dimension, keep it.
- if (dimSize != -1) {
- newShapeConstants.push_back(dimSize);
- continue;
- }
- auto *defOp = alloc.getOperand(dynamicDimPos).getDefiningOp();
- if (auto constantIndexOp = dyn_cast_or_null<ConstantIndexOp>(defOp)) {
- // Dynamic shape dimension will be folded.
- newShapeConstants.push_back(constantIndexOp.getValue());
- } else {
- // Dynamic shape dimension not folded; copy operand from old memref.
- newShapeConstants.push_back(-1);
- newOperands.push_back(alloc.getOperand(dynamicDimPos));
- }
- dynamicDimPos++;
- }
-
- // Create new memref type (which will have fewer dynamic dimensions).
- MemRefType newMemRefType =
- MemRefType::Builder(memrefType).setShape(newShapeConstants);
- assert(static_cast<int64_t>(newOperands.size()) ==
- newMemRefType.getNumDynamicDims());
-
- // Create and insert the alloc op for the new memref.
- auto newAlloc = rewriter.create<AllocLikeOp>(alloc.getLoc(), newMemRefType,
- newOperands, IntegerAttr());
- // Insert a cast so we have the same type as the old alloc.
- auto resultCast = rewriter.create<MemRefCastOp>(alloc.getLoc(), newAlloc,
- alloc.getType());
-
- rewriter.replaceOp(alloc, {resultCast});
- return success();
- }
-};
-
-/// Fold alloc operations with no uses. Alloc has side effects on the heap,
-/// but can still be deleted if it has zero uses.
-struct SimplifyDeadAlloc : public OpRewritePattern<AllocOp> {
- using OpRewritePattern<AllocOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(AllocOp alloc,
- PatternRewriter &rewriter) const override {
- if (alloc.use_empty()) {
- rewriter.eraseOp(alloc);
- return success();
- }
- return failure();
- }
-};
-} // end anonymous namespace.
-
-void AllocOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<SimplifyAllocConst<AllocOp>, SimplifyDeadAlloc>(context);
-}
-
-void AllocaOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<SimplifyAllocConst<AllocaOp>>(context);
-}
-
//===----------------------------------------------------------------------===//
// AndOp
//===----------------------------------------------------------------------===//
@@ -491,17 +329,6 @@ void AssertOp::getCanonicalizationPatterns(OwningRewritePatternList &patterns,
patterns.insert<EraseRedundantAssertions>(context);
}
-//===----------------------------------------------------------------------===//
-// AssumeAlignmentOp
-//===----------------------------------------------------------------------===//
-
-static LogicalResult verify(AssumeAlignmentOp op) {
- unsigned alignment = op.alignment();
- if (!llvm::isPowerOf2_32(alignment))
- return op.emitOpError("alignment must be power of 2");
- return success();
-}
-
//===----------------------------------------------------------------------===//
// AtomicRMWOp
//===----------------------------------------------------------------------===//
@@ -1347,222 +1174,6 @@ void ConstantIndexOp::build(OpBuilder &builder, OperationState &result,
ConstantOp::build(builder, result, type, builder.getIntegerAttr(type, value));
}
-//===----------------------------------------------------------------------===//
-// DeallocOp
-//===----------------------------------------------------------------------===//
-namespace {
-/// Fold Dealloc operations that are deallocating an AllocOp that is only used
-/// by other Dealloc operations.
-struct SimplifyDeadDealloc : public OpRewritePattern<DeallocOp> {
- using OpRewritePattern<DeallocOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(DeallocOp dealloc,
- PatternRewriter &rewriter) const override {
- // Check that the memref operand's defining operation is an AllocOp.
- Value memref = dealloc.memref();
- if (!isa_and_nonnull<AllocOp>(memref.getDefiningOp()))
- return failure();
-
- // Check that all of the uses of the AllocOp are other DeallocOps.
- for (auto *user : memref.getUsers())
- if (!isa<DeallocOp>(user))
- return failure();
-
- // Erase the dealloc operation.
- rewriter.eraseOp(dealloc);
- return success();
- }
-};
-} // end anonymous namespace.
-
-static LogicalResult verify(DeallocOp op) {
- if (!op.memref().getType().isa<MemRefType>())
- return op.emitOpError("operand must be a memref");
- return success();
-}
-
-void DeallocOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<SimplifyDeadDealloc>(context);
-}
-
-LogicalResult DeallocOp::fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results) {
- /// dealloc(memrefcast) -> dealloc
- return foldMemRefCast(*this);
-}
-
-//===----------------------------------------------------------------------===//
-// DimOp
-//===----------------------------------------------------------------------===//
-
-void DimOp::build(OpBuilder &builder, OperationState &result,
- Value memrefOrTensor, int64_t index) {
- auto loc = result.location;
- Value indexValue = builder.create<ConstantIndexOp>(loc, index);
- build(builder, result, memrefOrTensor, indexValue);
-}
-
-void DimOp::build(OpBuilder &builder, OperationState &result,
- Value memrefOrTensor, Value index) {
- auto indexTy = builder.getIndexType();
- build(builder, result, indexTy, memrefOrTensor, index);
-}
-
-Optional<int64_t> DimOp::getConstantIndex() {
- if (auto constantOp = index().getDefiningOp<ConstantOp>())
- return constantOp.getValue().cast<IntegerAttr>().getInt();
- return {};
-}
-
-static LogicalResult verify(DimOp op) {
- // Assume unknown index to be in range.
- Optional<int64_t> index = op.getConstantIndex();
- if (!index.hasValue())
- return success();
-
- // Check that constant index is not knowingly out of range.
- auto type = op.memrefOrTensor().getType();
- if (auto tensorType = type.dyn_cast<RankedTensorType>()) {
- if (index.getValue() >= tensorType.getRank())
- return op.emitOpError("index is out of range");
- } else if (auto memrefType = type.dyn_cast<MemRefType>()) {
- if (index.getValue() >= memrefType.getRank())
- return op.emitOpError("index is out of range");
- } else if (type.isa<UnrankedTensorType>() || type.isa<UnrankedMemRefType>()) {
- // Assume index to be in range.
- } else {
- llvm_unreachable("expected operand with tensor or memref type");
- }
-
- return success();
-}
-
-OpFoldResult DimOp::fold(ArrayRef<Attribute> operands) {
- auto index = operands[1].dyn_cast_or_null<IntegerAttr>();
-
- // All forms of folding require a known index.
- if (!index)
- return {};
-
- auto argTy = memrefOrTensor().getType();
- // Fold if the shape extent along the given index is known.
- if (auto shapedTy = argTy.dyn_cast<ShapedType>()) {
- // Folding for unranked types (UnrankedMemRefType, UnrankedTensorType) is
- // not supported.
- if (!shapedTy.hasRank())
- return {};
- if (!shapedTy.isDynamicDim(index.getInt())) {
- Builder builder(getContext());
- return builder.getIndexAttr(shapedTy.getShape()[index.getInt()]);
- }
- }
-
- Operation *definingOp = memrefOrTensor().getDefiningOp();
- // dim(tensor_load(memref)) -> dim(memref)
- if (auto tensorLoadOp = dyn_cast_or_null<TensorLoadOp>(definingOp)) {
- setOperand(0, tensorLoadOp.memref());
- return getResult();
- }
-
- // Fold dim to the operand of tensor.generate.
- if (auto fromElements = dyn_cast_or_null<tensor::GenerateOp>(definingOp)) {
- auto resultType =
- fromElements.getResult().getType().cast<RankedTensorType>();
- // The case where the type encodes the size of the dimension is handled
- // above.
- assert(resultType.getShape()[index.getInt()] ==
- RankedTensorType::kDynamicSize);
-
- // Find the operand of the fromElements that corresponds to this index.
- auto dynExtents = fromElements.dynamicExtents().begin();
- for (auto dim : resultType.getShape().take_front(index.getInt()))
- if (dim == RankedTensorType::kDynamicSize)
- dynExtents++;
-
- return Value{*dynExtents};
- }
-
- // The size at the given index is now known to be a dynamic size.
- unsigned unsignedIndex = index.getValue().getZExtValue();
-
- if (auto subtensor = dyn_cast_or_null<SubTensorOp>(definingOp)) {
- assert(subtensor.isDynamicSize(unsignedIndex) &&
- "Expected dynamic subtensor size");
- return subtensor.getDynamicSize(unsignedIndex);
- }
-
- // Fold dim to the size argument for an `AllocOp`, `ViewOp`, or `SubViewOp`.
- auto memrefType = argTy.dyn_cast<MemRefType>();
- if (!memrefType)
- return {};
-
- if (auto alloc = dyn_cast_or_null<AllocOp>(definingOp))
- return *(alloc.getDynamicSizes().begin() +
- memrefType.getDynamicDimIndex(unsignedIndex));
-
- if (auto view = dyn_cast_or_null<ViewOp>(definingOp))
- return *(view.getDynamicSizes().begin() +
- memrefType.getDynamicDimIndex(unsignedIndex));
-
- if (auto subview = dyn_cast_or_null<SubViewOp>(definingOp)) {
- assert(subview.isDynamicSize(unsignedIndex) &&
- "Expected dynamic subview size");
- return subview.getDynamicSize(unsignedIndex);
- }
-
- // dim(memrefcast) -> dim
- if (succeeded(foldMemRefCast(*this)))
- return getResult();
-
- return {};
-}
-
-namespace {
-/// Fold dim of a memref reshape operation to a load into the reshape's shape
-/// operand.
-struct DimOfMemRefReshape : public OpRewritePattern<DimOp> {
- using OpRewritePattern<DimOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(DimOp dim,
- PatternRewriter &rewriter) const override {
- auto reshape = dim.memrefOrTensor().getDefiningOp<MemRefReshapeOp>();
-
- if (!reshape)
- return failure();
-
- // Place the load directly after the reshape to ensure that the shape memref
- // was not mutated.
- rewriter.setInsertionPointAfter(reshape);
- rewriter.replaceOpWithNewOp<LoadOp>(dim, reshape.shape(),
- llvm::makeArrayRef({dim.index()}));
- return success();
- }
-};
-
-/// Fold dim of a dim of a cast into the dim of the source of the tensor cast.
-template <typename CastOpTy>
-struct DimOfCastOp : public OpRewritePattern<DimOp> {
- using OpRewritePattern<DimOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(DimOp dimOp,
- PatternRewriter &rewriter) const override {
- auto castOp = dimOp.memrefOrTensor().getDefiningOp<CastOpTy>();
- if (!castOp)
- return failure();
- Value newSource = castOp.getOperand();
- rewriter.replaceOpWithNewOp<DimOp>(dimOp, newSource, dimOp.index());
- return success();
- }
-};
-} // end anonymous namespace.
-
-void DimOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<DimOfMemRefReshape, DimOfCastOp<TensorToMemrefOp>,
- DimOfCastOp<tensor::CastOp>>(context);
-}
-
// ---------------------------------------------------------------------------
// DivFOp
// ---------------------------------------------------------------------------
@@ -1572,256 +1183,6 @@ OpFoldResult DivFOp::fold(ArrayRef<Attribute> operands) {
operands, [](APFloat a, APFloat b) { return a / b; });
}
-// ---------------------------------------------------------------------------
-// DmaStartOp
-// ---------------------------------------------------------------------------
-
-void DmaStartOp::build(OpBuilder &builder, OperationState &result,
- Value srcMemRef, ValueRange srcIndices, Value destMemRef,
- ValueRange destIndices, Value numElements,
- Value tagMemRef, ValueRange tagIndices, Value stride,
- Value elementsPerStride) {
- result.addOperands(srcMemRef);
- result.addOperands(srcIndices);
- result.addOperands(destMemRef);
- result.addOperands(destIndices);
- result.addOperands({numElements, tagMemRef});
- result.addOperands(tagIndices);
- if (stride)
- result.addOperands({stride, elementsPerStride});
-}
-
-void DmaStartOp::print(OpAsmPrinter &p) {
- p << "dma_start " << getSrcMemRef() << '[' << getSrcIndices() << "], "
- << getDstMemRef() << '[' << getDstIndices() << "], " << getNumElements()
- << ", " << getTagMemRef() << '[' << getTagIndices() << ']';
- if (isStrided())
- p << ", " << getStride() << ", " << getNumElementsPerStride();
-
- p.printOptionalAttrDict((*this)->getAttrs());
- p << " : " << getSrcMemRef().getType() << ", " << getDstMemRef().getType()
- << ", " << getTagMemRef().getType();
-}
-
-// Parse DmaStartOp.
-// Ex:
-// %dma_id = dma_start %src[%i, %j], %dst[%k, %l], %size,
-// %tag[%index], %stride, %num_elt_per_stride :
-// : memref<3076 x f32, 0>,
-// memref<1024 x f32, 2>,
-// memref<1 x i32>
-//
-ParseResult DmaStartOp::parse(OpAsmParser &parser, OperationState &result) {
- OpAsmParser::OperandType srcMemRefInfo;
- SmallVector<OpAsmParser::OperandType, 4> srcIndexInfos;
- OpAsmParser::OperandType dstMemRefInfo;
- SmallVector<OpAsmParser::OperandType, 4> dstIndexInfos;
- OpAsmParser::OperandType numElementsInfo;
- OpAsmParser::OperandType tagMemrefInfo;
- SmallVector<OpAsmParser::OperandType, 4> tagIndexInfos;
- SmallVector<OpAsmParser::OperandType, 2> strideInfo;
-
- SmallVector<Type, 3> types;
- auto indexType = parser.getBuilder().getIndexType();
-
- // Parse and resolve the following list of operands:
- // *) source memref followed by its indices (in square brackets).
- // *) destination memref followed by its indices (in square brackets).
- // *) dma size in KiB.
- if (parser.parseOperand(srcMemRefInfo) ||
- parser.parseOperandList(srcIndexInfos, OpAsmParser::Delimiter::Square) ||
- parser.parseComma() || parser.parseOperand(dstMemRefInfo) ||
- parser.parseOperandList(dstIndexInfos, OpAsmParser::Delimiter::Square) ||
- parser.parseComma() || parser.parseOperand(numElementsInfo) ||
- parser.parseComma() || parser.parseOperand(tagMemrefInfo) ||
- parser.parseOperandList(tagIndexInfos, OpAsmParser::Delimiter::Square))
- return failure();
-
- // Parse optional stride and elements per stride.
- if (parser.parseTrailingOperandList(strideInfo))
- return failure();
-
- bool isStrided = strideInfo.size() == 2;
- if (!strideInfo.empty() && !isStrided) {
- return parser.emitError(parser.getNameLoc(),
- "expected two stride related operands");
- }
-
- if (parser.parseColonTypeList(types))
- return failure();
- if (types.size() != 3)
- return parser.emitError(parser.getNameLoc(), "fewer/more types expected");
-
- if (parser.resolveOperand(srcMemRefInfo, types[0], result.operands) ||
- parser.resolveOperands(srcIndexInfos, indexType, result.operands) ||
- parser.resolveOperand(dstMemRefInfo, types[1], result.operands) ||
- parser.resolveOperands(dstIndexInfos, indexType, result.operands) ||
- // size should be an index.
- parser.resolveOperand(numElementsInfo, indexType, result.operands) ||
- parser.resolveOperand(tagMemrefInfo, types[2], result.operands) ||
- // tag indices should be index.
- parser.resolveOperands(tagIndexInfos, indexType, result.operands))
- return failure();
-
- if (isStrided) {
- if (parser.resolveOperands(strideInfo, indexType, result.operands))
- return failure();
- }
-
- return success();
-}
-
-LogicalResult DmaStartOp::verify() {
- unsigned numOperands = getNumOperands();
-
- // Mandatory non-variadic operands are: src memref, dst memref, tag memref and
- // the number of elements.
- if (numOperands < 4)
- return emitOpError("expected at least 4 operands");
-
- // Check types of operands. The order of these calls is important: the later
- // calls rely on some type properties to compute the operand position.
- // 1. Source memref.
- if (!getSrcMemRef().getType().isa<MemRefType>())
- return emitOpError("expected source to be of memref type");
- if (numOperands < getSrcMemRefRank() + 4)
- return emitOpError() << "expected at least " << getSrcMemRefRank() + 4
- << " operands";
- if (!getSrcIndices().empty() &&
- !llvm::all_of(getSrcIndices().getTypes(),
- [](Type t) { return t.isIndex(); }))
- return emitOpError("expected source indices to be of index type");
-
- // 2. Destination memref.
- if (!getDstMemRef().getType().isa<MemRefType>())
- return emitOpError("expected destination to be of memref type");
- unsigned numExpectedOperands = getSrcMemRefRank() + getDstMemRefRank() + 4;
- if (numOperands < numExpectedOperands)
- return emitOpError() << "expected at least " << numExpectedOperands
- << " operands";
- if (!getDstIndices().empty() &&
- !llvm::all_of(getDstIndices().getTypes(),
- [](Type t) { return t.isIndex(); }))
- return emitOpError("expected destination indices to be of index type");
-
- // 3. Number of elements.
- if (!getNumElements().getType().isIndex())
- return emitOpError("expected num elements to be of index type");
-
- // 4. Tag memref.
- if (!getTagMemRef().getType().isa<MemRefType>())
- return emitOpError("expected tag to be of memref type");
- numExpectedOperands += getTagMemRefRank();
- if (numOperands < numExpectedOperands)
- return emitOpError() << "expected at least " << numExpectedOperands
- << " operands";
- if (!getTagIndices().empty() &&
- !llvm::all_of(getTagIndices().getTypes(),
- [](Type t) { return t.isIndex(); }))
- return emitOpError("expected tag indices to be of index type");
-
- // DMAs from
diff erent memory spaces supported.
- if (getSrcMemorySpace() == getDstMemorySpace())
- return emitOpError("DMA should be between
diff erent memory spaces");
-
- // Optional stride-related operands must be either both present or both
- // absent.
- if (numOperands != numExpectedOperands &&
- numOperands != numExpectedOperands + 2)
- return emitOpError("incorrect number of operands");
-
- // 5. Strides.
- if (isStrided()) {
- if (!getStride().getType().isIndex() ||
- !getNumElementsPerStride().getType().isIndex())
- return emitOpError(
- "expected stride and num elements per stride to be of type index");
- }
-
- return success();
-}
-
-LogicalResult DmaStartOp::fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results) {
- /// dma_start(memrefcast) -> dma_start
- return foldMemRefCast(*this);
-}
-
-// ---------------------------------------------------------------------------
-// DmaWaitOp
-// ---------------------------------------------------------------------------
-
-void DmaWaitOp::build(OpBuilder &builder, OperationState &result,
- Value tagMemRef, ValueRange tagIndices,
- Value numElements) {
- result.addOperands(tagMemRef);
- result.addOperands(tagIndices);
- result.addOperands(numElements);
-}
-
-void DmaWaitOp::print(OpAsmPrinter &p) {
- p << "dma_wait " << getTagMemRef() << '[' << getTagIndices() << "], "
- << getNumElements();
- p.printOptionalAttrDict((*this)->getAttrs());
- p << " : " << getTagMemRef().getType();
-}
-
-// Parse DmaWaitOp.
-// Eg:
-// dma_wait %tag[%index], %num_elements : memref<1 x i32, (d0) -> (d0), 4>
-//
-ParseResult DmaWaitOp::parse(OpAsmParser &parser, OperationState &result) {
- OpAsmParser::OperandType tagMemrefInfo;
- SmallVector<OpAsmParser::OperandType, 2> tagIndexInfos;
- Type type;
- auto indexType = parser.getBuilder().getIndexType();
- OpAsmParser::OperandType numElementsInfo;
-
- // Parse tag memref, its indices, and dma size.
- if (parser.parseOperand(tagMemrefInfo) ||
- parser.parseOperandList(tagIndexInfos, OpAsmParser::Delimiter::Square) ||
- parser.parseComma() || parser.parseOperand(numElementsInfo) ||
- parser.parseColonType(type) ||
- parser.resolveOperand(tagMemrefInfo, type, result.operands) ||
- parser.resolveOperands(tagIndexInfos, indexType, result.operands) ||
- parser.resolveOperand(numElementsInfo, indexType, result.operands))
- return failure();
-
- return success();
-}
-
-LogicalResult DmaWaitOp::fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results) {
- /// dma_wait(memrefcast) -> dma_wait
- return foldMemRefCast(*this);
-}
-
-LogicalResult DmaWaitOp::verify() {
- // Mandatory non-variadic operands are tag and the number of elements.
- if (getNumOperands() < 2)
- return emitOpError() << "expected at least 2 operands";
-
- // Check types of operands. The order of these calls is important: the later
- // calls rely on some type properties to compute the operand position.
- if (!getTagMemRef().getType().isa<MemRefType>())
- return emitOpError() << "expected tag to be of memref type";
-
- if (getNumOperands() != 2 + getTagMemRefRank())
- return emitOpError() << "expected " << 2 + getTagMemRefRank()
- << " operands";
-
- if (!getTagIndices().empty() &&
- !llvm::all_of(getTagIndices().getTypes(),
- [](Type t) { return t.isIndex(); }))
- return emitOpError() << "expected tag indices to be of index type";
-
- if (!getNumElements().getType().isIndex())
- return emitOpError()
- << "expected the number of elements to be of index type";
-
- return success();
-}
-
//===----------------------------------------------------------------------===//
// FPExtOp
//===----------------------------------------------------------------------===//
@@ -1876,106 +1237,6 @@ bool FPTruncOp::areCastCompatible(TypeRange inputs, TypeRange outputs) {
return areVectorCastSimpleCompatible(a, b, areCastCompatible);
}
-//===----------------------------------------------------------------------===//
-// GlobalMemrefOp
-//===----------------------------------------------------------------------===//
-
-static void printGlobalMemrefOpTypeAndInitialValue(OpAsmPrinter &p,
- GlobalMemrefOp op,
- TypeAttr type,
- Attribute initialValue) {
- p << type;
- if (!op.isExternal()) {
- p << " = ";
- if (op.isUninitialized())
- p << "uninitialized";
- else
- p.printAttributeWithoutType(initialValue);
- }
-}
-
-static ParseResult
-parseGlobalMemrefOpTypeAndInitialValue(OpAsmParser &parser, TypeAttr &typeAttr,
- Attribute &initialValue) {
- Type type;
- if (parser.parseType(type))
- return failure();
-
- auto memrefType = type.dyn_cast<MemRefType>();
- if (!memrefType || !memrefType.hasStaticShape())
- return parser.emitError(parser.getNameLoc())
- << "type should be static shaped memref, but got " << type;
- typeAttr = TypeAttr::get(type);
-
- if (parser.parseOptionalEqual())
- return success();
-
- if (succeeded(parser.parseOptionalKeyword("uninitialized"))) {
- initialValue = UnitAttr::get(parser.getBuilder().getContext());
- return success();
- }
-
- Type tensorType = getTensorTypeFromMemRefType(memrefType);
- if (parser.parseAttribute(initialValue, tensorType))
- return failure();
- if (!initialValue.isa<ElementsAttr>())
- return parser.emitError(parser.getNameLoc())
- << "initial value should be a unit or elements attribute";
- return success();
-}
-
-static LogicalResult verify(GlobalMemrefOp op) {
- auto memrefType = op.type().dyn_cast<MemRefType>();
- if (!memrefType || !memrefType.hasStaticShape())
- return op.emitOpError("type should be static shaped memref, but got ")
- << op.type();
-
- // Verify that the initial value, if present, is either a unit attribute or
- // an elements attribute.
- if (op.initial_value().hasValue()) {
- Attribute initValue = op.initial_value().getValue();
- if (!initValue.isa<UnitAttr>() && !initValue.isa<ElementsAttr>())
- return op.emitOpError("initial value should be a unit or elements "
- "attribute, but got ")
- << initValue;
-
- // Check that the type of the initial value is compatible with the type of
- // the global variable.
- if (initValue.isa<ElementsAttr>()) {
- Type initType = initValue.getType();
- Type tensorType = getTensorTypeFromMemRefType(memrefType);
- if (initType != tensorType)
- return op.emitOpError("initial value expected to be of type ")
- << tensorType << ", but was of type " << initType;
- }
- }
-
- // TODO: verify visibility for declarations.
- return success();
-}
-
-//===----------------------------------------------------------------------===//
-// GetGlobalMemrefOp
-//===----------------------------------------------------------------------===//
-
-LogicalResult
-GetGlobalMemrefOp::verifySymbolUses(SymbolTableCollection &symbolTable) {
- // Verify that the result type is same as the type of the referenced
- // global_memref op.
- auto global =
- symbolTable.lookupNearestSymbolFrom<GlobalMemrefOp>(*this, nameAttr());
- if (!global)
- return emitOpError("'")
- << name() << "' does not reference a valid global memref";
-
- Type resultType = result().getType();
- if (global.type() != resultType)
- return emitOpError("result type ")
- << resultType << " does not match type " << global.type()
- << " of the global memref @" << name();
- return success();
-}
-
//===----------------------------------------------------------------------===//
// IndexCastOp
//===----------------------------------------------------------------------===//
@@ -2014,288 +1275,16 @@ OpFoldResult IndexCastOp::fold(ArrayRef<Attribute> cstOperands) {
}
//===----------------------------------------------------------------------===//
-// LoadOp
+// MulFOp
//===----------------------------------------------------------------------===//
-static LogicalResult verify(LoadOp op) {
- if (op.getNumOperands() != 1 + op.getMemRefType().getRank())
- return op.emitOpError("incorrect number of indices for load");
- return success();
-}
-
-OpFoldResult LoadOp::fold(ArrayRef<Attribute> cstOperands) {
- /// load(memrefcast) -> load
- if (succeeded(foldMemRefCast(*this)))
- return getResult();
- return OpFoldResult();
-}
-
-namespace {
-/// Fold a load on a tensor_to_memref operation into an tensor.extract on the
-/// corresponding tensor.
-struct LoadOfTensorToMemref : public OpRewritePattern<LoadOp> {
- using OpRewritePattern<LoadOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(LoadOp load,
- PatternRewriter &rewriter) const override {
- auto tensorToMemref = load.memref().getDefiningOp<TensorToMemrefOp>();
- if (!tensorToMemref)
- return failure();
-
- rewriter.replaceOpWithNewOp<tensor::ExtractOp>(
- load, tensorToMemref.tensor(), load.indices());
- return success();
- }
-};
-} // end anonymous namespace.
-
-void LoadOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<LoadOfTensorToMemref>(context);
+OpFoldResult MulFOp::fold(ArrayRef<Attribute> operands) {
+ return constFoldBinaryOp<FloatAttr>(
+ operands, [](APFloat a, APFloat b) { return a * b; });
}
//===----------------------------------------------------------------------===//
-// MemRefCastOp
-//===----------------------------------------------------------------------===//
-
-Value MemRefCastOp::getViewSource() { return source(); }
-
-bool MemRefCastOp::areCastCompatible(TypeRange inputs, TypeRange outputs) {
- if (inputs.size() != 1 || outputs.size() != 1)
- return false;
- Type a = inputs.front(), b = outputs.front();
- auto aT = a.dyn_cast<MemRefType>();
- auto bT = b.dyn_cast<MemRefType>();
-
- auto uaT = a.dyn_cast<UnrankedMemRefType>();
- auto ubT = b.dyn_cast<UnrankedMemRefType>();
-
- if (aT && bT) {
- if (aT.getElementType() != bT.getElementType())
- return false;
- if (aT.getAffineMaps() != bT.getAffineMaps()) {
- int64_t aOffset, bOffset;
- SmallVector<int64_t, 4> aStrides, bStrides;
- if (failed(getStridesAndOffset(aT, aStrides, aOffset)) ||
- failed(getStridesAndOffset(bT, bStrides, bOffset)) ||
- aStrides.size() != bStrides.size())
- return false;
-
- // Strides along a dimension/offset are compatible if the value in the
- // source memref is static and the value in the target memref is the
- // same. They are also compatible if either one is dynamic (see
- // description of MemRefCastOp for details).
- auto checkCompatible = [](int64_t a, int64_t b) {
- return (a == MemRefType::getDynamicStrideOrOffset() ||
- b == MemRefType::getDynamicStrideOrOffset() || a == b);
- };
- if (!checkCompatible(aOffset, bOffset))
- return false;
- for (auto aStride : enumerate(aStrides))
- if (!checkCompatible(aStride.value(), bStrides[aStride.index()]))
- return false;
- }
- if (aT.getMemorySpaceAsInt() != bT.getMemorySpaceAsInt())
- return false;
-
- // They must have the same rank, and any specified dimensions must match.
- if (aT.getRank() != bT.getRank())
- return false;
-
- for (unsigned i = 0, e = aT.getRank(); i != e; ++i) {
- int64_t aDim = aT.getDimSize(i), bDim = bT.getDimSize(i);
- if (aDim != -1 && bDim != -1 && aDim != bDim)
- return false;
- }
- return true;
- } else {
- if (!aT && !uaT)
- return false;
- if (!bT && !ubT)
- return false;
- // Unranked to unranked casting is unsupported
- if (uaT && ubT)
- return false;
-
- auto aEltType = (aT) ? aT.getElementType() : uaT.getElementType();
- auto bEltType = (bT) ? bT.getElementType() : ubT.getElementType();
- if (aEltType != bEltType)
- return false;
-
- auto aMemSpace =
- (aT) ? aT.getMemorySpaceAsInt() : uaT.getMemorySpaceAsInt();
- auto bMemSpace =
- (bT) ? bT.getMemorySpaceAsInt() : ubT.getMemorySpaceAsInt();
- if (aMemSpace != bMemSpace)
- return false;
-
- return true;
- }
-
- return false;
-}
-
-OpFoldResult MemRefCastOp::fold(ArrayRef<Attribute> operands) {
- return succeeded(foldMemRefCast(*this)) ? getResult() : Value();
-}
-
-//===----------------------------------------------------------------------===//
-// MemRefReinterpretCastOp
-//===----------------------------------------------------------------------===//
-
-/// Build a MemRefReinterpretCastOp with all dynamic entries: `staticOffsets`,
-/// `staticSizes` and `staticStrides` are automatically filled with
-/// source-memref-rank sentinel values that encode dynamic entries.
-void mlir::MemRefReinterpretCastOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- OpFoldResult offset,
- ArrayRef<OpFoldResult> sizes,
- ArrayRef<OpFoldResult> strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
- SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
- dispatchIndexOpFoldResults(offset, dynamicOffsets, staticOffsets,
- ShapedType::kDynamicStrideOrOffset);
- dispatchIndexOpFoldResults(sizes, dynamicSizes, staticSizes,
- ShapedType::kDynamicSize);
- dispatchIndexOpFoldResults(strides, dynamicStrides, staticStrides,
- ShapedType::kDynamicStrideOrOffset);
- build(b, result, resultType, source, dynamicOffsets, dynamicSizes,
- dynamicStrides, b.getI64ArrayAttr(staticOffsets),
- b.getI64ArrayAttr(staticSizes), b.getI64ArrayAttr(staticStrides));
- result.addAttributes(attrs);
-}
-
-void mlir::MemRefReinterpretCastOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- int64_t offset,
- ArrayRef<int64_t> sizes,
- ArrayRef<int64_t> strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<OpFoldResult> sizeValues =
- llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
- llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- build(b, result, resultType, source, b.getI64IntegerAttr(offset), sizeValues,
- strideValues, attrs);
-}
-
-void mlir::MemRefReinterpretCastOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- Value offset, ValueRange sizes,
- ValueRange strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<OpFoldResult> sizeValues = llvm::to_vector<4>(
- llvm::map_range(sizes, [](Value v) -> OpFoldResult { return v; }));
- SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
- llvm::map_range(strides, [](Value v) -> OpFoldResult { return v; }));
- build(b, result, resultType, source, offset, sizeValues, strideValues, attrs);
-}
-
-// TODO: ponder whether we want to allow missing trailing sizes/strides that are
-// completed automatically, like we have for subview and subtensor.
-static LogicalResult verify(MemRefReinterpretCastOp op) {
- // The source and result memrefs should be in the same memory space.
- auto srcType = op.source().getType().cast<BaseMemRefType>();
- auto resultType = op.getType().cast<MemRefType>();
- if (srcType.getMemorySpaceAsInt() != resultType.getMemorySpaceAsInt())
- return op.emitError("
diff erent memory spaces specified for source type ")
- << srcType << " and result memref type " << resultType;
- if (srcType.getElementType() != resultType.getElementType())
- return op.emitError("
diff erent element types specified for source type ")
- << srcType << " and result memref type " << resultType;
-
- // Match sizes in result memref type and in static_sizes attribute.
- for (auto &en :
- llvm::enumerate(llvm::zip(resultType.getShape(),
- extractFromI64ArrayAttr(op.static_sizes())))) {
- int64_t resultSize = std::get<0>(en.value());
- int64_t expectedSize = std::get<1>(en.value());
- if (resultSize != expectedSize)
- return op.emitError("expected result type with size = ")
- << expectedSize << " instead of " << resultSize
- << " in dim = " << en.index();
- }
-
- // Match offset and strides in static_offset and static_strides attributes if
- // result memref type has an affine map specified.
- if (!resultType.getAffineMaps().empty()) {
- int64_t resultOffset;
- SmallVector<int64_t, 4> resultStrides;
- if (failed(getStridesAndOffset(resultType, resultStrides, resultOffset)))
- return failure();
-
- // Match offset in result memref type and in static_offsets attribute.
- int64_t expectedOffset =
- extractFromI64ArrayAttr(op.static_offsets()).front();
- if (resultOffset != expectedOffset)
- return op.emitError("expected result type with offset = ")
- << resultOffset << " instead of " << expectedOffset;
-
- // Match strides in result memref type and in static_strides attribute.
- for (auto &en : llvm::enumerate(llvm::zip(
- resultStrides, extractFromI64ArrayAttr(op.static_strides())))) {
- int64_t resultStride = std::get<0>(en.value());
- int64_t expectedStride = std::get<1>(en.value());
- if (resultStride != expectedStride)
- return op.emitError("expected result type with stride = ")
- << expectedStride << " instead of " << resultStride
- << " in dim = " << en.index();
- }
- }
- return success();
-}
-
-//===----------------------------------------------------------------------===//
-// MemRefReshapeOp
-//===----------------------------------------------------------------------===//
-
-static LogicalResult verify(MemRefReshapeOp op) {
- Type operandType = op.source().getType();
- Type resultType = op.result().getType();
-
- Type operandElementType = operandType.cast<ShapedType>().getElementType();
- Type resultElementType = resultType.cast<ShapedType>().getElementType();
- if (operandElementType != resultElementType)
- return op.emitOpError("element types of source and destination memref "
- "types should be the same");
-
- if (auto operandMemRefType = operandType.dyn_cast<MemRefType>())
- if (!operandMemRefType.getAffineMaps().empty())
- return op.emitOpError(
- "source memref type should have identity affine map");
-
- int64_t shapeSize = op.shape().getType().cast<MemRefType>().getDimSize(0);
- auto resultMemRefType = resultType.dyn_cast<MemRefType>();
- if (resultMemRefType) {
- if (!resultMemRefType.getAffineMaps().empty())
- return op.emitOpError(
- "result memref type should have identity affine map");
- if (shapeSize == ShapedType::kDynamicSize)
- return op.emitOpError("cannot use shape operand with dynamic length to "
- "reshape to statically-ranked memref type");
- if (shapeSize != resultMemRefType.getRank())
- return op.emitOpError(
- "length of shape operand
diff ers from the result's memref rank");
- }
- return success();
-}
-
-//===----------------------------------------------------------------------===//
-// MulFOp
-//===----------------------------------------------------------------------===//
-
-OpFoldResult MulFOp::fold(ArrayRef<Attribute> operands) {
- return constFoldBinaryOp<FloatAttr>(
- operands, [](APFloat a, APFloat b) { return a * b; });
-}
-
-//===----------------------------------------------------------------------===//
-// MulIOp
+// MulIOp
//===----------------------------------------------------------------------===//
OpFoldResult MulIOp::fold(ArrayRef<Attribute> operands) {
@@ -2327,76 +1316,6 @@ OpFoldResult OrOp::fold(ArrayRef<Attribute> operands) {
[](APInt a, APInt b) { return a | b; });
}
-//===----------------------------------------------------------------------===//
-// PrefetchOp
-//===----------------------------------------------------------------------===//
-
-static void print(OpAsmPrinter &p, PrefetchOp op) {
- p << PrefetchOp::getOperationName() << " " << op.memref() << '[';
- p.printOperands(op.indices());
- p << ']' << ", " << (op.isWrite() ? "write" : "read");
- p << ", locality<" << op.localityHint();
- p << ">, " << (op.isDataCache() ? "data" : "instr");
- p.printOptionalAttrDict(
- op->getAttrs(),
- /*elidedAttrs=*/{"localityHint", "isWrite", "isDataCache"});
- p << " : " << op.getMemRefType();
-}
-
-static ParseResult parsePrefetchOp(OpAsmParser &parser,
- OperationState &result) {
- OpAsmParser::OperandType memrefInfo;
- SmallVector<OpAsmParser::OperandType, 4> indexInfo;
- IntegerAttr localityHint;
- MemRefType type;
- StringRef readOrWrite, cacheType;
-
- auto indexTy = parser.getBuilder().getIndexType();
- auto i32Type = parser.getBuilder().getIntegerType(32);
- if (parser.parseOperand(memrefInfo) ||
- parser.parseOperandList(indexInfo, OpAsmParser::Delimiter::Square) ||
- parser.parseComma() || parser.parseKeyword(&readOrWrite) ||
- parser.parseComma() || parser.parseKeyword("locality") ||
- parser.parseLess() ||
- parser.parseAttribute(localityHint, i32Type, "localityHint",
- result.attributes) ||
- parser.parseGreater() || parser.parseComma() ||
- parser.parseKeyword(&cacheType) || parser.parseColonType(type) ||
- parser.resolveOperand(memrefInfo, type, result.operands) ||
- parser.resolveOperands(indexInfo, indexTy, result.operands))
- return failure();
-
- if (!readOrWrite.equals("read") && !readOrWrite.equals("write"))
- return parser.emitError(parser.getNameLoc(),
- "rw specifier has to be 'read' or 'write'");
- result.addAttribute(
- PrefetchOp::getIsWriteAttrName(),
- parser.getBuilder().getBoolAttr(readOrWrite.equals("write")));
-
- if (!cacheType.equals("data") && !cacheType.equals("instr"))
- return parser.emitError(parser.getNameLoc(),
- "cache type has to be 'data' or 'instr'");
-
- result.addAttribute(
- PrefetchOp::getIsDataCacheAttrName(),
- parser.getBuilder().getBoolAttr(cacheType.equals("data")));
-
- return success();
-}
-
-static LogicalResult verify(PrefetchOp op) {
- if (op.getNumOperands() != 1 + op.getMemRefType().getRank())
- return op.emitOpError("too few indices");
-
- return success();
-}
-
-LogicalResult PrefetchOp::fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results) {
- // prefetch(memrefcast) -> prefetch
- return foldMemRefCast(*this);
-}
-
//===----------------------------------------------------------------------===//
// RankOp
//===----------------------------------------------------------------------===//
@@ -2735,23 +1654,6 @@ OpFoldResult SplatOp::fold(ArrayRef<Attribute> operands) {
return SplatElementsAttr::get(shapedType, {constOperand});
}
-//===----------------------------------------------------------------------===//
-// StoreOp
-//===----------------------------------------------------------------------===//
-
-static LogicalResult verify(StoreOp op) {
- if (op.getNumOperands() != 2 + op.getMemRefType().getRank())
- return op.emitOpError("store index operand count not equal to memref rank");
-
- return success();
-}
-
-LogicalResult StoreOp::fold(ArrayRef<Attribute> cstOperands,
- SmallVectorImpl<OpFoldResult> &results) {
- /// store(memrefcast) -> store
- return foldMemRefCast(*this);
-}
-
//===----------------------------------------------------------------------===//
// SubFOp
//===----------------------------------------------------------------------===//
@@ -2792,97 +1694,33 @@ bool UIToFPOp::areCastCompatible(TypeRange inputs, TypeRange outputs) {
}
//===----------------------------------------------------------------------===//
-// SubViewOp
+// SubTensorOp
//===----------------------------------------------------------------------===//
-namespace {
-/// Helpers to write more idiomatic operations.
-namespace saturated_arith {
-struct Wrapper {
- explicit Wrapper(int64_t v) : v(v) {}
- operator int64_t() { return v; }
- int64_t v;
-};
-Wrapper operator+(Wrapper a, int64_t b) {
- if (ShapedType::isDynamicStrideOrOffset(a) ||
- ShapedType::isDynamicStrideOrOffset(b))
- return Wrapper(ShapedType::kDynamicStrideOrOffset);
- return Wrapper(a.v + b);
-}
-Wrapper operator*(Wrapper a, int64_t b) {
- if (ShapedType::isDynamicStrideOrOffset(a) ||
- ShapedType::isDynamicStrideOrOffset(b))
- return Wrapper(ShapedType::kDynamicStrideOrOffset);
- return Wrapper(a.v * b);
-}
-} // end namespace saturated_arith
-} // end namespace
-
-/// A subview result type can be fully inferred from the source type and the
+/// A subtensor result type can be fully inferred from the source type and the
/// static representation of offsets, sizes and strides. Special sentinels
/// encode the dynamic case.
-Type SubViewOp::inferResultType(MemRefType sourceMemRefType,
- ArrayRef<int64_t> leadingStaticOffsets,
- ArrayRef<int64_t> leadingStaticSizes,
- ArrayRef<int64_t> leadingStaticStrides) {
- // A subview may specify only a leading subset of offset/sizes/strides in
+Type SubTensorOp::inferResultType(RankedTensorType sourceRankedTensorType,
+ ArrayRef<int64_t> leadingStaticOffsets,
+ ArrayRef<int64_t> leadingStaticSizes,
+ ArrayRef<int64_t> leadingStaticStrides) {
+ // A subtensor may specify only a leading subset of offset/sizes/strides in
// which case we complete with offset=0, sizes from memref type and strides=1.
- unsigned rank = sourceMemRefType.getRank();
- assert(leadingStaticOffsets.size() <= rank &&
- "unexpected leadingStaticOffsets overflow");
+ unsigned rank = sourceRankedTensorType.getRank();
assert(leadingStaticSizes.size() <= rank &&
"unexpected leadingStaticSizes overflow");
- assert(leadingStaticStrides.size() <= rank &&
- "unexpected leadingStaticStrides overflow");
- auto staticOffsets = llvm::to_vector<4>(leadingStaticOffsets);
auto staticSizes = llvm::to_vector<4>(leadingStaticSizes);
- auto staticStrides = llvm::to_vector<4>(leadingStaticStrides);
- unsigned numTrailingOffsets = rank - staticOffsets.size();
unsigned numTrailingSizes = rank - staticSizes.size();
- unsigned numTrailingStrides = rank - staticStrides.size();
- staticOffsets.append(numTrailingOffsets, 0);
- llvm::append_range(staticSizes,
- sourceMemRefType.getShape().take_back(numTrailingSizes));
- staticStrides.append(numTrailingStrides, 1);
-
- // Extract source offset and strides.
- int64_t sourceOffset;
- SmallVector<int64_t, 4> sourceStrides;
- auto res = getStridesAndOffset(sourceMemRefType, sourceStrides, sourceOffset);
- assert(succeeded(res) && "SubViewOp expected strided memref type");
- (void)res;
-
- // Compute target offset whose value is:
- // `sourceOffset + sum_i(staticOffset_i * sourceStrides_i)`.
- int64_t targetOffset = sourceOffset;
- for (auto it : llvm::zip(staticOffsets, sourceStrides)) {
- auto staticOffset = std::get<0>(it), targetStride = std::get<1>(it);
- using namespace saturated_arith;
- targetOffset = Wrapper(targetOffset) + Wrapper(staticOffset) * targetStride;
- }
-
- // Compute target stride whose value is:
- // `sourceStrides_i * staticStrides_i`.
- SmallVector<int64_t, 4> targetStrides;
- targetStrides.reserve(staticOffsets.size());
- for (auto it : llvm::zip(sourceStrides, staticStrides)) {
- auto sourceStride = std::get<0>(it), staticStride = std::get<1>(it);
- using namespace saturated_arith;
- targetStrides.push_back(Wrapper(sourceStride) * staticStride);
- }
-
- // The type is now known.
- return MemRefType::get(
- staticSizes, sourceMemRefType.getElementType(),
- makeStridedLinearLayoutMap(targetStrides, targetOffset,
- sourceMemRefType.getContext()),
- sourceMemRefType.getMemorySpaceAsInt());
+ llvm::append_range(staticSizes, sourceRankedTensorType.getShape().take_back(
+ numTrailingSizes));
+ return RankedTensorType::get(staticSizes,
+ sourceRankedTensorType.getElementType());
}
-Type SubViewOp::inferResultType(MemRefType sourceMemRefType,
- ArrayRef<OpFoldResult> leadingStaticOffsets,
- ArrayRef<OpFoldResult> leadingStaticSizes,
- ArrayRef<OpFoldResult> leadingStaticStrides) {
+Type SubTensorOp::inferResultType(RankedTensorType sourceRankedTensorType,
+ ArrayRef<OpFoldResult> leadingStaticOffsets,
+ ArrayRef<OpFoldResult> leadingStaticSizes,
+ ArrayRef<OpFoldResult> leadingStaticStrides) {
SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
dispatchIndexOpFoldResults(leadingStaticOffsets, dynamicOffsets,
@@ -2891,55 +1729,39 @@ Type SubViewOp::inferResultType(MemRefType sourceMemRefType,
ShapedType::kDynamicSize);
dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
staticStrides, ShapedType::kDynamicStrideOrOffset);
- return SubViewOp::inferResultType(sourceMemRefType, staticOffsets,
- staticSizes, staticStrides);
-}
-
-static void
-getPositionsOfShapeOne(unsigned rank, ArrayRef<int64_t> shape,
- llvm::SmallDenseSet<unsigned> &dimsToProject) {
- dimsToProject.reserve(rank);
- for (unsigned pos = 0, e = shape.size(); pos < e && rank > 0; ++pos) {
- if (shape[pos] == 1) {
- dimsToProject.insert(pos);
- --rank;
- }
- }
+ return SubTensorOp::inferResultType(sourceRankedTensorType, staticOffsets,
+ staticSizes, staticStrides);
}
-Type SubViewOp::inferRankReducedResultType(
- unsigned resultRank, MemRefType sourceRankedTensorType,
+/// A subtensor result type can be fully inferred from the source type and the
+/// static representation of offsets, sizes and strides. Special sentinels
+/// encode the dynamic case.
+Type SubTensorOp::inferRankReducedResultType(
+ unsigned resultRank, RankedTensorType sourceRankedTensorType,
ArrayRef<int64_t> leadingStaticOffsets,
ArrayRef<int64_t> leadingStaticSizes,
ArrayRef<int64_t> leadingStaticStrides) {
auto inferredType =
inferResultType(sourceRankedTensorType, leadingStaticOffsets,
leadingStaticSizes, leadingStaticStrides)
- .cast<MemRefType>();
- assert(inferredType.getRank() >= resultRank && "expected ");
+ .cast<RankedTensorType>();
int rankDiff = inferredType.getRank() - resultRank;
if (rankDiff > 0) {
auto shape = inferredType.getShape();
llvm::SmallDenseSet<unsigned> dimsToProject;
- getPositionsOfShapeOne(rankDiff, shape, dimsToProject);
+ mlir::getPositionsOfShapeOne(rankDiff, shape, dimsToProject);
SmallVector<int64_t> projectedShape;
for (unsigned pos = 0, e = shape.size(); pos < e; ++pos)
if (!dimsToProject.contains(pos))
projectedShape.push_back(shape[pos]);
-
- AffineMap map;
- auto maps = inferredType.getAffineMaps();
- if (!maps.empty() && maps.front())
- map = getProjectedMap(maps.front(), dimsToProject);
inferredType =
- MemRefType::get(projectedShape, inferredType.getElementType(), map,
- inferredType.getMemorySpaceAsInt());
+ RankedTensorType::get(projectedShape, inferredType.getElementType());
}
return inferredType;
}
-Type SubViewOp::inferRankReducedResultType(
- unsigned resultRank, MemRefType sourceRankedTensorType,
+Type SubTensorOp::inferRankReducedResultType(
+ unsigned resultRank, RankedTensorType sourceRankedTensorType,
ArrayRef<OpFoldResult> leadingStaticOffsets,
ArrayRef<OpFoldResult> leadingStaticSizes,
ArrayRef<OpFoldResult> leadingStaticStrides) {
@@ -2951,19 +1773,19 @@ Type SubViewOp::inferRankReducedResultType(
ShapedType::kDynamicSize);
dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
staticStrides, ShapedType::kDynamicStrideOrOffset);
- return SubViewOp::inferRankReducedResultType(
+ return SubTensorOp::inferRankReducedResultType(
resultRank, sourceRankedTensorType, staticOffsets, staticSizes,
staticStrides);
}
-// Build a SubViewOp with mixed static and dynamic entries and custom result
+// Build a SubTensorOp with mixed static and dynamic entries and custom result
// type. If the type passed is nullptr, it is inferred.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- ArrayRef<OpFoldResult> offsets,
- ArrayRef<OpFoldResult> sizes,
- ArrayRef<OpFoldResult> strides,
- ArrayRef<NamedAttribute> attrs) {
+void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
+ RankedTensorType resultType, Value source,
+ ArrayRef<OpFoldResult> offsets,
+ ArrayRef<OpFoldResult> sizes,
+ ArrayRef<OpFoldResult> strides,
+ ArrayRef<NamedAttribute> attrs) {
SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
dispatchIndexOpFoldResults(offsets, dynamicOffsets, staticOffsets,
@@ -2972,12 +1794,13 @@ void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
ShapedType::kDynamicSize);
dispatchIndexOpFoldResults(strides, dynamicStrides, staticStrides,
ShapedType::kDynamicStrideOrOffset);
- auto sourceMemRefType = source.getType().cast<MemRefType>();
+ auto sourceRankedTensorType = source.getType().cast<RankedTensorType>();
// Structuring implementation this way avoids duplication between builders.
if (!resultType) {
- resultType = SubViewOp::inferResultType(sourceMemRefType, staticOffsets,
- staticSizes, staticStrides)
- .cast<MemRefType>();
+ resultType =
+ SubTensorOp::inferResultType(sourceRankedTensorType, staticOffsets,
+ staticSizes, staticStrides)
+ .cast<RankedTensorType>();
}
build(b, result, resultType, source, dynamicOffsets, dynamicSizes,
dynamicStrides, b.getI64ArrayAttr(staticOffsets),
@@ -2985,66 +1808,23 @@ void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
result.addAttributes(attrs);
}
-// Build a SubViewOp with mixed static and dynamic entries and inferred result
+// Build a SubTensorOp with mixed static and dynamic entries and inferred result
// type.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
- ArrayRef<OpFoldResult> offsets,
- ArrayRef<OpFoldResult> sizes,
- ArrayRef<OpFoldResult> strides,
- ArrayRef<NamedAttribute> attrs) {
- build(b, result, MemRefType(), source, offsets, sizes, strides, attrs);
-}
-
-// Build a SubViewOp with static entries and inferred result type.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
- ArrayRef<int64_t> offsets, ArrayRef<int64_t> sizes,
- ArrayRef<int64_t> strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
- llvm::map_range(offsets, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- SmallVector<OpFoldResult> sizeValues =
- llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
- llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- build(b, result, source, offsetValues, sizeValues, strideValues, attrs);
-}
-
-// Build a SubViewOp with dynamic entries and custom result type. If the
-// type passed is nullptr, it is inferred.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- ArrayRef<int64_t> offsets, ArrayRef<int64_t> sizes,
- ArrayRef<int64_t> strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
- llvm::map_range(offsets, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- SmallVector<OpFoldResult> sizeValues =
- llvm::to_vector<4>(llvm::map_range(sizes, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
- llvm::map_range(strides, [&](int64_t v) -> OpFoldResult {
- return b.getI64IntegerAttr(v);
- }));
- build(b, result, resultType, source, offsetValues, sizeValues, strideValues,
- attrs);
+void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
+ Value source, ArrayRef<OpFoldResult> offsets,
+ ArrayRef<OpFoldResult> sizes,
+ ArrayRef<OpFoldResult> strides,
+ ArrayRef<NamedAttribute> attrs) {
+ build(b, result, RankedTensorType(), source, offsets, sizes, strides, attrs);
}
-// Build a SubViewOp with dynamic entries and custom result type. If the type
+// Build a SubTensorOp with dynamic entries and custom result type. If the type
// passed is nullptr, it is inferred.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
- MemRefType resultType, Value source,
- ValueRange offsets, ValueRange sizes,
- ValueRange strides,
- ArrayRef<NamedAttribute> attrs) {
+void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
+ RankedTensorType resultType, Value source,
+ ValueRange offsets, ValueRange sizes,
+ ValueRange strides,
+ ArrayRef<NamedAttribute> attrs) {
SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
llvm::map_range(offsets, [](Value v) -> OpFoldResult { return v; }));
SmallVector<OpFoldResult> sizeValues = llvm::to_vector<4>(
@@ -3054,75 +1834,34 @@ void mlir::SubViewOp::build(OpBuilder &b, OperationState &result,
build(b, result, resultType, source, offsetValues, sizeValues, strideValues);
}
-// Build a SubViewOp with dynamic entries and inferred result type.
-void mlir::SubViewOp::build(OpBuilder &b, OperationState &result, Value source,
- ValueRange offsets, ValueRange sizes,
- ValueRange strides,
- ArrayRef<NamedAttribute> attrs) {
- build(b, result, MemRefType(), source, offsets, sizes, strides, attrs);
-}
-
-/// For ViewLikeOpInterface.
-Value SubViewOp::getViewSource() { return source(); }
-
-/// Given an `originalShape` and a `reducedShape` assumed to be a subset of
-/// `originalShape` with some `1` entries erased, return the set of indices
-/// that specifies which of the entries of `originalShape` are dropped to obtain
-/// `reducedShape`. The returned mask can be applied as a projection to
-/// `originalShape` to obtain the `reducedShape`. This mask is useful to track
-/// which dimensions must be kept when e.g. compute MemRef strides under
-/// rank-reducing operations. Return None if reducedShape cannot be obtained
-/// by dropping only `1` entries in `originalShape`.
-llvm::Optional<llvm::SmallDenseSet<unsigned>>
-mlir::computeRankReductionMask(ArrayRef<int64_t> originalShape,
- ArrayRef<int64_t> reducedShape) {
- size_t originalRank = originalShape.size(), reducedRank = reducedShape.size();
- llvm::SmallDenseSet<unsigned> unusedDims;
- unsigned reducedIdx = 0;
- for (unsigned originalIdx = 0; originalIdx < originalRank; ++originalIdx) {
- // Greedily insert `originalIdx` if no match.
- if (reducedIdx < reducedRank &&
- originalShape[originalIdx] == reducedShape[reducedIdx]) {
- reducedIdx++;
- continue;
- }
-
- unusedDims.insert(originalIdx);
- // If no match on `originalIdx`, the `originalShape` at this dimension
- // must be 1, otherwise we bail.
- if (originalShape[originalIdx] != 1)
- return llvm::None;
- }
- // The whole reducedShape must be scanned, otherwise we bail.
- if (reducedIdx != reducedRank)
- return llvm::None;
- return unusedDims;
+// Build a SubTensorOp with dynamic entries and inferred result type.
+void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
+ Value source, ValueRange offsets,
+ ValueRange sizes, ValueRange strides,
+ ArrayRef<NamedAttribute> attrs) {
+ build(b, result, RankedTensorType(), source, offsets, sizes, strides, attrs);
}
-enum SubViewVerificationResult {
+enum SubTensorVerificationResult {
Success,
RankTooLarge,
SizeMismatch,
ElemTypeMismatch,
- MemSpaceMismatch,
- AffineMapMismatch
};
/// Checks if `original` Type type can be rank reduced to `reduced` type.
/// This function is slight variant of `is subsequence` algorithm where
/// not matching dimension must be 1.
-static SubViewVerificationResult
+static SubTensorVerificationResult
isRankReducedType(Type originalType, Type candidateReducedType,
std::string *errMsg = nullptr) {
if (originalType == candidateReducedType)
- return SubViewVerificationResult::Success;
- if (!originalType.isa<RankedTensorType>() && !originalType.isa<MemRefType>())
- return SubViewVerificationResult::Success;
+ return SubTensorVerificationResult::Success;
+ if (!originalType.isa<RankedTensorType>())
+ return SubTensorVerificationResult::Success;
if (originalType.isa<RankedTensorType>() &&
!candidateReducedType.isa<RankedTensorType>())
- return SubViewVerificationResult::Success;
- if (originalType.isa<MemRefType>() && !candidateReducedType.isa<MemRefType>())
- return SubViewVerificationResult::Success;
+ return SubTensorVerificationResult::Success;
ShapedType originalShapedType = originalType.cast<ShapedType>();
ShapedType candidateReducedShapedType =
@@ -3135,553 +1874,75 @@ isRankReducedType(Type originalType, Type candidateReducedType,
unsigned originalRank = originalShape.size(),
candidateReducedRank = candidateReducedShape.size();
if (candidateReducedRank > originalRank)
- return SubViewVerificationResult::RankTooLarge;
+ return SubTensorVerificationResult::RankTooLarge;
auto optionalUnusedDimsMask =
computeRankReductionMask(originalShape, candidateReducedShape);
// Sizes cannot be matched in case empty vector is returned.
if (!optionalUnusedDimsMask.hasValue())
- return SubViewVerificationResult::SizeMismatch;
+ return SubTensorVerificationResult::SizeMismatch;
if (originalShapedType.getElementType() !=
candidateReducedShapedType.getElementType())
- return SubViewVerificationResult::ElemTypeMismatch;
+ return SubTensorVerificationResult::ElemTypeMismatch;
// We are done for the tensor case.
if (originalType.isa<RankedTensorType>())
- return SubViewVerificationResult::Success;
-
- // Strided layout logic is relevant for MemRefType only.
- MemRefType original = originalType.cast<MemRefType>();
- MemRefType candidateReduced = candidateReducedType.cast<MemRefType>();
- if (original.getMemorySpaceAsInt() != candidateReduced.getMemorySpaceAsInt())
- return SubViewVerificationResult::MemSpaceMismatch;
+ return SubTensorVerificationResult::Success;
- llvm::SmallDenseSet<unsigned> unusedDims = optionalUnusedDimsMask.getValue();
- auto inferredType =
- getProjectedMap(getStridedLinearLayoutMap(original), unusedDims);
- AffineMap candidateLayout;
- if (candidateReduced.getAffineMaps().empty())
- candidateLayout = getStridedLinearLayoutMap(candidateReduced);
- else
- candidateLayout = candidateReduced.getAffineMaps().front();
- assert(inferredType.getNumResults() == 1 &&
- candidateLayout.getNumResults() == 1);
- if (inferredType.getNumSymbols() != candidateLayout.getNumSymbols() ||
- inferredType.getNumDims() != candidateLayout.getNumDims()) {
- if (errMsg) {
- llvm::raw_string_ostream os(*errMsg);
- os << "inferred type: " << inferredType;
- }
- return SubViewVerificationResult::AffineMapMismatch;
- }
- // Check that the
diff erence of the affine maps simplifies to 0.
- AffineExpr
diff Expr =
- inferredType.getResult(0) - candidateLayout.getResult(0);
-
diff Expr = simplifyAffineExpr(
diff Expr, inferredType.getNumDims(),
- inferredType.getNumSymbols());
- auto cst =
diff Expr.dyn_cast<AffineConstantExpr>();
- if (!(cst && cst.getValue() == 0)) {
- if (errMsg) {
- llvm::raw_string_ostream os(*errMsg);
- os << "inferred type: " << inferredType;
- }
- return SubViewVerificationResult::AffineMapMismatch;
- }
- return SubViewVerificationResult::Success;
+ return SubTensorVerificationResult::Success;
}
template <typename OpTy>
-static LogicalResult produceSubViewErrorMsg(SubViewVerificationResult result,
- OpTy op, Type expectedType,
- StringRef errMsg = "") {
+static LogicalResult
+produceSubTensorErrorMsg(SubTensorVerificationResult result, OpTy op,
+ Type expectedType, StringRef errMsg = "") {
auto memrefType = expectedType.cast<ShapedType>();
switch (result) {
- case SubViewVerificationResult::Success:
+ case SubTensorVerificationResult::Success:
return success();
- case SubViewVerificationResult::RankTooLarge:
+ case SubTensorVerificationResult::RankTooLarge:
return op.emitError("expected result rank to be smaller or equal to ")
<< "the source rank. " << errMsg;
- case SubViewVerificationResult::SizeMismatch:
+ case SubTensorVerificationResult::SizeMismatch:
return op.emitError("expected result type to be ")
<< expectedType
<< " or a rank-reduced version. (mismatch of result sizes) "
<< errMsg;
- case SubViewVerificationResult::ElemTypeMismatch:
+ case SubTensorVerificationResult::ElemTypeMismatch:
return op.emitError("expected result element type to be ")
<< memrefType.getElementType() << errMsg;
- case SubViewVerificationResult::MemSpaceMismatch:
- return op.emitError("expected result and source memory spaces to match.")
- << errMsg;
- case SubViewVerificationResult::AffineMapMismatch:
- return op.emitError("expected result type to be ")
- << expectedType
- << " or a rank-reduced version. (mismatch of result affine map) "
- << errMsg;
}
- llvm_unreachable("unexpected subview verification result");
+ llvm_unreachable("unexpected subtensor verification result");
}
-
-/// Verifier for SubViewOp.
-static LogicalResult verify(SubViewOp op) {
- MemRefType baseType = op.getSourceType();
- MemRefType subViewType = op.getType();
-
- // The base memref and the view memref should be in the same memory space.
- if (baseType.getMemorySpaceAsInt() != subViewType.getMemorySpaceAsInt())
- return op.emitError("
diff erent memory spaces specified for base memref "
- "type ")
- << baseType << " and subview memref type " << subViewType;
-
- // Verify that the base memref type has a strided layout map.
- if (!isStrided(baseType))
- return op.emitError("base type ") << baseType << " is not strided";
-
+/// Verifier for SubTensorOp.
+static LogicalResult verify(SubTensorOp op) {
// Verify result type against inferred type.
- auto expectedType = SubViewOp::inferResultType(
- baseType, extractFromI64ArrayAttr(op.static_offsets()),
+ auto expectedType = SubTensorOp::inferResultType(
+ op.getSourceType(), extractFromI64ArrayAttr(op.static_offsets()),
extractFromI64ArrayAttr(op.static_sizes()),
extractFromI64ArrayAttr(op.static_strides()));
-
- std::string errMsg;
- auto result = isRankReducedType(expectedType, subViewType, &errMsg);
- return produceSubViewErrorMsg(result, op, expectedType, errMsg);
-}
-
-raw_ostream &mlir::operator<<(raw_ostream &os, Range &range) {
- return os << "range " << range.offset << ":" << range.size << ":"
- << range.stride;
-}
-
-/// Return the list of Range (i.e. offset, size, stride). Each Range
-/// entry contains either the dynamic value or a ConstantIndexOp constructed
-/// with `b` at location `loc`.
-SmallVector<Range, 8> mlir::getOrCreateRanges(OffsetSizeAndStrideOpInterface op,
- OpBuilder &b, Location loc) {
- std::array<unsigned, 3> ranks = op.getArrayAttrMaxRanks();
- assert(ranks[0] == ranks[1] && "expected offset and sizes of equal ranks");
- assert(ranks[1] == ranks[2] && "expected sizes and strides of equal ranks");
- SmallVector<Range, 8> res;
- unsigned rank = ranks[0];
- res.reserve(rank);
- for (unsigned idx = 0; idx < rank; ++idx) {
- Value offset =
- op.isDynamicOffset(idx)
- ? op.getDynamicOffset(idx)
- : b.create<ConstantIndexOp>(loc, op.getStaticOffset(idx));
- Value size = op.isDynamicSize(idx)
- ? op.getDynamicSize(idx)
- : b.create<ConstantIndexOp>(loc, op.getStaticSize(idx));
- Value stride =
- op.isDynamicStride(idx)
- ? op.getDynamicStride(idx)
- : b.create<ConstantIndexOp>(loc, op.getStaticStride(idx));
- res.emplace_back(Range{offset, size, stride});
- }
- return res;
+ auto result = isRankReducedType(expectedType, op.getType());
+ return produceSubTensorErrorMsg(result, op, expectedType);
}
namespace {
-
-/// Detects the `values` produced by a ConstantIndexOp and places the new
-/// constant in place of the corresponding sentinel value.
-void canonicalizeSubViewPart(SmallVectorImpl<OpFoldResult> &values,
- llvm::function_ref<bool(int64_t)> isDynamic) {
- for (OpFoldResult &ofr : values) {
- if (ofr.is<Attribute>())
- continue;
- // Newly static, move from Value to constant.
- if (auto cstOp = ofr.dyn_cast<Value>().getDefiningOp<ConstantIndexOp>())
- ofr = OpBuilder(cstOp).getIndexAttr(cstOp.getValue());
- }
-}
-
-static void replaceWithNewOp(PatternRewriter &rewriter, SubViewOp op,
- SubViewOp newOp) {
- rewriter.replaceOpWithNewOp<MemRefCastOp>(op, newOp, op.getType());
-}
-
-static void replaceWithNewOp(PatternRewriter &rewriter, SubTensorOp op,
- SubTensorOp newOp) {
- Value replacement = newOp.getResult();
- if (replacement.getType() != op.getType())
- replacement =
- rewriter.create<tensor::CastOp>(op.getLoc(), op.getType(), replacement);
- rewriter.replaceOp(op, replacement);
-}
-
-/// Pattern to rewrite a subview op with constant arguments.
-template <typename OpType>
-class OpWithOffsetSizesAndStridesConstantArgumentFolder final
- : public OpRewritePattern<OpType> {
-public:
- using OpRewritePattern<OpType>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(OpType op,
- PatternRewriter &rewriter) const override {
- // No constant operand, just return;
- if (llvm::none_of(op.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
- }))
- return failure();
-
- // At least one of offsets/sizes/strides is a new constant.
- // Form the new list of operands and constant attributes from the existing.
- SmallVector<OpFoldResult> mixedOffsets(op.getMixedOffsets());
- SmallVector<OpFoldResult> mixedSizes(op.getMixedSizes());
- SmallVector<OpFoldResult> mixedStrides(op.getMixedStrides());
- canonicalizeSubViewPart(mixedOffsets, ShapedType::isDynamicStrideOrOffset);
- canonicalizeSubViewPart(mixedSizes, ShapedType::isDynamic);
- canonicalizeSubViewPart(mixedStrides, ShapedType::isDynamicStrideOrOffset);
-
- // Create the new op in canonical form.
- auto newOp = rewriter.create<OpType>(op.getLoc(), op.source(), mixedOffsets,
- mixedSizes, mixedStrides);
-
- replaceWithNewOp(rewriter, op, newOp);
-
- return success();
- }
-};
-
-} // end anonymous namespace
-
-/// Determines whether MemRefCastOp casts to a more dynamic version of the
-/// source memref. This is useful to to fold a memref_cast into a consuming op
-/// and implement canonicalization patterns for ops in
diff erent dialects that
-/// may consume the results of memref_cast operations. Such foldable memref_cast
-/// operations are typically inserted as `view` and `subview` ops are
-/// canonicalized, to preserve the type compatibility of their uses.
-///
-/// Returns true when all conditions are met:
-/// 1. source and result are ranked memrefs with strided semantics and same
-/// element type and rank.
-/// 2. each of the source's size, offset or stride has more static information
-/// than the corresponding result's size, offset or stride.
-///
-/// Example 1:
-/// ```mlir
-/// %1 = memref_cast %0 : memref<8x16xf32> to memref<?x?xf32>
-/// %2 = consumer %1 ... : memref<?x?xf32> ...
-/// ```
-///
-/// may fold into:
-///
-/// ```mlir
-/// %2 = consumer %0 ... : memref<8x16xf32> ...
-/// ```
-///
-/// Example 2:
-/// ```
-/// %1 = memref_cast %0 : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
-/// to memref<?x?xf32>
-/// consumer %1 : memref<?x?xf32> ...
-/// ```
-///
-/// may fold into:
-///
-/// ```
-/// consumer %0 ... : memref<?x16xf32, affine_map<(i, j)->(16 * i + j)>>
-/// ```
-bool mlir::canFoldIntoConsumerOp(MemRefCastOp castOp) {
- MemRefType sourceType = castOp.source().getType().dyn_cast<MemRefType>();
- MemRefType resultType = castOp.getType().dyn_cast<MemRefType>();
-
- // Requires ranked MemRefType.
- if (!sourceType || !resultType)
- return false;
-
- // Requires same elemental type.
- if (sourceType.getElementType() != resultType.getElementType())
- return false;
-
- // Requires same rank.
- if (sourceType.getRank() != resultType.getRank())
- return false;
-
- // Only fold casts between strided memref forms.
- int64_t sourceOffset, resultOffset;
- SmallVector<int64_t, 4> sourceStrides, resultStrides;
- if (failed(getStridesAndOffset(sourceType, sourceStrides, sourceOffset)) ||
- failed(getStridesAndOffset(resultType, resultStrides, resultOffset)))
- return false;
-
- // If cast is towards more static sizes along any dimension, don't fold.
- for (auto it : llvm::zip(sourceType.getShape(), resultType.getShape())) {
- auto ss = std::get<0>(it), st = std::get<1>(it);
- if (ss != st)
- if (MemRefType::isDynamic(ss) && !MemRefType::isDynamic(st))
- return false;
- }
-
- // If cast is towards more static offset along any dimension, don't fold.
- if (sourceOffset != resultOffset)
- if (MemRefType::isDynamicStrideOrOffset(sourceOffset) &&
- !MemRefType::isDynamicStrideOrOffset(resultOffset))
- return false;
-
- // If cast is towards more static strides along any dimension, don't fold.
- for (auto it : llvm::zip(sourceStrides, resultStrides)) {
- auto ss = std::get<0>(it), st = std::get<1>(it);
- if (ss != st)
- if (MemRefType::isDynamicStrideOrOffset(ss) &&
- !MemRefType::isDynamicStrideOrOffset(st))
- return false;
- }
-
- return true;
-}
-
-namespace {
-/// Pattern to rewrite a subview op with MemRefCast arguments.
-/// This essentially pushes memref_cast past its consuming subview when
-/// `canFoldIntoConsumerOp` is true.
-///
-/// Example:
-/// ```
-/// %0 = memref_cast %V : memref<16x16xf32> to memref<?x?xf32>
-/// %1 = subview %0[0, 0][3, 4][1, 1] :
-/// memref<?x?xf32> to memref<3x4xf32, offset:?, strides:[?, 1]>
-/// ```
-/// is rewritten into:
-/// ```
-/// %0 = subview %V: memref<16x16xf32> to memref<3x4xf32, #[[map0]]>
-/// %1 = memref_cast %0: memref<3x4xf32, offset:0, strides:[16, 1]> to
-/// memref<3x4xf32, offset:?, strides:[?, 1]>
-/// ```
-class SubViewOpMemRefCastFolder final : public OpRewritePattern<SubViewOp> {
-public:
- using OpRewritePattern<SubViewOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(SubViewOp subViewOp,
- PatternRewriter &rewriter) const override {
- // Any constant operand, just return to let SubViewOpConstantFolder kick in.
- if (llvm::any_of(subViewOp.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
- }))
- return failure();
-
- auto castOp = subViewOp.source().getDefiningOp<MemRefCastOp>();
- if (!castOp)
- return failure();
-
- if (!canFoldIntoConsumerOp(castOp))
- return failure();
-
- /// Deduce the resultType of the SubViewOp using `inferSubViewResultType` on
- /// the cast source operand type and the SubViewOp static information. This
- /// is the resulting type if the MemRefCastOp were folded.
- auto resultType = SubViewOp::inferRankReducedResultType(
- subViewOp.getType().getRank(),
- castOp.source().getType().cast<MemRefType>(),
- subViewOp.getMixedOffsets(), subViewOp.getMixedSizes(),
- subViewOp.getMixedStrides());
- Value newSubView = rewriter.create<SubViewOp>(
- subViewOp.getLoc(), resultType, castOp.source(), subViewOp.offsets(),
- subViewOp.sizes(), subViewOp.strides(), subViewOp.static_offsets(),
- subViewOp.static_sizes(), subViewOp.static_strides());
- rewriter.replaceOpWithNewOp<MemRefCastOp>(subViewOp, subViewOp.getType(),
- newSubView);
- return success();
- }
-};
-} // namespace
-
-void SubViewOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<OpWithOffsetSizesAndStridesConstantArgumentFolder<SubViewOp>,
- SubViewOpMemRefCastFolder>(context);
-}
-
-OpFoldResult SubViewOp::fold(ArrayRef<Attribute> operands) {
- auto resultShapedType = getResult().getType().cast<ShapedType>();
- auto sourceShapedType = source().getType().cast<ShapedType>();
-
- if (resultShapedType.hasStaticShape() &&
- resultShapedType == sourceShapedType) {
- return getViewSource();
- }
-
- return {};
-}
-
-//===----------------------------------------------------------------------===//
-// SubTensorOp
-//===----------------------------------------------------------------------===//
-
-/// A subtensor result type can be fully inferred from the source type and the
-/// static representation of offsets, sizes and strides. Special sentinels
-/// encode the dynamic case.
-Type SubTensorOp::inferResultType(RankedTensorType sourceRankedTensorType,
- ArrayRef<int64_t> leadingStaticOffsets,
- ArrayRef<int64_t> leadingStaticSizes,
- ArrayRef<int64_t> leadingStaticStrides) {
- // A subtensor may specify only a leading subset of offset/sizes/strides in
- // which case we complete with offset=0, sizes from memref type and strides=1.
- unsigned rank = sourceRankedTensorType.getRank();
- assert(leadingStaticSizes.size() <= rank &&
- "unexpected leadingStaticSizes overflow");
- auto staticSizes = llvm::to_vector<4>(leadingStaticSizes);
- unsigned numTrailingSizes = rank - staticSizes.size();
- llvm::append_range(staticSizes, sourceRankedTensorType.getShape().take_back(
- numTrailingSizes));
- return RankedTensorType::get(staticSizes,
- sourceRankedTensorType.getElementType());
-}
-
-Type SubTensorOp::inferResultType(RankedTensorType sourceRankedTensorType,
- ArrayRef<OpFoldResult> leadingStaticOffsets,
- ArrayRef<OpFoldResult> leadingStaticSizes,
- ArrayRef<OpFoldResult> leadingStaticStrides) {
- SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
- SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
- dispatchIndexOpFoldResults(leadingStaticOffsets, dynamicOffsets,
- staticOffsets, ShapedType::kDynamicStrideOrOffset);
- dispatchIndexOpFoldResults(leadingStaticSizes, dynamicSizes, staticSizes,
- ShapedType::kDynamicSize);
- dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
- staticStrides, ShapedType::kDynamicStrideOrOffset);
- return SubTensorOp::inferResultType(sourceRankedTensorType, staticOffsets,
- staticSizes, staticStrides);
-}
-
-/// A subtensor result type can be fully inferred from the source type and the
-/// static representation of offsets, sizes and strides. Special sentinels
-/// encode the dynamic case.
-Type SubTensorOp::inferRankReducedResultType(
- unsigned resultRank, RankedTensorType sourceRankedTensorType,
- ArrayRef<int64_t> leadingStaticOffsets,
- ArrayRef<int64_t> leadingStaticSizes,
- ArrayRef<int64_t> leadingStaticStrides) {
- auto inferredType =
- inferResultType(sourceRankedTensorType, leadingStaticOffsets,
- leadingStaticSizes, leadingStaticStrides)
- .cast<RankedTensorType>();
- int rankDiff = inferredType.getRank() - resultRank;
- if (rankDiff > 0) {
- auto shape = inferredType.getShape();
- llvm::SmallDenseSet<unsigned> dimsToProject;
- getPositionsOfShapeOne(rankDiff, shape, dimsToProject);
- SmallVector<int64_t> projectedShape;
- for (unsigned pos = 0, e = shape.size(); pos < e; ++pos)
- if (!dimsToProject.contains(pos))
- projectedShape.push_back(shape[pos]);
- inferredType =
- RankedTensorType::get(projectedShape, inferredType.getElementType());
- }
- return inferredType;
-}
-
-Type SubTensorOp::inferRankReducedResultType(
- unsigned resultRank, RankedTensorType sourceRankedTensorType,
- ArrayRef<OpFoldResult> leadingStaticOffsets,
- ArrayRef<OpFoldResult> leadingStaticSizes,
- ArrayRef<OpFoldResult> leadingStaticStrides) {
- SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
- SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
- dispatchIndexOpFoldResults(leadingStaticOffsets, dynamicOffsets,
- staticOffsets, ShapedType::kDynamicStrideOrOffset);
- dispatchIndexOpFoldResults(leadingStaticSizes, dynamicSizes, staticSizes,
- ShapedType::kDynamicSize);
- dispatchIndexOpFoldResults(leadingStaticStrides, dynamicStrides,
- staticStrides, ShapedType::kDynamicStrideOrOffset);
- return SubTensorOp::inferRankReducedResultType(
- resultRank, sourceRankedTensorType, staticOffsets, staticSizes,
- staticStrides);
-}
-
-// Build a SubTensorOp with mixed static and dynamic entries and custom result
-// type. If the type passed is nullptr, it is inferred.
-void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
- RankedTensorType resultType, Value source,
- ArrayRef<OpFoldResult> offsets,
- ArrayRef<OpFoldResult> sizes,
- ArrayRef<OpFoldResult> strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<int64_t> staticOffsets, staticSizes, staticStrides;
- SmallVector<Value> dynamicOffsets, dynamicSizes, dynamicStrides;
- dispatchIndexOpFoldResults(offsets, dynamicOffsets, staticOffsets,
- ShapedType::kDynamicStrideOrOffset);
- dispatchIndexOpFoldResults(sizes, dynamicSizes, staticSizes,
- ShapedType::kDynamicSize);
- dispatchIndexOpFoldResults(strides, dynamicStrides, staticStrides,
- ShapedType::kDynamicStrideOrOffset);
- auto sourceRankedTensorType = source.getType().cast<RankedTensorType>();
- // Structuring implementation this way avoids duplication between builders.
- if (!resultType) {
- resultType =
- SubTensorOp::inferResultType(sourceRankedTensorType, staticOffsets,
- staticSizes, staticStrides)
- .cast<RankedTensorType>();
- }
- build(b, result, resultType, source, dynamicOffsets, dynamicSizes,
- dynamicStrides, b.getI64ArrayAttr(staticOffsets),
- b.getI64ArrayAttr(staticSizes), b.getI64ArrayAttr(staticStrides));
- result.addAttributes(attrs);
-}
-
-// Build a SubTensorOp with mixed static and dynamic entries and inferred result
-// type.
-void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
- Value source, ArrayRef<OpFoldResult> offsets,
- ArrayRef<OpFoldResult> sizes,
- ArrayRef<OpFoldResult> strides,
- ArrayRef<NamedAttribute> attrs) {
- build(b, result, RankedTensorType(), source, offsets, sizes, strides, attrs);
-}
-
-// Build a SubTensorOp with dynamic entries and custom result type. If the type
-// passed is nullptr, it is inferred.
-void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
- RankedTensorType resultType, Value source,
- ValueRange offsets, ValueRange sizes,
- ValueRange strides,
- ArrayRef<NamedAttribute> attrs) {
- SmallVector<OpFoldResult> offsetValues = llvm::to_vector<4>(
- llvm::map_range(offsets, [](Value v) -> OpFoldResult { return v; }));
- SmallVector<OpFoldResult> sizeValues = llvm::to_vector<4>(
- llvm::map_range(sizes, [](Value v) -> OpFoldResult { return v; }));
- SmallVector<OpFoldResult> strideValues = llvm::to_vector<4>(
- llvm::map_range(strides, [](Value v) -> OpFoldResult { return v; }));
- build(b, result, resultType, source, offsetValues, sizeValues, strideValues);
-}
-
-// Build a SubTensorOp with dynamic entries and inferred result type.
-void mlir::SubTensorOp::build(OpBuilder &b, OperationState &result,
- Value source, ValueRange offsets,
- ValueRange sizes, ValueRange strides,
- ArrayRef<NamedAttribute> attrs) {
- build(b, result, RankedTensorType(), source, offsets, sizes, strides, attrs);
-}
-
-/// Verifier for SubTensorOp.
-static LogicalResult verify(SubTensorOp op) {
- // Verify result type against inferred type.
- auto expectedType = SubTensorOp::inferResultType(
- op.getSourceType(), extractFromI64ArrayAttr(op.static_offsets()),
- extractFromI64ArrayAttr(op.static_sizes()),
- extractFromI64ArrayAttr(op.static_strides()));
- auto result = isRankReducedType(expectedType, op.getType());
- return produceSubViewErrorMsg(result, op, expectedType);
-}
-
-namespace {
-/// Pattern to rewrite a subtensor op with tensor::Cast arguments.
-/// This essentially pushes memref_cast past its consuming subtensor when
-/// `canFoldIntoConsumerOp` is true.
-///
-/// Example:
-/// ```
-/// %0 = tensorcast %V : tensor<16x16xf32> to tensor<?x?xf32>
-/// %1 = subtensor %0[0, 0][3, 4][1, 1] : tensor<?x?xf32> to tensor<3x4xf32>
-/// ```
-/// is rewritten into:
-/// ```
-/// %0 = subtensor %V[0, 0][3, 4][1, 1] : tensor<16x16xf32> to tensor<3x4xf32>
-/// %1 = tensor.cast %0: tensor<3x4xf32> to tensor<3x4xf32>
-/// ```
-class SubTensorOpCastFolder final : public OpRewritePattern<SubTensorOp> {
+/// Pattern to rewrite a subtensor op with tensor::Cast arguments.
+/// This essentially pushes memref_cast past its consuming subtensor when
+/// `canFoldIntoConsumerOp` is true.
+///
+/// Example:
+/// ```
+/// %0 = tensorcast %V : tensor<16x16xf32> to tensor<?x?xf32>
+/// %1 = subtensor %0[0, 0][3, 4][1, 1] : tensor<?x?xf32> to tensor<3x4xf32>
+/// ```
+/// is rewritten into:
+/// ```
+/// %0 = subtensor %V[0, 0][3, 4][1, 1] : tensor<16x16xf32> to tensor<3x4xf32>
+/// %1 = tensor.cast %0: tensor<3x4xf32> to tensor<3x4xf32>
+/// ```
+class SubTensorOpCastFolder final : public OpRewritePattern<SubTensorOp> {
public:
using OpRewritePattern<SubTensorOp>::OpRewritePattern;
@@ -3689,7 +1950,7 @@ class SubTensorOpCastFolder final : public OpRewritePattern<SubTensorOp> {
PatternRewriter &rewriter) const override {
// Any constant operand, just return to let SubViewOpConstantFolder kick in.
if (llvm::any_of(subTensorOp.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
+ return matchPattern(operand, matchConstantIndex());
}))
return failure();
@@ -3721,9 +1982,22 @@ class SubTensorOpCastFolder final : public OpRewritePattern<SubTensorOp> {
};
} // namespace
+/// A canonicalizer wrapper to replace SubTensorOps.
+struct SubTensorCanonicalizer {
+ void operator()(PatternRewriter &rewriter, SubTensorOp op,
+ SubTensorOp newOp) {
+ Value replacement = newOp.getResult();
+ if (replacement.getType() != op.getType())
+ replacement = rewriter.create<tensor::CastOp>(op.getLoc(), op.getType(),
+ replacement);
+ rewriter.replaceOp(op, replacement);
+ }
+};
+
void SubTensorOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
MLIRContext *context) {
- results.insert<OpWithOffsetSizesAndStridesConstantArgumentFolder<SubTensorOp>,
+ results.insert<OpWithOffsetSizesAndStridesConstantArgumentFolder<
+ SubTensorOp, SubTensorCanonicalizer>,
SubTensorOpCastFolder>(context);
}
@@ -3814,12 +2088,13 @@ class SubTensorInsertOpConstantArgumentFolder final
PatternRewriter &rewriter) const override {
// No constant operand, just return.
if (llvm::none_of(subTensorInsertOp.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
+ return matchPattern(operand, matchConstantIndex());
}))
return failure();
// At least one of offsets/sizes/strides is a new constant.
- // Form the new list of operands and constant attributes from the existing.
+ // Form the new list of operands and constant attributes from the
+ // existing.
SmallVector<OpFoldResult> mixedOffsets(subTensorInsertOp.getMixedOffsets());
SmallVector<OpFoldResult> mixedSizes(subTensorInsertOp.getMixedSizes());
SmallVector<OpFoldResult> mixedStrides(subTensorInsertOp.getMixedStrides());
@@ -3856,7 +2131,7 @@ struct SubTensorInsertOpCastFolder final
LogicalResult matchAndRewrite(SubTensorInsertOp subTensorInsertOp,
PatternRewriter &rewriter) const override {
if (llvm::any_of(subTensorInsertOp.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
+ return matchPattern(operand, matchConstantIndex());
}))
return failure();
@@ -3896,175 +2171,6 @@ void SubTensorInsertOp::getCanonicalizationPatterns(
SubTensorInsertOpCastFolder>(context);
}
-//===----------------------------------------------------------------------===//
-// TensorLoadOp
-//===----------------------------------------------------------------------===//
-
-OpFoldResult TensorLoadOp::fold(ArrayRef<Attribute>) {
- if (auto tensorToMemref = memref().getDefiningOp<TensorToMemrefOp>())
- // Approximate alias analysis by conservatively folding only when no there
- // is no interleaved operation.
- if (tensorToMemref->getBlock() == this->getOperation()->getBlock() &&
- tensorToMemref->getNextNode() == this->getOperation())
- return tensorToMemref.tensor();
- return {};
-}
-
-//===----------------------------------------------------------------------===//
-// TensorToMemrefOp
-//===----------------------------------------------------------------------===//
-
-OpFoldResult TensorToMemrefOp::fold(ArrayRef<Attribute>) {
- if (auto tensorLoad = tensor().getDefiningOp<TensorLoadOp>())
- if (tensorLoad.memref().getType() == getType())
- return tensorLoad.memref();
- return {};
-}
-
-namespace {
-/// Replace tensor_cast + tensor_to_memref by tensor_to_memref + memref_cast.
-struct TensorCastToMemref : public OpRewritePattern<TensorToMemrefOp> {
- using OpRewritePattern<TensorToMemrefOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(TensorToMemrefOp tensorToMemRef,
- PatternRewriter &rewriter) const final {
- auto tensorCastOperand =
- tensorToMemRef.getOperand().getDefiningOp<tensor::CastOp>();
- if (!tensorCastOperand)
- return failure();
- auto srcTensorType =
- tensorCastOperand.getOperand().getType().dyn_cast<RankedTensorType>();
- if (!srcTensorType)
- return failure();
- auto memrefType = MemRefType::get(srcTensorType.getShape(),
- srcTensorType.getElementType());
- Value memref = rewriter.create<TensorToMemrefOp>(
- tensorToMemRef.getLoc(), memrefType, tensorCastOperand.getOperand());
- rewriter.replaceOpWithNewOp<MemRefCastOp>(tensorToMemRef,
- tensorToMemRef.getType(), memref);
- return success();
- }
-};
-
-/// Canonicalize tensor_load + tensor_to_memref to memref_cast when type
-/// mismatches prevent `TensorToMemrefOp::fold` to kick in.
-struct TensorLoadToMemref : public OpRewritePattern<TensorToMemrefOp> {
- using OpRewritePattern<TensorToMemrefOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(TensorToMemrefOp tensorToMemRef,
- PatternRewriter &rewriter) const final {
- auto tensorLoad = tensorToMemRef.tensor().getDefiningOp<TensorLoadOp>();
- // Bail unless we have a tensor_load + tensor_to_memref with
diff erent
- // types. `TensorToMemrefOp::fold` handles the same type case.
- if (!tensorLoad ||
- tensorLoad.memref().getType() == tensorToMemRef.getType())
- return failure();
- // If types are not cast-compatible, bail.
- if (!MemRefCastOp::areCastCompatible(tensorLoad.memref().getType(),
- tensorToMemRef.getType()))
- return failure();
- rewriter.replaceOpWithNewOp<MemRefCastOp>(
- tensorToMemRef, tensorToMemRef.getType(), tensorLoad.memref());
- return success();
- }
-};
-} // namespace
-
-void TensorToMemrefOp::getCanonicalizationPatterns(
- OwningRewritePatternList &results, MLIRContext *context) {
- results.insert<TensorCastToMemref, TensorLoadToMemref>(context);
-}
-
-//===----------------------------------------------------------------------===//
-// TransposeOp
-//===----------------------------------------------------------------------===//
-
-/// Build a strided memref type by applying `permutationMap` tp `memRefType`.
-static MemRefType inferTransposeResultType(MemRefType memRefType,
- AffineMap permutationMap) {
- auto rank = memRefType.getRank();
- auto originalSizes = memRefType.getShape();
- // Compute permuted sizes.
- SmallVector<int64_t, 4> sizes(rank, 0);
- for (auto en : llvm::enumerate(permutationMap.getResults()))
- sizes[en.index()] =
- originalSizes[en.value().cast<AffineDimExpr>().getPosition()];
-
- // Compute permuted strides.
- int64_t offset;
- SmallVector<int64_t, 4> strides;
- auto res = getStridesAndOffset(memRefType, strides, offset);
- assert(succeeded(res) && strides.size() == static_cast<unsigned>(rank));
- (void)res;
- auto map =
- makeStridedLinearLayoutMap(strides, offset, memRefType.getContext());
- map = permutationMap ? map.compose(permutationMap) : map;
- return MemRefType::Builder(memRefType).setShape(sizes).setAffineMaps(map);
-}
-
-void TransposeOp::build(OpBuilder &b, OperationState &result, Value in,
- AffineMapAttr permutation,
- ArrayRef<NamedAttribute> attrs) {
- auto permutationMap = permutation.getValue();
- assert(permutationMap);
-
- auto memRefType = in.getType().cast<MemRefType>();
- // Compute result type.
- MemRefType resultType = inferTransposeResultType(memRefType, permutationMap);
-
- build(b, result, resultType, in, attrs);
- result.addAttribute(TransposeOp::getPermutationAttrName(), permutation);
-}
-
-// transpose $in $permutation attr-dict : type($in) `to` type(results)
-static void print(OpAsmPrinter &p, TransposeOp op) {
- p << "transpose " << op.in() << " " << op.permutation();
- p.printOptionalAttrDict(op->getAttrs(),
- {TransposeOp::getPermutationAttrName()});
- p << " : " << op.in().getType() << " to " << op.getType();
-}
-
-static ParseResult parseTransposeOp(OpAsmParser &parser,
- OperationState &result) {
- OpAsmParser::OperandType in;
- AffineMap permutation;
- MemRefType srcType, dstType;
- if (parser.parseOperand(in) || parser.parseAffineMap(permutation) ||
- parser.parseOptionalAttrDict(result.attributes) ||
- parser.parseColonType(srcType) ||
- parser.resolveOperand(in, srcType, result.operands) ||
- parser.parseKeywordType("to", dstType) ||
- parser.addTypeToList(dstType, result.types))
- return failure();
-
- result.addAttribute(TransposeOp::getPermutationAttrName(),
- AffineMapAttr::get(permutation));
- return success();
-}
-
-static LogicalResult verify(TransposeOp op) {
- if (!op.permutation().isPermutation())
- return op.emitOpError("expected a permutation map");
- if (op.permutation().getNumDims() != op.getShapedType().getRank())
- return op.emitOpError(
- "expected a permutation map of same rank as the input");
-
- auto srcType = op.in().getType().cast<MemRefType>();
- auto dstType = op.getType().cast<MemRefType>();
- auto transposedType = inferTransposeResultType(srcType, op.permutation());
- if (dstType != transposedType)
- return op.emitOpError("output type ")
- << dstType << " does not match transposed input type " << srcType
- << ", " << transposedType;
- return success();
-}
-
-OpFoldResult TransposeOp::fold(ArrayRef<Attribute>) {
- if (succeeded(foldMemRefCast(*this)))
- return getResult();
- return {};
-}
-
//===----------------------------------------------------------------------===//
// TruncateIOp
//===----------------------------------------------------------------------===//
@@ -4141,172 +2247,6 @@ OpFoldResult UnsignedRemIOp::fold(ArrayRef<Attribute> operands) {
return IntegerAttr::get(lhs.getType(), lhs.getValue().urem(rhsValue));
}
-//===----------------------------------------------------------------------===//
-// ViewOp
-//===----------------------------------------------------------------------===//
-
-static ParseResult parseViewOp(OpAsmParser &parser, OperationState &result) {
- OpAsmParser::OperandType srcInfo;
- SmallVector<OpAsmParser::OperandType, 1> offsetInfo;
- SmallVector<OpAsmParser::OperandType, 4> sizesInfo;
- auto indexType = parser.getBuilder().getIndexType();
- Type srcType, dstType;
- llvm::SMLoc offsetLoc;
- if (parser.parseOperand(srcInfo) || parser.getCurrentLocation(&offsetLoc) ||
- parser.parseOperandList(offsetInfo, OpAsmParser::Delimiter::Square))
- return failure();
-
- if (offsetInfo.size() != 1)
- return parser.emitError(offsetLoc) << "expects 1 offset operand";
-
- return failure(
- parser.parseOperandList(sizesInfo, OpAsmParser::Delimiter::Square) ||
- parser.parseOptionalAttrDict(result.attributes) ||
- parser.parseColonType(srcType) ||
- parser.resolveOperand(srcInfo, srcType, result.operands) ||
- parser.resolveOperands(offsetInfo, indexType, result.operands) ||
- parser.resolveOperands(sizesInfo, indexType, result.operands) ||
- parser.parseKeywordType("to", dstType) ||
- parser.addTypeToList(dstType, result.types));
-}
-
-static void print(OpAsmPrinter &p, ViewOp op) {
- p << op.getOperationName() << ' ' << op.getOperand(0) << '[';
- p.printOperand(op.byte_shift());
- p << "][" << op.sizes() << ']';
- p.printOptionalAttrDict(op->getAttrs());
- p << " : " << op.getOperand(0).getType() << " to " << op.getType();
-}
-
-static LogicalResult verify(ViewOp op) {
- auto baseType = op.getOperand(0).getType().cast<MemRefType>();
- auto viewType = op.getType();
-
- // The base memref should have identity layout map (or none).
- if (baseType.getAffineMaps().size() > 1 ||
- (baseType.getAffineMaps().size() == 1 &&
- !baseType.getAffineMaps()[0].isIdentity()))
- return op.emitError("unsupported map for base memref type ") << baseType;
-
- // The result memref should have identity layout map (or none).
- if (viewType.getAffineMaps().size() > 1 ||
- (viewType.getAffineMaps().size() == 1 &&
- !viewType.getAffineMaps()[0].isIdentity()))
- return op.emitError("unsupported map for result memref type ") << viewType;
-
- // The base memref and the view memref should be in the same memory space.
- if (baseType.getMemorySpaceAsInt() != viewType.getMemorySpaceAsInt())
- return op.emitError("
diff erent memory spaces specified for base memref "
- "type ")
- << baseType << " and view memref type " << viewType;
-
- // Verify that we have the correct number of sizes for the result type.
- unsigned numDynamicDims = viewType.getNumDynamicDims();
- if (op.sizes().size() != numDynamicDims)
- return op.emitError("incorrect number of size operands for type ")
- << viewType;
-
- return success();
-}
-
-Value ViewOp::getViewSource() { return source(); }
-
-namespace {
-
-struct ViewOpShapeFolder : public OpRewritePattern<ViewOp> {
- using OpRewritePattern<ViewOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(ViewOp viewOp,
- PatternRewriter &rewriter) const override {
- // Return if none of the operands are constants.
- if (llvm::none_of(viewOp.getOperands(), [](Value operand) {
- return matchPattern(operand, m_ConstantIndex());
- }))
- return failure();
-
- // Get result memref type.
- auto memrefType = viewOp.getType();
-
- // Get offset from old memref view type 'memRefType'.
- int64_t oldOffset;
- SmallVector<int64_t, 4> oldStrides;
- if (failed(getStridesAndOffset(memrefType, oldStrides, oldOffset)))
- return failure();
- assert(oldOffset == 0 && "Expected 0 offset");
-
- SmallVector<Value, 4> newOperands;
-
- // Offset cannot be folded into result type.
-
- // Fold any dynamic dim operands which are produced by a constant.
- SmallVector<int64_t, 4> newShapeConstants;
- newShapeConstants.reserve(memrefType.getRank());
-
- unsigned dynamicDimPos = 0;
- unsigned rank = memrefType.getRank();
- for (unsigned dim = 0, e = rank; dim < e; ++dim) {
- int64_t dimSize = memrefType.getDimSize(dim);
- // If this is already static dimension, keep it.
- if (!ShapedType::isDynamic(dimSize)) {
- newShapeConstants.push_back(dimSize);
- continue;
- }
- auto *defOp = viewOp.sizes()[dynamicDimPos].getDefiningOp();
- if (auto constantIndexOp = dyn_cast_or_null<ConstantIndexOp>(defOp)) {
- // Dynamic shape dimension will be folded.
- newShapeConstants.push_back(constantIndexOp.getValue());
- } else {
- // Dynamic shape dimension not folded; copy operand from old memref.
- newShapeConstants.push_back(dimSize);
- newOperands.push_back(viewOp.sizes()[dynamicDimPos]);
- }
- dynamicDimPos++;
- }
-
- // Create new memref type with constant folded dims.
- MemRefType newMemRefType =
- MemRefType::Builder(memrefType).setShape(newShapeConstants);
- // Nothing new, don't fold.
- if (newMemRefType == memrefType)
- return failure();
-
- // Create new ViewOp.
- auto newViewOp = rewriter.create<ViewOp>(viewOp.getLoc(), newMemRefType,
- viewOp.getOperand(0),
- viewOp.byte_shift(), newOperands);
- // Insert a cast so we have the same type as the old memref type.
- rewriter.replaceOpWithNewOp<MemRefCastOp>(viewOp, newViewOp,
- viewOp.getType());
- return success();
- }
-};
-
-struct ViewOpMemrefCastFolder : public OpRewritePattern<ViewOp> {
- using OpRewritePattern<ViewOp>::OpRewritePattern;
-
- LogicalResult matchAndRewrite(ViewOp viewOp,
- PatternRewriter &rewriter) const override {
- Value memrefOperand = viewOp.getOperand(0);
- MemRefCastOp memrefCastOp = memrefOperand.getDefiningOp<MemRefCastOp>();
- if (!memrefCastOp)
- return failure();
- Value allocOperand = memrefCastOp.getOperand();
- AllocOp allocOp = allocOperand.getDefiningOp<AllocOp>();
- if (!allocOp)
- return failure();
- rewriter.replaceOpWithNewOp<ViewOp>(viewOp, viewOp.getType(), allocOperand,
- viewOp.byte_shift(), viewOp.sizes());
- return success();
- }
-};
-
-} // end anonymous namespace
-
-void ViewOp::getCanonicalizationPatterns(OwningRewritePatternList &results,
- MLIRContext *context) {
- results.insert<ViewOpShapeFolder, ViewOpMemrefCastFolder>(context);
-}
-
//===----------------------------------------------------------------------===//
// XOrOp
//===----------------------------------------------------------------------===//
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/Bufferize.cpp b/mlir/lib/Dialect/StandardOps/Transforms/Bufferize.cpp
index 2a3a464cd0a8..c2b9c938e0b8 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/StandardOps/Transforms/Bufferize.cpp
@@ -12,6 +12,7 @@
#include "mlir/Transforms/Bufferize.h"
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/Passes.h"
@@ -21,15 +22,15 @@
using namespace mlir;
namespace {
-class BufferizeDimOp : public OpConversionPattern<DimOp> {
+class BufferizeDimOp : public OpConversionPattern<memref::DimOp> {
public:
using OpConversionPattern::OpConversionPattern;
LogicalResult
- matchAndRewrite(DimOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::DimOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- DimOp::Adaptor adaptor(operands);
- rewriter.replaceOpWithNewOp<DimOp>(op, adaptor.memrefOrTensor(),
- adaptor.index());
+ memref::DimOp::Adaptor adaptor(operands);
+ rewriter.replaceOpWithNewOp<memref::DimOp>(op, adaptor.memrefOrTensor(),
+ adaptor.index());
return success();
}
};
@@ -67,6 +68,7 @@ struct StdBufferizePass : public StdBufferizeBase<StdBufferizePass> {
OwningRewritePatternList patterns;
ConversionTarget target(*context);
+ target.addLegalDialect<memref::MemRefDialect>();
target.addLegalDialect<StandardOpsDialect>();
target.addLegalDialect<scf::SCFDialect>();
@@ -78,8 +80,8 @@ struct StdBufferizePass : public StdBufferizeBase<StdBufferizePass> {
return typeConverter.isLegal(op.getType()) ||
!op.condition().getType().isa<IntegerType>();
});
- target.addDynamicallyLegalOp<DimOp>(
- [&](DimOp op) { return typeConverter.isLegal(op); });
+ target.addDynamicallyLegalOp<memref::DimOp>(
+ [&](memref::DimOp op) { return typeConverter.isLegal(op); });
if (failed(
applyPartialConversion(getFunction(), target, std::move(patterns))))
signalPassFailure();
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/CMakeLists.txt b/mlir/lib/Dialect/StandardOps/Transforms/CMakeLists.txt
index d6e2f028a5a4..b01ff954d7f3 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/StandardOps/Transforms/CMakeLists.txt
@@ -14,6 +14,7 @@ add_mlir_dialect_library(MLIRStandardOpsTransforms
LINK_LIBS PUBLIC
MLIRIR
+ MLIRMemRef
MLIRPass
MLIRSCF
MLIRStandard
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/ExpandOps.cpp b/mlir/lib/Dialect/StandardOps/Transforms/ExpandOps.cpp
index 9d1378f06bcb..98b261c8044b 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/ExpandOps.cpp
+++ b/mlir/lib/Dialect/StandardOps/Transforms/ExpandOps.cpp
@@ -13,6 +13,7 @@
//===----------------------------------------------------------------------===//
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/Passes.h"
#include "mlir/IR/PatternMatch.h"
@@ -70,13 +71,13 @@ struct AtomicRMWOpConverter : public OpRewritePattern<AtomicRMWOp> {
}
};
-/// Converts `memref_reshape` that has a target shape of a statically-known
-/// size to `memref_reinterpret_cast`.
-struct MemRefReshapeOpConverter : public OpRewritePattern<MemRefReshapeOp> {
+/// Converts `memref.reshape` that has a target shape of a statically-known
+/// size to `memref.reinterpret_cast`.
+struct MemRefReshapeOpConverter : public OpRewritePattern<memref::ReshapeOp> {
public:
using OpRewritePattern::OpRewritePattern;
- LogicalResult matchAndRewrite(MemRefReshapeOp op,
+ LogicalResult matchAndRewrite(memref::ReshapeOp op,
PatternRewriter &rewriter) const final {
auto shapeType = op.shape().getType().cast<MemRefType>();
if (!shapeType.hasStaticShape())
@@ -91,7 +92,7 @@ struct MemRefReshapeOpConverter : public OpRewritePattern<MemRefReshapeOp> {
Value stride = rewriter.create<ConstantIndexOp>(loc, 1);
for (int i = rank - 1; i >= 0; --i) {
Value index = rewriter.create<ConstantIndexOp>(loc, i);
- Value size = rewriter.create<LoadOp>(loc, op.shape(), index);
+ Value size = rewriter.create<memref::LoadOp>(loc, op.shape(), index);
if (!size.getType().isa<IndexType>())
size = rewriter.create<IndexCastOp>(loc, size, rewriter.getIndexType());
sizes[i] = size;
@@ -99,7 +100,7 @@ struct MemRefReshapeOpConverter : public OpRewritePattern<MemRefReshapeOp> {
if (i > 0)
stride = rewriter.create<MulIOp>(loc, stride, size);
}
- rewriter.replaceOpWithNewOp<MemRefReinterpretCastOp>(
+ rewriter.replaceOpWithNewOp<memref::ReinterpretCastOp>(
op, op.getType(), op.source(), /*offset=*/rewriter.getIndexAttr(0),
sizes, strides);
return success();
@@ -215,12 +216,12 @@ struct StdExpandOpsPass : public StdExpandOpsBase<StdExpandOpsPass> {
ConversionTarget target(getContext());
- target.addLegalDialect<StandardOpsDialect>();
+ target.addLegalDialect<memref::MemRefDialect, StandardOpsDialect>();
target.addDynamicallyLegalOp<AtomicRMWOp>([](AtomicRMWOp op) {
return op.kind() != AtomicRMWKind::maxf &&
op.kind() != AtomicRMWKind::minf;
});
- target.addDynamicallyLegalOp<MemRefReshapeOp>([](MemRefReshapeOp op) {
+ target.addDynamicallyLegalOp<memref::ReshapeOp>([](memref::ReshapeOp op) {
return !op.shape().getType().cast<MemRefType>().hasStaticShape();
});
target.addIllegalOp<SignedCeilDivIOp>();
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/FuncBufferize.cpp b/mlir/lib/Dialect/StandardOps/Transforms/FuncBufferize.cpp
index 4b5a2d632670..d38a564ac2a9 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/FuncBufferize.cpp
+++ b/mlir/lib/Dialect/StandardOps/Transforms/FuncBufferize.cpp
@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/FuncConversions.h"
#include "mlir/Dialect/StandardOps/Transforms/Passes.h"
@@ -22,7 +23,6 @@ using namespace mlir;
namespace {
struct FuncBufferizePass : public FuncBufferizeBase<FuncBufferizePass> {
using FuncBufferizeBase<FuncBufferizePass>::FuncBufferizeBase;
-
void runOnOperation() override {
auto module = getOperation();
auto *context = &getContext();
@@ -43,8 +43,8 @@ struct FuncBufferizePass : public FuncBufferizeBase<FuncBufferizePass> {
populateBranchOpInterfaceTypeConversionPattern(patterns, context,
typeConverter);
populateReturnOpTypeConversionPattern(patterns, context, typeConverter);
- target.addLegalOp<ModuleOp, ModuleTerminatorOp, TensorLoadOp,
- TensorToMemrefOp>();
+ target.addLegalOp<ModuleOp, ModuleTerminatorOp, memref::TensorLoadOp,
+ memref::BufferCastOp>();
target.markUnknownOpDynamicallyLegal([&](Operation *op) {
return isNotBranchOpInterfaceOrReturnLikeOp(op) ||
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/PassDetail.h b/mlir/lib/Dialect/StandardOps/Transforms/PassDetail.h
index 9e3bbbb45462..7cff00a0a06c 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/PassDetail.h
+++ b/mlir/lib/Dialect/StandardOps/Transforms/PassDetail.h
@@ -16,6 +16,10 @@ namespace mlir {
class AtomicRMWOp;
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
#define GEN_PASS_CLASSES
#include "mlir/Dialect/StandardOps/Transforms/Passes.h.inc"
diff --git a/mlir/lib/Dialect/StandardOps/Transforms/TensorConstantBufferize.cpp b/mlir/lib/Dialect/StandardOps/Transforms/TensorConstantBufferize.cpp
index a024d6c1f58b..18c3be94685b 100644
--- a/mlir/lib/Dialect/StandardOps/Transforms/TensorConstantBufferize.cpp
+++ b/mlir/lib/Dialect/StandardOps/Transforms/TensorConstantBufferize.cpp
@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/Passes.h"
#include "mlir/IR/BlockAndValueMapping.h"
@@ -26,13 +27,13 @@ namespace {
class GlobalCreator {
public:
explicit GlobalCreator(ModuleOp module);
- GlobalMemrefOp getGlobalFor(Attribute attr) {
+ memref::GlobalOp getGlobalFor(Attribute attr) {
assert(globals.find(attr) != globals.end() && "unknown constant attr");
return globals[attr];
}
private:
- DenseMap<Attribute, GlobalMemrefOp> globals;
+ DenseMap<Attribute, memref::GlobalOp> globals;
};
GlobalCreator::GlobalCreator(ModuleOp module) {
@@ -58,7 +59,7 @@ GlobalCreator::GlobalCreator(ModuleOp module) {
interleave(type.getShape(), os, "x");
os << "x" << type.getElementType();
- auto global = globalBuilder.create<GlobalMemrefOp>(
+ auto global = globalBuilder.create<memref::GlobalOp>(
op.getLoc(), (Twine("__constant_") + os.str()).str(),
/*sym_visibility=*/globalBuilder.getStringAttr("private"),
/*type=*/typeConverter.convertType(type),
@@ -89,8 +90,8 @@ class BufferizeTensorConstantOp : public OpConversionPattern<ConstantOp> {
return failure();
auto globalMemref = globals.getGlobalFor(op.value());
- rewriter.replaceOpWithNewOp<GetGlobalMemrefOp>(op, globalMemref.type(),
- globalMemref.getName());
+ rewriter.replaceOpWithNewOp<memref::GetGlobalOp>(op, globalMemref.type(),
+ globalMemref.getName());
return success();
}
GlobalCreator &globals;
@@ -109,7 +110,7 @@ struct TensorConstantBufferizePass
OwningRewritePatternList patterns;
ConversionTarget target(*context);
- target.addLegalDialect<StandardOpsDialect>();
+ target.addLegalDialect<memref::MemRefDialect>();
patterns.insert<BufferizeTensorConstantOp>(globals, typeConverter, context);
target.addDynamicallyLegalOp<ConstantOp>(
[&](ConstantOp op) { return typeConverter.isLegal(op.getType()); });
diff --git a/mlir/lib/Dialect/StandardOps/Utils/Utils.cpp b/mlir/lib/Dialect/StandardOps/Utils/Utils.cpp
index d01eb130b543..6036dfa6fd2c 100644
--- a/mlir/lib/Dialect/StandardOps/Utils/Utils.cpp
+++ b/mlir/lib/Dialect/StandardOps/Utils/Utils.cpp
@@ -12,6 +12,7 @@
#include "mlir/Dialect/StandardOps/Utils/Utils.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
using namespace mlir;
@@ -21,8 +22,41 @@ SmallVector<Value, 4> mlir::getDynOperands(Location loc, Value val,
SmallVector<Value, 4> dynOperands;
auto shapedType = val.getType().cast<ShapedType>();
for (auto dim : llvm::enumerate(shapedType.getShape())) {
- if (dim.value() == TensorType::kDynamicSize)
- dynOperands.push_back(b.create<DimOp>(loc, val, dim.index()));
+ if (dim.value() == MemRefType::kDynamicSize)
+ dynOperands.push_back(b.create<memref::DimOp>(loc, val, dim.index()));
}
return dynOperands;
}
+
+/// Matches a ConstantIndexOp.
+/// TODO: This should probably just be a general matcher that uses matchConstant
+/// and checks the operation for an index type.
+detail::op_matcher<ConstantIndexOp> mlir::matchConstantIndex() {
+ return detail::op_matcher<ConstantIndexOp>();
+}
+
+/// Detects the `values` produced by a ConstantIndexOp and places the new
+/// constant in place of the corresponding sentinel value.
+void mlir::canonicalizeSubViewPart(
+ SmallVectorImpl<OpFoldResult> &values,
+ llvm::function_ref<bool(int64_t)> isDynamic) {
+ for (OpFoldResult &ofr : values) {
+ if (ofr.is<Attribute>())
+ continue;
+ // Newly static, move from Value to constant.
+ if (auto cstOp = ofr.dyn_cast<Value>().getDefiningOp<ConstantIndexOp>())
+ ofr = OpBuilder(cstOp).getIndexAttr(cstOp.getValue());
+ }
+}
+
+void mlir::getPositionsOfShapeOne(
+ unsigned rank, ArrayRef<int64_t> shape,
+ llvm::SmallDenseSet<unsigned> &dimsToProject) {
+ dimsToProject.reserve(rank);
+ for (unsigned pos = 0, e = shape.size(); pos < e && rank > 0; ++pos) {
+ if (shape[pos] == 1) {
+ dimsToProject.insert(pos);
+ --rank;
+ }
+ }
+}
diff --git a/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
index 66de78758692..1ef742e84424 100644
--- a/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
@@ -12,6 +12,7 @@
#include "mlir/Transforms/Bufferize.h"
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
@@ -28,7 +29,7 @@ class BufferizeCastOp : public OpConversionPattern<tensor::CastOp> {
matchAndRewrite(tensor::CastOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
auto resultType = getTypeConverter()->convertType(op.getType());
- rewriter.replaceOpWithNewOp<MemRefCastOp>(op, resultType, operands[0]);
+ rewriter.replaceOpWithNewOp<memref::CastOp>(op, resultType, operands[0]);
return success();
}
};
@@ -42,8 +43,8 @@ class BufferizeExtractOp : public OpConversionPattern<tensor::ExtractOp> {
matchAndRewrite(tensor::ExtractOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
tensor::ExtractOp::Adaptor adaptor(operands);
- rewriter.replaceOpWithNewOp<LoadOp>(op, adaptor.tensor(),
- adaptor.indices());
+ rewriter.replaceOpWithNewOp<memref::LoadOp>(op, adaptor.tensor(),
+ adaptor.indices());
return success();
}
};
@@ -60,11 +61,12 @@ class BufferizeFromElementsOp
int numberOfElements = op.elements().size();
auto resultType = MemRefType::get(
{numberOfElements}, op.getType().cast<TensorType>().getElementType());
- Value result = rewriter.create<AllocOp>(op.getLoc(), resultType);
+ Value result = rewriter.create<memref::AllocOp>(op.getLoc(), resultType);
for (auto element : llvm::enumerate(op.elements())) {
Value index =
rewriter.create<ConstantIndexOp>(op.getLoc(), element.index());
- rewriter.create<StoreOp>(op.getLoc(), element.value(), result, index);
+ rewriter.create<memref::StoreOp>(op.getLoc(), element.value(), result,
+ index);
}
rewriter.replaceOp(op, {result});
return success();
@@ -86,8 +88,8 @@ class BufferizeGenerateOp : public OpConversionPattern<tensor::GenerateOp> {
RankedTensorType tensorType = op.getType().cast<RankedTensorType>();
MemRefType memrefType =
MemRefType::get(tensorType.getShape(), tensorType.getElementType());
- Value result =
- rewriter.create<AllocOp>(loc, memrefType, transformed.dynamicExtents());
+ Value result = rewriter.create<memref::AllocOp>(
+ loc, memrefType, transformed.dynamicExtents());
// Collect loop bounds.
int64_t rank = tensorType.getRank();
@@ -125,9 +127,9 @@ class BufferizeGenerateOp : public OpConversionPattern<tensor::GenerateOp> {
// about creating that.
Operation *elementYield = parallelBody->getTerminator()->getPrevNode();
rewriter.setInsertionPointAfter(elementYield);
- rewriter.replaceOpWithNewOp<StoreOp>(elementYield,
- elementYield->getOperands()[0], result,
- parallelBody->getArguments());
+ rewriter.replaceOpWithNewOp<memref::StoreOp>(
+ elementYield, elementYield->getOperands()[0], result,
+ parallelBody->getArguments());
rewriter.replaceOp(op, {result});
return success();
@@ -155,6 +157,7 @@ struct TensorBufferizePass : public TensorBufferizeBase<TensorBufferizePass> {
populateTensorBufferizePatterns(context, typeConverter, patterns);
target.addIllegalOp<tensor::CastOp, tensor::ExtractOp,
tensor::FromElementsOp, tensor::GenerateOp>();
+ target.addLegalDialect<memref::MemRefDialect>();
target.addLegalDialect<StandardOpsDialect>();
target.addLegalDialect<scf::SCFDialect>();
diff --git a/mlir/lib/Dialect/Tensor/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Tensor/Transforms/CMakeLists.txt
index 6d29bd56dca6..f9d9047d75af 100644
--- a/mlir/lib/Dialect/Tensor/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Tensor/Transforms/CMakeLists.txt
@@ -9,6 +9,7 @@ add_mlir_dialect_library(MLIRTensorTransforms
LINK_LIBS PUBLIC
MLIRIR
+ MLIRMemRef
MLIRPass
MLIRSCF
MLIRTensor
diff --git a/mlir/lib/Dialect/Vector/CMakeLists.txt b/mlir/lib/Dialect/Vector/CMakeLists.txt
index f3e9af2b0c0e..1c895f950c28 100644
--- a/mlir/lib/Dialect/Vector/CMakeLists.txt
+++ b/mlir/lib/Dialect/Vector/CMakeLists.txt
@@ -20,6 +20,7 @@ add_mlir_dialect_library(MLIRVector
MLIRStandard
MLIRAffine
MLIRLinalg
+ MLIRMemRef
MLIRSCF
MLIRLoopAnalysis
MLIRSideEffectInterfaces
diff --git a/mlir/lib/Dialect/Vector/VectorOps.cpp b/mlir/lib/Dialect/Vector/VectorOps.cpp
index 39d016b8d0fe..6ca28ba681ef 100644
--- a/mlir/lib/Dialect/Vector/VectorOps.cpp
+++ b/mlir/lib/Dialect/Vector/VectorOps.cpp
@@ -12,6 +12,7 @@
//===----------------------------------------------------------------------===//
#include "mlir/Dialect/Vector/VectorOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Tensor/IR/Tensor.h"
#include "mlir/Dialect/Utils/StructuredOpsUtils.h"
@@ -2396,12 +2397,12 @@ static LogicalResult verify(TransferReadOp op) {
/// ```
/// someop(memrefcast) -> someop
/// ```
-/// It folds the source of the memref_cast into the root operation directly.
+/// It folds the source of the memref.cast into the root operation directly.
static LogicalResult foldMemRefCast(Operation *op) {
bool folded = false;
for (OpOperand &operand : op->getOpOperands()) {
- auto castOp = operand.get().getDefiningOp<MemRefCastOp>();
- if (castOp && canFoldIntoConsumerOp(castOp)) {
+ auto castOp = operand.get().getDefiningOp<memref::CastOp>();
+ if (castOp && memref::CastOp::canFoldIntoConsumerOp(castOp)) {
operand.set(castOp.getOperand());
folded = true;
}
diff --git a/mlir/lib/Dialect/Vector/VectorTransforms.cpp b/mlir/lib/Dialect/Vector/VectorTransforms.cpp
index 090afda01fe4..57602a562d9a 100644
--- a/mlir/lib/Dialect/Vector/VectorTransforms.cpp
+++ b/mlir/lib/Dialect/Vector/VectorTransforms.cpp
@@ -16,6 +16,8 @@
#include "mlir/Dialect/Affine/EDSC/Intrinsics.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Linalg/EDSC/Intrinsics.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
@@ -2282,7 +2284,7 @@ static Value createScopedInBoundsCond(VectorTransferOpInterface xferOp) {
// Fold or create the check that `index + vector_size` <= `memref_size`.
Value sum = xferOp.indices()[indicesIdx] + std_constant_index(vectorSize);
Value cond =
- createScopedFoldedSLE(sum, std_dim(xferOp.source(), indicesIdx));
+ createScopedFoldedSLE(sum, memref_dim(xferOp.source(), indicesIdx));
if (!cond)
return;
// Conjunction over all dims for which we are in-bounds.
@@ -2317,7 +2319,7 @@ LogicalResult mlir::vector::splitFullAndPartialTransferPrecondition(
/// b. using a dynamic shape and/or stride for the dimensions that don't
/// agree.
static MemRefType getCastCompatibleMemRefType(MemRefType aT, MemRefType bT) {
- if (MemRefCastOp::areCastCompatible(aT, bT))
+ if (memref::CastOp::areCastCompatible(aT, bT))
return aT;
if (aT.getRank() != bT.getRank())
return MemRefType();
@@ -2362,8 +2364,8 @@ static Value createScopedSubViewIntersection(VectorTransferOpInterface xferOp,
sizes.append(leadingIndices.begin(), leadingIndices.end());
xferOp.zipResultAndIndexing([&](int64_t resultIdx, int64_t indicesIdx) {
using MapList = ArrayRef<ArrayRef<AffineExpr>>;
- Value dimMemRef = std_dim(xferOp.source(), indicesIdx);
- Value dimAlloc = std_dim(alloc, resultIdx);
+ Value dimMemRef = memref_dim(xferOp.source(), indicesIdx);
+ Value dimAlloc = memref_dim(alloc, resultIdx);
Value index = xferOp.indices()[indicesIdx];
AffineExpr i, j, k;
bindDims(xferOp.getContext(), i, j, k);
@@ -2377,7 +2379,7 @@ static Value createScopedSubViewIntersection(VectorTransferOpInterface xferOp,
SmallVector<OpFoldResult, 4> indices = llvm::to_vector<4>(llvm::map_range(
xferOp.indices(), [](Value idx) -> OpFoldResult { return idx; }));
- return std_sub_view(
+ return memref_sub_view(
xferOp.source(), indices, sizes,
SmallVector<OpFoldResult>(memrefRank, OpBuilder(xferOp).getIndexAttr(1)));
}
@@ -2388,13 +2390,13 @@ static Value createScopedSubViewIntersection(VectorTransferOpInterface xferOp,
/// Produce IR resembling:
/// ```
/// %1:3 = scf.if (%inBounds) {
-/// memref_cast %A: memref<A...> to compatibleMemRefType
+/// memref.cast %A: memref<A...> to compatibleMemRefType
/// scf.yield %view, ... : compatibleMemRefType, index, index
/// } else {
/// %2 = linalg.fill(%alloc, %pad)
/// %3 = subview %view [...][...][...]
/// linalg.copy(%3, %alloc)
-/// memref_cast %alloc: memref<B...> to compatibleMemRefType
+/// memref.cast %alloc: memref<B...> to compatibleMemRefType
/// scf.yield %4, ... : compatibleMemRefType, index, index
/// }
/// ```
@@ -2412,7 +2414,7 @@ static scf::IfOp createScopedFullPartialLinalgCopy(
[&]() -> scf::ValueVector {
Value res = memref;
if (compatibleMemRefType != xferOp.getShapedType())
- res = std_memref_cast(memref, compatibleMemRefType);
+ res = memref_cast(memref, compatibleMemRefType);
scf::ValueVector viewAndIndices{res};
viewAndIndices.insert(viewAndIndices.end(), xferOp.indices().begin(),
xferOp.indices().end());
@@ -2425,7 +2427,7 @@ static scf::IfOp createScopedFullPartialLinalgCopy(
Value memRefSubView = createScopedSubViewIntersection(
cast<VectorTransferOpInterface>(xferOp.getOperation()), alloc);
linalg_copy(memRefSubView, alloc);
- Value casted = std_memref_cast(alloc, compatibleMemRefType);
+ Value casted = memref_cast(alloc, compatibleMemRefType);
scf::ValueVector viewAndIndices{casted};
viewAndIndices.insert(viewAndIndices.end(), xferOp.getTransferRank(),
zero);
@@ -2441,14 +2443,14 @@ static scf::IfOp createScopedFullPartialLinalgCopy(
/// Produce IR resembling:
/// ```
/// %1:3 = scf.if (%inBounds) {
-/// memref_cast %A: memref<A...> to compatibleMemRefType
+/// memref.cast %A: memref<A...> to compatibleMemRefType
/// scf.yield %view, ... : compatibleMemRefType, index, index
/// } else {
/// %2 = vector.transfer_read %view[...], %pad : memref<A...>, vector<...>
/// %3 = vector.type_cast %extra_alloc :
/// memref<...> to memref<vector<...>>
/// store %2, %3[] : memref<vector<...>>
-/// %4 = memref_cast %alloc: memref<B...> to compatibleMemRefType
+/// %4 = memref.cast %alloc: memref<B...> to compatibleMemRefType
/// scf.yield %4, ... : compatibleMemRefType, index, index
/// }
/// ```
@@ -2466,7 +2468,7 @@ static scf::IfOp createScopedFullPartialVectorTransferRead(
[&]() -> scf::ValueVector {
Value res = memref;
if (compatibleMemRefType != xferOp.getShapedType())
- res = std_memref_cast(memref, compatibleMemRefType);
+ res = memref_cast(memref, compatibleMemRefType);
scf::ValueVector viewAndIndices{res};
viewAndIndices.insert(viewAndIndices.end(), xferOp.indices().begin(),
xferOp.indices().end());
@@ -2476,10 +2478,10 @@ static scf::IfOp createScopedFullPartialVectorTransferRead(
Operation *newXfer =
ScopedContext::getBuilderRef().clone(*xferOp.getOperation());
Value vector = cast<VectorTransferOpInterface>(newXfer).vector();
- std_store(vector, vector_type_cast(
- MemRefType::get({}, vector.getType()), alloc));
+ memref_store(vector, vector_type_cast(
+ MemRefType::get({}, vector.getType()), alloc));
- Value casted = std_memref_cast(alloc, compatibleMemRefType);
+ Value casted = memref_cast(alloc, compatibleMemRefType);
scf::ValueVector viewAndIndices{casted};
viewAndIndices.insert(viewAndIndices.end(), xferOp.getTransferRank(),
zero);
@@ -2506,11 +2508,11 @@ static scf::IfOp createScopedFullPartialVectorTransferRead(
/// ```
/// %1:3 = scf.if (%inBounds) {
/// // fastpath, direct cast
-/// memref_cast %A: memref<A...> to compatibleMemRefType
+/// memref.cast %A: memref<A...> to compatibleMemRefType
/// scf.yield %view : compatibleMemRefType, index, index
/// } else {
/// // slowpath, masked vector.transfer or linalg.copy.
-/// memref_cast %alloc: memref<B...> to compatibleMemRefType
+/// memref.cast %alloc: memref<B...> to compatibleMemRefType
/// scf.yield %4 : compatibleMemRefType, index, index
// }
/// %0 = vector.transfer_read %1#0[%1#1, %1#2] {masked = [false ... false]}
@@ -2565,8 +2567,8 @@ LogicalResult mlir::vector::splitFullAndPartialTransfer(
b.setInsertionPointToStart(&funcOp.getRegion().front());
auto shape = xferOp.getVectorType().getShape();
Type elementType = xferOp.getVectorType().getElementType();
- alloc = std_alloca(MemRefType::get(shape, elementType), ValueRange{},
- b.getI64IntegerAttr(32));
+ alloc = memref_alloca(MemRefType::get(shape, elementType), ValueRange{},
+ b.getI64IntegerAttr(32));
}
MemRefType compatibleMemRefType =
@@ -2780,8 +2782,8 @@ struct TransferReadToVectorLoadLowering
if (!broadcastedDims.empty() &&
unbroadcastedVectorType.getNumElements() == 1) {
// If broadcasting is required and the number of loaded elements is 1 then
- // we can create `std.load` instead of `vector.load`.
- loadOp = rewriter.create<mlir::LoadOp>(read.getLoc(), read.source(),
+ // we can create `memref.load` instead of `vector.load`.
+ loadOp = rewriter.create<memref::LoadOp>(read.getLoc(), read.source(),
read.indices());
} else {
// Otherwise create `vector.load`.
diff --git a/mlir/lib/Transforms/BufferDeallocation.cpp b/mlir/lib/Transforms/BufferDeallocation.cpp
index a4155695fdf7..aa837cb0e77c 100644
--- a/mlir/lib/Transforms/BufferDeallocation.cpp
+++ b/mlir/lib/Transforms/BufferDeallocation.cpp
@@ -53,6 +53,7 @@
#include "PassDetail.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Utils/Utils.h"
#include "mlir/IR/Operation.h"
@@ -425,8 +426,8 @@ class BufferDeallocation : BufferPlacementTransformationBase {
// TODO: provide a generic interface to create dialect-specific
// Alloc and CopyOp nodes.
- auto alloc = builder.create<AllocOp>(terminator->getLoc(), memRefType,
- dynamicOperands);
+ auto alloc = builder.create<memref::AllocOp>(terminator->getLoc(),
+ memRefType, dynamicOperands);
// Create a new copy operation that copies to contents of the old
// allocation to the new one.
@@ -499,7 +500,7 @@ class BufferDeallocation : BufferPlacementTransformationBase {
continue;
// If there is no dealloc node, insert one in the right place.
OpBuilder builder(nextOp);
- builder.create<DeallocOp>(alloc.getLoc(), alloc);
+ builder.create<memref::DeallocOp>(alloc.getLoc(), alloc);
}
}
}
diff --git a/mlir/lib/Transforms/BufferOptimizations.cpp b/mlir/lib/Transforms/BufferOptimizations.cpp
index 9ed4bb46e239..b681c860de05 100644
--- a/mlir/lib/Transforms/BufferOptimizations.cpp
+++ b/mlir/lib/Transforms/BufferOptimizations.cpp
@@ -12,6 +12,7 @@
// convert heap-based allocations to stack-based allocations, if possible.
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/Operation.h"
#include "mlir/Interfaces/LoopLikeInterface.h"
#include "mlir/Pass/Pass.h"
@@ -33,7 +34,7 @@ static bool defaultIsSmallAlloc(Value alloc, unsigned maximumSizeInBytes,
unsigned bitwidthOfIndexType,
unsigned maxRankOfAllocatedMemRef) {
auto type = alloc.getType().dyn_cast<ShapedType>();
- if (!type || !alloc.getDefiningOp<AllocOp>())
+ if (!type || !alloc.getDefiningOp<memref::AllocOp>())
return false;
if (!type.hasStaticShape()) {
// Check if the dynamic shape dimension of the alloc is produced by RankOp.
@@ -317,7 +318,7 @@ class BufferPlacementPromotion : BufferPlacementTransformationBase {
// `AutomaticAllocationScope` determined during the initialization phase.
OpBuilder builder(startOperation);
Operation *allocOp = alloc.getDefiningOp();
- Operation *alloca = builder.create<AllocaOp>(
+ Operation *alloca = builder.create<memref::AllocaOp>(
alloc.getLoc(), alloc.getType().cast<MemRefType>(),
allocOp->getOperands());
diff --git a/mlir/lib/Transforms/BufferResultsToOutParams.cpp b/mlir/lib/Transforms/BufferResultsToOutParams.cpp
index 87f1a2ef936f..0920d1321e42 100644
--- a/mlir/lib/Transforms/BufferResultsToOutParams.cpp
+++ b/mlir/lib/Transforms/BufferResultsToOutParams.cpp
@@ -8,6 +8,7 @@
#include "PassDetail.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/Operation.h"
#include "mlir/Pass/Pass.h"
@@ -99,7 +100,7 @@ static LogicalResult updateCalls(ModuleOp module) {
didFail = true;
return;
}
- Value outParam = builder.create<AllocOp>(
+ Value outParam = builder.create<memref::AllocOp>(
op.getLoc(), memref.getType().cast<MemRefType>());
memref.replaceAllUsesWith(outParam);
outParams.push_back(outParam);
diff --git a/mlir/lib/Transforms/Bufferize.cpp b/mlir/lib/Transforms/Bufferize.cpp
index 66b1cc65646c..74de8613646a 100644
--- a/mlir/lib/Transforms/Bufferize.cpp
+++ b/mlir/lib/Transforms/Bufferize.cpp
@@ -8,6 +8,7 @@
#include "mlir/Transforms/Bufferize.h"
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/Operation.h"
#include "mlir/Transforms/Passes.h"
@@ -21,7 +22,7 @@ static Value materializeTensorLoad(OpBuilder &builder, TensorType type,
ValueRange inputs, Location loc) {
assert(inputs.size() == 1);
assert(inputs[0].getType().isa<BaseMemRefType>());
- return builder.create<TensorLoadOp>(loc, type, inputs[0]);
+ return builder.create<memref::TensorLoadOp>(loc, type, inputs[0]);
}
/// Registers conversions into BufferizeTypeConverter
@@ -42,24 +43,24 @@ BufferizeTypeConverter::BufferizeTypeConverter() {
ValueRange inputs, Location loc) -> Value {
assert(inputs.size() == 1);
assert(inputs[0].getType().isa<TensorType>());
- return builder.create<TensorToMemrefOp>(loc, type, inputs[0]);
+ return builder.create<memref::BufferCastOp>(loc, type, inputs[0]);
});
}
void mlir::populateBufferizeMaterializationLegality(ConversionTarget &target) {
- target.addLegalOp<TensorLoadOp, TensorToMemrefOp>();
+ target.addLegalOp<memref::TensorLoadOp, memref::BufferCastOp>();
}
namespace {
// In a finalizing bufferize conversion, we know that all tensors have been
// converted to memrefs, thus, this op becomes an identity.
-class BufferizeTensorLoadOp : public OpConversionPattern<TensorLoadOp> {
+class BufferizeTensorLoadOp : public OpConversionPattern<memref::TensorLoadOp> {
public:
using OpConversionPattern::OpConversionPattern;
LogicalResult
- matchAndRewrite(TensorLoadOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::TensorLoadOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- TensorLoadOp::Adaptor adaptor(operands);
+ memref::TensorLoadOp::Adaptor adaptor(operands);
rewriter.replaceOp(op, adaptor.memref());
return success();
}
@@ -69,13 +70,13 @@ class BufferizeTensorLoadOp : public OpConversionPattern<TensorLoadOp> {
namespace {
// In a finalizing bufferize conversion, we know that all tensors have been
// converted to memrefs, thus, this op becomes an identity.
-class BufferizeTensorToMemrefOp : public OpConversionPattern<TensorToMemrefOp> {
+class BufferizeCastOp : public OpConversionPattern<memref::BufferCastOp> {
public:
using OpConversionPattern::OpConversionPattern;
LogicalResult
- matchAndRewrite(TensorToMemrefOp op, ArrayRef<Value> operands,
+ matchAndRewrite(memref::BufferCastOp op, ArrayRef<Value> operands,
ConversionPatternRewriter &rewriter) const override {
- TensorToMemrefOp::Adaptor adaptor(operands);
+ memref::BufferCastOp::Adaptor adaptor(operands);
rewriter.replaceOp(op, adaptor.tensor());
return success();
}
@@ -85,8 +86,8 @@ class BufferizeTensorToMemrefOp : public OpConversionPattern<TensorToMemrefOp> {
void mlir::populateEliminateBufferizeMaterializationsPatterns(
MLIRContext *context, BufferizeTypeConverter &typeConverter,
OwningRewritePatternList &patterns) {
- patterns.insert<BufferizeTensorLoadOp, BufferizeTensorToMemrefOp>(
- typeConverter, context);
+ patterns.insert<BufferizeTensorLoadOp, BufferizeCastOp>(typeConverter,
+ context);
}
namespace {
diff --git a/mlir/lib/Transforms/CMakeLists.txt b/mlir/lib/Transforms/CMakeLists.txt
index 95f71e1e0371..36f9e5b832be 100644
--- a/mlir/lib/Transforms/CMakeLists.txt
+++ b/mlir/lib/Transforms/CMakeLists.txt
@@ -38,6 +38,7 @@ add_mlir_library(MLIRTransforms
MLIRCopyOpInterface
MLIRLinalg
MLIRLoopLikeInterface
+ MLIRMemRef
MLIRSCF
MLIRPass
MLIRTransformUtils
diff --git a/mlir/lib/Transforms/Canonicalizer.cpp b/mlir/lib/Transforms/Canonicalizer.cpp
index ba6ffe72a248..cd99681a0283 100644
--- a/mlir/lib/Transforms/Canonicalizer.cpp
+++ b/mlir/lib/Transforms/Canonicalizer.cpp
@@ -12,6 +12,7 @@
//===----------------------------------------------------------------------===//
#include "PassDetail.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
#include "mlir/Transforms/Passes.h"
diff --git a/mlir/lib/Transforms/LoopFusion.cpp b/mlir/lib/Transforms/LoopFusion.cpp
index d6d18b3c6f7a..4eb23f841fd0 100644
--- a/mlir/lib/Transforms/LoopFusion.cpp
+++ b/mlir/lib/Transforms/LoopFusion.cpp
@@ -16,6 +16,7 @@
#include "mlir/Analysis/LoopAnalysis.h"
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/AffineExpr.h"
#include "mlir/IR/AffineMap.h"
#include "mlir/IR/Builders.h"
@@ -958,7 +959,7 @@ static Value createPrivateMemRef(AffineForOp forOp, Operation *srcStoreOpInst,
// consumer loop nests to reduce their live range. Currently they are added
// at the beginning of the function, because loop nests can be reordered
// during the fusion pass.
- Value newMemRef = top.create<AllocOp>(forOp.getLoc(), newMemRefType);
+ Value newMemRef = top.create<memref::AllocOp>(forOp.getLoc(), newMemRefType);
// Build an AffineMap to remap access functions based on lower bound offsets.
SmallVector<AffineExpr, 4> remapExprs;
@@ -1937,7 +1938,7 @@ struct GreedyFusion {
continue;
// Use list expected to match the dep graph info.
auto *op = memref.getDefiningOp();
- if (isa_and_nonnull<AllocOp>(op))
+ if (isa_and_nonnull<memref::AllocOp>(op))
op->erase();
}
}
diff --git a/mlir/lib/Transforms/MemRefDataFlowOpt.cpp b/mlir/lib/Transforms/MemRefDataFlowOpt.cpp
index 7924b46a83d5..401aaaf1c7cd 100644
--- a/mlir/lib/Transforms/MemRefDataFlowOpt.cpp
+++ b/mlir/lib/Transforms/MemRefDataFlowOpt.cpp
@@ -17,6 +17,7 @@
#include "mlir/Analysis/AffineAnalysis.h"
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/Dominance.h"
#include "mlir/Transforms/Passes.h"
@@ -168,7 +169,7 @@ void MemRefDataFlowOpt::forwardStoreToLoad(AffineReadOpInterface loadOp) {
// Perform the actual store to load forwarding.
Value storeVal =
- cast<AffineWriteOpInterface>(lastWriteStoreOp).getValueToStore();
+ cast<AffineWriteOpInterface>(lastWriteStoreOp).getValueToStore();
loadOp.getValue().replaceAllUsesWith(storeVal);
// Record the memref for a later sweep to optimize away.
memrefsToErase.insert(loadOp.getMemRef());
@@ -203,12 +204,12 @@ void MemRefDataFlowOpt::runOnFunction() {
for (auto memref : memrefsToErase) {
// If the memref hasn't been alloc'ed in this function, skip.
Operation *defOp = memref.getDefiningOp();
- if (!defOp || !isa<AllocOp>(defOp))
+ if (!defOp || !isa<memref::AllocOp>(defOp))
// TODO: if the memref was returned by a 'call' operation, we
// could still erase it if the call had no side-effects.
continue;
if (llvm::any_of(memref.getUsers(), [&](Operation *ownerOp) {
- return !isa<AffineWriteOpInterface, DeallocOp>(ownerOp);
+ return !isa<AffineWriteOpInterface, memref::DeallocOp>(ownerOp);
}))
continue;
diff --git a/mlir/lib/Transforms/NormalizeMemRefs.cpp b/mlir/lib/Transforms/NormalizeMemRefs.cpp
index d4adcfcb8410..831c613a4118 100644
--- a/mlir/lib/Transforms/NormalizeMemRefs.cpp
+++ b/mlir/lib/Transforms/NormalizeMemRefs.cpp
@@ -13,6 +13,7 @@
#include "PassDetail.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Transforms/Passes.h"
#include "mlir/Transforms/Utils.h"
#include "llvm/ADT/SmallSet.h"
@@ -152,7 +153,7 @@ bool NormalizeMemRefs::areMemRefsNormalizable(FuncOp funcOp) {
return true;
if (funcOp
- .walk([&](AllocOp allocOp) -> WalkResult {
+ .walk([&](memref::AllocOp allocOp) -> WalkResult {
Value oldMemRef = allocOp.getResult();
if (!isMemRefNormalizable(oldMemRef.getUsers()))
return WalkResult::interrupt();
@@ -326,10 +327,10 @@ void NormalizeMemRefs::normalizeFuncOpMemRefs(FuncOp funcOp,
// Turn memrefs' non-identity layouts maps into ones with identity. Collect
// alloc ops first and then process since normalizeMemRef replaces/erases ops
// during memref rewriting.
- SmallVector<AllocOp, 4> allocOps;
- funcOp.walk([&](AllocOp op) { allocOps.push_back(op); });
- for (AllocOp allocOp : allocOps)
- (void)normalizeMemRef(allocOp);
+ SmallVector<memref::AllocOp, 4> allocOps;
+ funcOp.walk([&](memref::AllocOp op) { allocOps.push_back(op); });
+ for (memref::AllocOp allocOp : allocOps)
+ (void)normalizeMemRef(&allocOp);
// We use this OpBuilder to create new memref layout later.
OpBuilder b(funcOp);
diff --git a/mlir/lib/Transforms/PassDetail.h b/mlir/lib/Transforms/PassDetail.h
index 220ed1aac407..81b9530c4835 100644
--- a/mlir/lib/Transforms/PassDetail.h
+++ b/mlir/lib/Transforms/PassDetail.h
@@ -20,6 +20,10 @@ namespace linalg {
class LinalgDialect;
} // end namespace linalg
+namespace memref {
+class MemRefDialect;
+} // end namespace memref
+
#define GEN_PASS_CLASSES
#include "mlir/Transforms/Passes.h.inc"
diff --git a/mlir/lib/Transforms/PipelineDataTransfer.cpp b/mlir/lib/Transforms/PipelineDataTransfer.cpp
index 025b1b217476..d15be19b2411 100644
--- a/mlir/lib/Transforms/PipelineDataTransfer.cpp
+++ b/mlir/lib/Transforms/PipelineDataTransfer.cpp
@@ -17,6 +17,7 @@
#include "mlir/Analysis/LoopAnalysis.h"
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/Utils/Utils.h"
#include "mlir/IR/Builders.h"
#include "mlir/Transforms/LoopUtils.h"
@@ -88,8 +89,8 @@ static bool doubleBuffer(Value oldMemRef, AffineForOp forOp) {
auto allocOperands = getDynOperands(forOp.getLoc(), oldMemRef, bOuter);
// Create and place the alloc right before the 'affine.for' operation.
- Value newMemRef =
- bOuter.create<AllocOp>(forOp.getLoc(), newMemRefType, allocOperands);
+ Value newMemRef = bOuter.create<memref::AllocOp>(
+ forOp.getLoc(), newMemRefType, allocOperands);
// Create 'iv mod 2' value to index the leading dimension.
auto d0 = bInner.getAffineDimExpr(0);
@@ -115,7 +116,7 @@ static bool doubleBuffer(Value oldMemRef, AffineForOp forOp) {
}
// Insert the dealloc op right after the for loop.
bOuter.setInsertionPointAfter(forOp);
- bOuter.create<DeallocOp>(forOp.getLoc(), newMemRef);
+ bOuter.create<memref::DeallocOp>(forOp.getLoc(), newMemRef);
return true;
}
@@ -201,7 +202,7 @@ static void findMatchingStartFinishInsts(
bool escapingUses = false;
for (auto *user : memref.getUsers()) {
// We can double buffer regardless of dealloc's outside the loop.
- if (isa<DeallocOp>(user))
+ if (isa<memref::DeallocOp>(user))
continue;
if (!forOp.getBody()->findAncestorOpInBlock(*user)) {
LLVM_DEBUG(llvm::dbgs()
@@ -274,7 +275,8 @@ void PipelineDataTransfer::runOnAffineForOp(AffineForOp forOp) {
if (oldMemRef.use_empty()) {
allocOp->erase();
} else if (oldMemRef.hasOneUse()) {
- if (auto dealloc = dyn_cast<DeallocOp>(*oldMemRef.user_begin())) {
+ if (auto dealloc =
+ dyn_cast<memref::DeallocOp>(*oldMemRef.user_begin())) {
dealloc.erase();
allocOp->erase();
}
@@ -296,7 +298,8 @@ void PipelineDataTransfer::runOnAffineForOp(AffineForOp forOp) {
if (oldTagMemRef.use_empty()) {
tagAllocOp->erase();
} else if (oldTagMemRef.hasOneUse()) {
- if (auto dealloc = dyn_cast<DeallocOp>(*oldTagMemRef.user_begin())) {
+ if (auto dealloc =
+ dyn_cast<memref::DeallocOp>(*oldTagMemRef.user_begin())) {
dealloc.erase();
tagAllocOp->erase();
}
diff --git a/mlir/lib/Transforms/Utils/CMakeLists.txt b/mlir/lib/Transforms/Utils/CMakeLists.txt
index 6dbb74b4905a..704f431d7526 100644
--- a/mlir/lib/Transforms/Utils/CMakeLists.txt
+++ b/mlir/lib/Transforms/Utils/CMakeLists.txt
@@ -18,6 +18,7 @@ add_mlir_library(MLIRTransformUtils
MLIRAffine
MLIRAnalysis
MLIRLoopAnalysis
+ MLIRMemRef
MLIRSCF
MLIRPass
MLIRRewrite
diff --git a/mlir/lib/Transforms/Utils/LoopUtils.cpp b/mlir/lib/Transforms/Utils/LoopUtils.cpp
index 3fa04514e81f..a9b5979278ff 100644
--- a/mlir/lib/Transforms/Utils/LoopUtils.cpp
+++ b/mlir/lib/Transforms/Utils/LoopUtils.cpp
@@ -18,6 +18,7 @@
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Affine/IR/AffineValueMap.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/IR/AffineMap.h"
#include "mlir/IR/BlockAndValueMapping.h"
@@ -2487,7 +2488,8 @@ static LogicalResult generateCopy(
// Create the fast memory space buffer just before the 'affine.for'
// operation.
- fastMemRef = prologue.create<AllocOp>(loc, fastMemRefType).getResult();
+ fastMemRef =
+ prologue.create<memref::AllocOp>(loc, fastMemRefType).getResult();
// Record it.
fastBufferMap[memref] = fastMemRef;
// fastMemRefType is a constant shaped memref.
@@ -2557,7 +2559,7 @@ static LogicalResult generateCopy(
// Create a tag (single element 1-d memref) for the DMA.
auto tagMemRefType = MemRefType::get({1}, top.getIntegerType(32), {},
copyOptions.tagMemorySpace);
- auto tagMemRef = prologue.create<AllocOp>(loc, tagMemRefType);
+ auto tagMemRef = prologue.create<memref::AllocOp>(loc, tagMemRefType);
SmallVector<Value, 4> tagIndices({zeroIndex});
auto tagAffineMap = b.getMultiDimIdentityMap(tagIndices.size());
@@ -2585,7 +2587,7 @@ static LogicalResult generateCopy(
numElementsSSA);
// Generate dealloc for the tag.
- auto tagDeallocOp = epilogue.create<DeallocOp>(loc, tagMemRef);
+ auto tagDeallocOp = epilogue.create<memref::DeallocOp>(loc, tagMemRef);
if (*nEnd == end && isCopyOutAtEndOfBlock)
// Since new ops are being appended (for outgoing DMAs), adjust the end to
// mark end of range of the original.
@@ -2594,7 +2596,7 @@ static LogicalResult generateCopy(
// Generate dealloc for the buffer.
if (!existingBuf) {
- auto bufDeallocOp = epilogue.create<DeallocOp>(loc, fastMemRef);
+ auto bufDeallocOp = epilogue.create<memref::DeallocOp>(loc, fastMemRef);
// When generating pointwise copies, `nEnd' has to be set to deallocOp on
// the fast buffer (since it marks the new end insertion point).
if (!copyOptions.generateDma && *nEnd == end && isCopyOutAtEndOfBlock)
diff --git a/mlir/lib/Transforms/Utils/Utils.cpp b/mlir/lib/Transforms/Utils/Utils.cpp
index f99159b25054..58e8d8cd5006 100644
--- a/mlir/lib/Transforms/Utils/Utils.cpp
+++ b/mlir/lib/Transforms/Utils/Utils.cpp
@@ -17,6 +17,7 @@
#include "mlir/Analysis/AffineStructures.h"
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/IR/Builders.h"
#include "mlir/IR/BuiltinOps.h"
#include "mlir/IR/Dominance.h"
@@ -253,7 +254,7 @@ LogicalResult mlir::replaceAllMemRefUsesWith(
// Skip dealloc's - no replacement is necessary, and a memref replacement
// at other uses doesn't hurt these dealloc's.
- if (isa<DeallocOp>(op) && !replaceInDeallocOp)
+ if (isa<memref::DeallocOp>(op) && !replaceInDeallocOp)
continue;
// Check if the memref was used in a non-dereferencing context. It is fine
@@ -380,24 +381,24 @@ void mlir::createAffineComputationSlice(
}
// TODO: Currently works for static memrefs with a single layout map.
-LogicalResult mlir::normalizeMemRef(AllocOp allocOp) {
- MemRefType memrefType = allocOp.getType();
- OpBuilder b(allocOp);
+LogicalResult mlir::normalizeMemRef(memref::AllocOp *allocOp) {
+ MemRefType memrefType = allocOp->getType();
+ OpBuilder b(*allocOp);
// Fetch a new memref type after normalizing the old memref to have an
// identity map layout.
MemRefType newMemRefType =
- normalizeMemRefType(memrefType, b, allocOp.symbolOperands().size());
+ normalizeMemRefType(memrefType, b, allocOp->symbolOperands().size());
if (newMemRefType == memrefType)
// Either memrefType already had an identity map or the map couldn't be
// transformed to an identity map.
return failure();
- Value oldMemRef = allocOp.getResult();
+ Value oldMemRef = allocOp->getResult();
- SmallVector<Value, 4> symbolOperands(allocOp.symbolOperands());
- AllocOp newAlloc = b.create<AllocOp>(allocOp.getLoc(), newMemRefType,
- allocOp.alignmentAttr());
+ SmallVector<Value, 4> symbolOperands(allocOp->symbolOperands());
+ memref::AllocOp newAlloc = b.create<memref::AllocOp>(
+ allocOp->getLoc(), newMemRefType, allocOp->alignmentAttr());
AffineMap layoutMap = memrefType.getAffineMaps().front();
// Replace all uses of the old memref.
if (failed(replaceAllMemRefUsesWith(oldMemRef, /*newMemRef=*/newAlloc,
@@ -414,10 +415,11 @@ LogicalResult mlir::normalizeMemRef(AllocOp allocOp) {
}
// Replace any uses of the original alloc op and erase it. All remaining uses
// have to be dealloc's; RAMUW above would've failed otherwise.
- assert(llvm::all_of(oldMemRef.getUsers(),
- [](Operation *op) { return isa<DeallocOp>(op); }));
+ assert(llvm::all_of(oldMemRef.getUsers(), [](Operation *op) {
+ return isa<memref::DeallocOp>(op);
+ }));
oldMemRef.replaceAllUsesWith(newAlloc);
- allocOp.erase();
+ allocOp->erase();
return success();
}
diff --git a/mlir/test/Analysis/test-alias-analysis.mlir b/mlir/test/Analysis/test-alias-analysis.mlir
index cb0f6484f1e3..a53e9e1f05f2 100644
--- a/mlir/test/Analysis/test-alias-analysis.mlir
+++ b/mlir/test/Analysis/test-alias-analysis.mlir
@@ -23,10 +23,10 @@
// CHECK-DAG: alloc_2#0 <-> func.region0#0: MayAlias
// CHECK-DAG: alloc_2#0 <-> func.region0#1: MayAlias
func @simple(%arg: memref<2xf32>, %arg1: memref<2xf32>) attributes {test.ptr = "func"} {
- %0 = alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
- %1 = alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
- %2 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
- %3 = alloc() {test.ptr = "alloc_2"} : memref<8x64xf32>
+ %0 = memref.alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
+ %1 = memref.alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
+ %2 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %3 = memref.alloc() {test.ptr = "alloc_2"} : memref<8x64xf32>
return
}
@@ -50,9 +50,9 @@ func @simple(%arg: memref<2xf32>, %arg1: memref<2xf32>) attributes {test.ptr = "
// CHECK-DAG: func.region0.block1#0 <-> func.region0.block2#0: MustAlias
func @control_flow(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr = "func"} {
- %0 = alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
- %1 = alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
- %2 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %0 = memref.alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
+ %1 = memref.alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
+ %2 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
cond_br %cond, ^bb1(%0 : memref<8x64xf32>), ^bb2(%0 : memref<8x64xf32>)
@@ -83,9 +83,9 @@ func @control_flow(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr = "func"
// CHECK-DAG: func.region0.block1#0 <-> func.region0.block2#0: MayAlias
func @control_flow_merge(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr = "func"} {
- %0 = alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
- %1 = alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
- %2 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %0 = memref.alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
+ %1 = memref.alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
+ %2 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
cond_br %cond, ^bb1(%0 : memref<8x64xf32>), ^bb2(%2 : memref<8x64xf32>)
@@ -123,9 +123,9 @@ func @control_flow_merge(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr =
// CHECK-DAG: if_alloc#0 <-> func.region0#0: MayAlias
// CHECK-DAG: if_alloc#0 <-> func.region0#1: MayAlias
func @region_control_flow(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr = "func"} {
- %0 = alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
- %1 = alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
- %2 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %0 = memref.alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
+ %1 = memref.alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
+ %2 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
%3 = scf.if %cond -> (memref<8x64xf32>) {
scf.yield %0 : memref<8x64xf32>
@@ -181,9 +181,9 @@ func @region_control_flow(%arg: memref<2xf32>, %cond: i1) attributes {test.ptr =
// CHECK-DAG: for_alloca.region0#1 <-> func.region0#3: NoAlias
func @region_loop_control_flow(%arg: memref<2xf32>, %loopI0 : index,
%loopI1 : index, %loopI2 : index) attributes {test.ptr = "func"} {
- %0 = alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
- %1 = alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
- %2 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %0 = memref.alloca() {test.ptr = "alloca_1"} : memref<8x64xf32>
+ %1 = memref.alloca() {test.ptr = "alloca_2"} : memref<8x64xf32>
+ %2 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
%result = scf.for %i0 = %loopI0 to %loopI1 step %loopI2 iter_args(%si = %0) -> (memref<8x64xf32>) {
scf.yield %si : memref<8x64xf32>
@@ -201,11 +201,11 @@ func @region_loop_control_flow(%arg: memref<2xf32>, %loopI0 : index,
// CHECK-DAG: view#0 <-> func.region0#0: NoAlias
// CHECK-DAG: view#0 <-> func.region0#1: NoAlias
func @view_like(%arg: memref<2xf32>, %size: index) attributes {test.ptr = "func"} {
- %1 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %1 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
%c0 = constant 0 : index
- %2 = alloca (%size) {test.ptr = "alloca_1"} : memref<?xi8>
- %3 = view %2[%c0][] {test.ptr = "view"} : memref<?xi8> to memref<8x64xf32>
+ %2 = memref.alloca (%size) {test.ptr = "alloca_1"} : memref<?xi8>
+ %3 = memref.view %2[%c0][] {test.ptr = "view"} : memref<?xi8> to memref<8x64xf32>
return
}
@@ -225,7 +225,7 @@ func @view_like(%arg: memref<2xf32>, %size: index) attributes {test.ptr = "func"
// CHECK-DAG: constant_3#0 <-> func.region0#0: MayAlias
func @constants(%arg: memref<2xf32>) attributes {test.ptr = "func"} {
- %1 = alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
+ %1 = memref.alloc() {test.ptr = "alloc_1"} : memref<8x64xf32>
%c0 = constant {test.ptr = "constant_1"} 0 : index
%c0_2 = constant {test.ptr = "constant_2"} 0 : index
diff --git a/mlir/test/Analysis/test-liveness.mlir b/mlir/test/Analysis/test-liveness.mlir
index 11648e4819a3..ffb64dbd93df 100644
--- a/mlir/test/Analysis/test-liveness.mlir
+++ b/mlir/test/Analysis/test-liveness.mlir
@@ -221,7 +221,7 @@ func @nested_region(
// CHECK-NEXT: LiveOut:{{ *$}}
%2 = addi %0, %arg5 : i32
%3 = addi %2, %0 : i32
- store %3, %buffer[] : memref<i32>
+ memref.store %3, %buffer[] : memref<i32>
}
return %1 : i32
}
@@ -265,7 +265,7 @@ func @nested_region2(
%2 = addi %0, %arg5 : i32
scf.for %arg7 = %arg0 to %arg1 step %arg2 {
%3 = addi %2, %0 : i32
- store %3, %buffer[] : memref<i32>
+ memref.store %3, %buffer[] : memref<i32>
}
}
return %1 : i32
@@ -299,7 +299,7 @@ func @nested_region3(
// CHECK-NEXT: LiveIn: arg5 at 0 arg6 at 0 val_7
// CHECK-NEXT: LiveOut:{{ *$}}
%2 = addi %0, %arg5 : i32
- store %2, %buffer[] : memref<i32>
+ memref.store %2, %buffer[] : memref<i32>
}
br ^exit
@@ -312,7 +312,7 @@ func @nested_region3(
// CHECK-NEXT: LiveIn: arg6 at 0 val_7 val_8
// CHECK-NEXT: LiveOut:{{ *$}}
%2 = addi %0, %1 : i32
- store %2, %buffer[] : memref<i32>
+ memref.store %2, %buffer[] : memref<i32>
}
return %1 : i32
}
diff --git a/mlir/test/CAPI/ir.c b/mlir/test/CAPI/ir.c
index 38d200ec7597..b72a6e6cf2fd 100644
--- a/mlir/test/CAPI/ir.c
+++ b/mlir/test/CAPI/ir.c
@@ -35,7 +35,7 @@ void populateLoopBody(MlirContext ctx, MlirBlock loopBody,
mlirTypeParseGet(ctx, mlirStringRefCreateFromCString("f32"));
MlirOperationState loadLHSState = mlirOperationStateGet(
- mlirStringRefCreateFromCString("std.load"), location);
+ mlirStringRefCreateFromCString("memref.load"), location);
MlirValue loadLHSOperands[] = {funcArg0, iv};
mlirOperationStateAddOperands(&loadLHSState, 2, loadLHSOperands);
mlirOperationStateAddResults(&loadLHSState, 1, &f32Type);
@@ -43,7 +43,7 @@ void populateLoopBody(MlirContext ctx, MlirBlock loopBody,
mlirBlockAppendOwnedOperation(loopBody, loadLHS);
MlirOperationState loadRHSState = mlirOperationStateGet(
- mlirStringRefCreateFromCString("std.load"), location);
+ mlirStringRefCreateFromCString("memref.load"), location);
MlirValue loadRHSOperands[] = {funcArg1, iv};
mlirOperationStateAddOperands(&loadRHSState, 2, loadRHSOperands);
mlirOperationStateAddResults(&loadRHSState, 1, &f32Type);
@@ -60,7 +60,7 @@ void populateLoopBody(MlirContext ctx, MlirBlock loopBody,
mlirBlockAppendOwnedOperation(loopBody, add);
MlirOperationState storeState = mlirOperationStateGet(
- mlirStringRefCreateFromCString("std.store"), location);
+ mlirStringRefCreateFromCString("memref.store"), location);
MlirValue storeOperands[] = {mlirOperationGetResult(add, 0), funcArg0, iv};
mlirOperationStateAddOperands(&storeState, 3, storeOperands);
MlirOperation store = mlirOperationCreate(&storeState);
@@ -121,7 +121,7 @@ MlirModule makeAndDumpAdd(MlirContext ctx, MlirLocation location) {
MlirValue constZeroValue = mlirOperationGetResult(constZero, 0);
MlirValue dimOperands[] = {funcArg0, constZeroValue};
MlirOperationState dimState = mlirOperationStateGet(
- mlirStringRefCreateFromCString("std.dim"), location);
+ mlirStringRefCreateFromCString("memref.dim"), location);
mlirOperationStateAddOperands(&dimState, 2, dimOperands);
mlirOperationStateAddResults(&dimState, 1, &indexType);
MlirOperation dim = mlirOperationCreate(&dimState);
@@ -167,13 +167,13 @@ MlirModule makeAndDumpAdd(MlirContext ctx, MlirLocation location) {
// CHECK: module {
// CHECK: func @add(%[[ARG0:.*]]: memref<?xf32>, %[[ARG1:.*]]: memref<?xf32>) {
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[DIM:.*]] = dim %[[ARG0]], %[[C0]] : memref<?xf32>
+ // CHECK: %[[DIM:.*]] = memref.dim %[[ARG0]], %[[C0]] : memref<?xf32>
// CHECK: %[[C1:.*]] = constant 1 : index
// CHECK: scf.for %[[I:.*]] = %[[C0]] to %[[DIM]] step %[[C1]] {
- // CHECK: %[[LHS:.*]] = load %[[ARG0]][%[[I]]] : memref<?xf32>
- // CHECK: %[[RHS:.*]] = load %[[ARG1]][%[[I]]] : memref<?xf32>
+ // CHECK: %[[LHS:.*]] = memref.load %[[ARG0]][%[[I]]] : memref<?xf32>
+ // CHECK: %[[RHS:.*]] = memref.load %[[ARG1]][%[[I]]] : memref<?xf32>
// CHECK: %[[SUM:.*]] = addf %[[LHS]], %[[RHS]] : f32
- // CHECK: store %[[SUM]], %[[ARG0]][%[[I]]] : memref<?xf32>
+ // CHECK: memref.store %[[SUM]], %[[ARG0]][%[[I]]] : memref<?xf32>
// CHECK: }
// CHECK: return
// CHECK: }
@@ -330,7 +330,7 @@ static void printFirstOfEach(MlirContext ctx, MlirOperation operation) {
// CHECK: Block eq: 1
// In the module we created, the first operation of the first function is
- // an "std.dim", which has an attribute and a single result that we can
+ // an "memref.dim", which has an attribute and a single result that we can
// use to test the printing mechanism.
mlirBlockPrint(block, printToStderr, NULL);
fprintf(stderr, "\n");
@@ -339,13 +339,13 @@ static void printFirstOfEach(MlirContext ctx, MlirOperation operation) {
fprintf(stderr, "\n");
// clang-format off
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[DIM:.*]] = dim %{{.*}}, %[[C0]] : memref<?xf32>
+ // CHECK: %[[DIM:.*]] = memref.dim %{{.*}}, %[[C0]] : memref<?xf32>
// CHECK: %[[C1:.*]] = constant 1 : index
// CHECK: scf.for %[[I:.*]] = %[[C0]] to %[[DIM]] step %[[C1]] {
- // CHECK: %[[LHS:.*]] = load %{{.*}}[%[[I]]] : memref<?xf32>
- // CHECK: %[[RHS:.*]] = load %{{.*}}[%[[I]]] : memref<?xf32>
+ // CHECK: %[[LHS:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf32>
+ // CHECK: %[[RHS:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf32>
// CHECK: %[[SUM:.*]] = addf %[[LHS]], %[[RHS]] : f32
- // CHECK: store %[[SUM]], %{{.*}}[%[[I]]] : memref<?xf32>
+ // CHECK: memref.store %[[SUM]], %{{.*}}[%[[I]]] : memref<?xf32>
// CHECK: }
// CHECK: return
// CHECK: First operation: {{.*}} = constant 0 : index
@@ -1453,13 +1453,15 @@ static int testBackreferences() {
mlirContextSetAllowUnregisteredDialects(ctx, true);
MlirLocation loc = mlirLocationUnknownGet(ctx);
- MlirOperationState opState = mlirOperationStateGet(mlirStringRefCreateFromCString("invalid.op"), loc);
+ MlirOperationState opState =
+ mlirOperationStateGet(mlirStringRefCreateFromCString("invalid.op"), loc);
MlirRegion region = mlirRegionCreate();
MlirBlock block = mlirBlockCreate(0, NULL);
mlirRegionAppendOwnedBlock(region, block);
mlirOperationStateAddOwnedRegions(&opState, 1, ®ion);
MlirOperation op = mlirOperationCreate(&opState);
- MlirIdentifier ident = mlirIdentifierGet(ctx, mlirStringRefCreateFromCString("identifier"));
+ MlirIdentifier ident =
+ mlirIdentifierGet(ctx, mlirStringRefCreateFromCString("identifier"));
if (!mlirContextEqual(ctx, mlirOperationGetContext(op))) {
fprintf(stderr, "ERROR: Getting context from operation failed\n");
@@ -1473,7 +1475,7 @@ static int testBackreferences() {
fprintf(stderr, "ERROR: Getting context from identifier failed\n");
return 3;
}
-
+
mlirOperationDestroy(op);
mlirContextDestroy(ctx);
diff --git a/mlir/test/Conversion/AffineToStandard/lower-affine-gpu.mlir b/mlir/test/Conversion/AffineToStandard/lower-affine-gpu.mlir
index 8e0d4f4fe8a4..382c3359278e 100644
--- a/mlir/test/Conversion/AffineToStandard/lower-affine-gpu.mlir
+++ b/mlir/test/Conversion/AffineToStandard/lower-affine-gpu.mlir
@@ -4,12 +4,12 @@
gpu.module @kernels {
gpu.func @foo(%arg0 : index, %arg1 : memref<?xf32>) -> f32 {
%0 = affine.apply #map0gpufunc(%arg0)
- %1 = load %arg1[%0] : memref<?xf32>
+ %1 = memref.load %arg1[%0] : memref<?xf32>
gpu.return %1 : f32
}
// CHECK: gpu.func
// CHECK-SAME: %[[ARG0:.*]]: index
// CHECK-NOT: affine.apply
-// CHECK: load %{{.*}}[%[[ARG0]]]
+// CHECK: memref.load %{{.*}}[%[[ARG0]]]
}
diff --git a/mlir/test/Conversion/AffineToStandard/lower-affine-to-vector.mlir b/mlir/test/Conversion/AffineToStandard/lower-affine-to-vector.mlir
index 3df9bb33ece2..ce87504e90d0 100644
--- a/mlir/test/Conversion/AffineToStandard/lower-affine-to-vector.mlir
+++ b/mlir/test/Conversion/AffineToStandard/lower-affine-to-vector.mlir
@@ -1,13 +1,12 @@
// RUN: mlir-opt -lower-affine --split-input-file %s | FileCheck %s
-
// CHECK-LABEL: func @affine_vector_load
func @affine_vector_load(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
affine.for %i0 = 0 to 16 {
%1 = affine.vector_load %0[%i0 + symbol(%arg0) + 7] : memref<100xf32>, vector<8xf32>
}
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK: %[[a:.*]] = addi %{{.*}}, %{{.*}} : index
// CHECK-NEXT: %[[c7:.*]] = constant 7 : index
// CHECK-NEXT: %[[b:.*]] = addi %[[a]], %[[c7]] : index
@@ -19,12 +18,12 @@ func @affine_vector_load(%arg0 : index) {
// CHECK-LABEL: func @affine_vector_store
func @affine_vector_store(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%1 = constant dense<11.0> : vector<4xf32>
affine.for %i0 = 0 to 16 {
affine.vector_store %1, %0[%i0 - symbol(%arg0) + 7] : memref<100xf32>, vector<4xf32>
}
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK: %[[val:.*]] = constant dense
// CHECK: %[[c_1:.*]] = constant -1 : index
// CHECK-NEXT: %[[a:.*]] = muli %arg0, %[[c_1]] : index
@@ -39,11 +38,11 @@ func @affine_vector_store(%arg0 : index) {
// CHECK-LABEL: func @vector_load_2d
func @vector_load_2d() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 16 step 2{
affine.for %i1 = 0 to 16 step 8 {
%1 = affine.vector_load %0[%i0, %i1] : memref<100x100xf32>, vector<2x8xf32>
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK: scf.for %[[i0:.*]] =
// CHECK: scf.for %[[i1:.*]] =
// CHECK-NEXT: vector.load %[[buf]][%[[i0]], %[[i1]]] : memref<100x100xf32>, vector<2x8xf32>
@@ -56,12 +55,12 @@ func @vector_load_2d() {
// CHECK-LABEL: func @vector_store_2d
func @vector_store_2d() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%1 = constant dense<11.0> : vector<2x8xf32>
affine.for %i0 = 0 to 16 step 2{
affine.for %i1 = 0 to 16 step 8 {
affine.vector_store %1, %0[%i0, %i1] : memref<100x100xf32>, vector<2x8xf32>
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK: %[[val:.*]] = constant dense
// CHECK: scf.for %[[i0:.*]] =
// CHECK: scf.for %[[i1:.*]] =
diff --git a/mlir/test/Conversion/AffineToStandard/lower-affine.mlir b/mlir/test/Conversion/AffineToStandard/lower-affine.mlir
index cac7d17a8348..74c54ebe45b0 100644
--- a/mlir/test/Conversion/AffineToStandard/lower-affine.mlir
+++ b/mlir/test/Conversion/AffineToStandard/lower-affine.mlir
@@ -42,7 +42,7 @@ func @for_with_yield(%buffer: memref<1024xf32>) -> (f32) {
// CHECK-NEXT: %[[UPPER:.*]] = constant 10 : index
// CHECK-NEXT: %[[STEP:.*]] = constant 2 : index
// CHECK-NEXT: %[[SUM:.*]] = scf.for %[[IV:.*]] = %[[LOWER]] to %[[UPPER]] step %[[STEP]] iter_args(%[[SUM_ITER:.*]] = %[[INIT_SUM]]) -> (f32) {
-// CHECK-NEXT: load
+// CHECK-NEXT: memref.load
// CHECK-NEXT: %[[SUM_NEXT:.*]] = addf
// CHECK-NEXT: scf.yield %[[SUM_NEXT]] : f32
// CHECK-NEXT: }
@@ -533,20 +533,20 @@ func @affine_apply_ceildiv(%arg0 : index) -> (index) {
// CHECK-LABEL: func @affine_load
func @affine_load(%arg0 : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
%1 = affine.load %0[%i0 + symbol(%arg0) + 7] : memref<10xf32>
}
// CHECK: %[[a:.*]] = addi %{{.*}}, %{{.*}} : index
// CHECK-NEXT: %[[c7:.*]] = constant 7 : index
// CHECK-NEXT: %[[b:.*]] = addi %[[a]], %[[c7]] : index
-// CHECK-NEXT: %{{.*}} = load %[[v0:.*]][%[[b]]] : memref<10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %[[v0:.*]][%[[b]]] : memref<10xf32>
return
}
// CHECK-LABEL: func @affine_store
func @affine_store(%arg0 : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%1 = constant 11.0 : f32
affine.for %i0 = 0 to 10 {
affine.store %1, %0[%i0 - symbol(%arg0) + 7] : memref<10xf32>
@@ -564,29 +564,29 @@ func @affine_store(%arg0 : index) {
func @affine_load_store_zero_dim(%arg0 : memref<i32>, %arg1 : memref<i32>) {
%0 = affine.load %arg0[] : memref<i32>
affine.store %0, %arg1[] : memref<i32>
-// CHECK: %[[x:.*]] = load %arg0[] : memref<i32>
+// CHECK: %[[x:.*]] = memref.load %arg0[] : memref<i32>
// CHECK: store %[[x]], %arg1[] : memref<i32>
return
}
// CHECK-LABEL: func @affine_prefetch
func @affine_prefetch(%arg0 : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.prefetch %0[%i0 + symbol(%arg0) + 7], read, locality<3>, data : memref<10xf32>
}
// CHECK: %[[a:.*]] = addi %{{.*}}, %{{.*}} : index
// CHECK-NEXT: %[[c7:.*]] = constant 7 : index
// CHECK-NEXT: %[[b:.*]] = addi %[[a]], %[[c7]] : index
-// CHECK-NEXT: prefetch %[[v0:.*]][%[[b]]], read, locality<3>, data : memref<10xf32>
+// CHECK-NEXT: memref.prefetch %[[v0:.*]][%[[b]]], read, locality<3>, data : memref<10xf32>
return
}
// CHECK-LABEL: func @affine_dma_start
func @affine_dma_start(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
- %1 = alloc() : memref<100xf32, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100xf32>
+ %1 = memref.alloc() : memref<100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -603,7 +603,7 @@ func @affine_dma_start(%arg0 : index) {
// CHECK-LABEL: func @affine_dma_wait
func @affine_dma_wait(%arg0 : index) {
- %2 = alloc() : memref<1xi32>
+ %2 = memref.alloc() : memref<1xi32>
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
affine.dma_wait %2[%i0 + %arg0 + 17], %c64 : memref<1xi32>
@@ -694,15 +694,15 @@ func @affine_parallel_tiled(%o: memref<100x100xf32>, %a: memref<100x100xf32>, %b
// CHECK-DAG: %[[C1_9:.*]] = constant 1
// CHECK-DAG: %[[C1_10:.*]] = constant 1
// CHECK: scf.parallel (%[[arg6:.*]], %[[arg7:.*]], %[[arg8:.*]]) = (%[[arg3]], %[[arg4]], %[[arg5]]) to (%[[A0]], %[[A1]], %[[A2]]) step (%[[C1]], %[[C1_9]], %[[C1_10]]) {
-// CHECK: %[[A3:.*]] = load %[[ARG1]][%[[arg6]], %[[arg8]]] : memref<100x100xf32>
-// CHECK: %[[A4:.*]] = load %[[ARG2]][%[[arg8]], %[[arg7]]] : memref<100x100xf32>
+// CHECK: %[[A3:.*]] = memref.load %[[ARG1]][%[[arg6]], %[[arg8]]] : memref<100x100xf32>
+// CHECK: %[[A4:.*]] = memref.load %[[ARG2]][%[[arg8]], %[[arg7]]] : memref<100x100xf32>
// CHECK: mulf %[[A3]], %[[A4]] : f32
// CHECK: scf.yield
/////////////////////////////////////////////////////////////////////
func @affine_parallel_simple(%arg0: memref<3x3xf32>, %arg1: memref<3x3xf32>) -> (memref<3x3xf32>) {
- %O = alloc() : memref<3x3xf32>
+ %O = memref.alloc() : memref<3x3xf32>
affine.parallel (%kx, %ky) = (0, 0) to (2, 2) {
%1 = affine.load %arg0[%kx, %ky] : memref<3x3xf32>
%2 = affine.load %arg1[%kx, %ky] : memref<3x3xf32>
@@ -719,8 +719,8 @@ func @affine_parallel_simple(%arg0: memref<3x3xf32>, %arg1: memref<3x3xf32>) ->
// CHECK-NEXT: %[[STEP_1:.*]] = constant 1 : index
// CHECK-NEXT: %[[STEP_2:.*]] = constant 1 : index
// CHECK-NEXT: scf.parallel (%[[I:.*]], %[[J:.*]]) = (%[[LOWER_1]], %[[LOWER_2]]) to (%[[UPPER_1]], %[[UPPER_2]]) step (%[[STEP_1]], %[[STEP_2]]) {
-// CHECK-NEXT: %[[VAL_1:.*]] = load
-// CHECK-NEXT: %[[VAL_2:.*]] = load
+// CHECK-NEXT: %[[VAL_1:.*]] = memref.load
+// CHECK-NEXT: %[[VAL_2:.*]] = memref.load
// CHECK-NEXT: %[[PRODUCT:.*]] = mulf
// CHECK-NEXT: store
// CHECK-NEXT: scf.yield
@@ -732,7 +732,7 @@ func @affine_parallel_simple(%arg0: memref<3x3xf32>, %arg1: memref<3x3xf32>) ->
func @affine_parallel_simple_dynamic_bounds(%arg0: memref<?x?xf32>, %arg1: memref<?x?xf32>, %arg2: memref<?x?xf32>) {
%c_0 = constant 0 : index
- %output_dim = dim %arg0, %c_0 : memref<?x?xf32>
+ %output_dim = memref.dim %arg0, %c_0 : memref<?x?xf32>
affine.parallel (%kx, %ky) = (%c_0, %c_0) to (%output_dim, %output_dim) {
%1 = affine.load %arg0[%kx, %ky] : memref<?x?xf32>
%2 = affine.load %arg1[%kx, %ky] : memref<?x?xf32>
@@ -744,14 +744,14 @@ func @affine_parallel_simple_dynamic_bounds(%arg0: memref<?x?xf32>, %arg1: memre
// CHECK-LABEL: func @affine_parallel_simple_dynamic_bounds
// CHECK-SAME: %[[ARG_0:.*]]: memref<?x?xf32>, %[[ARG_1:.*]]: memref<?x?xf32>, %[[ARG_2:.*]]: memref<?x?xf32>
// CHECK: %[[DIM_INDEX:.*]] = constant 0 : index
-// CHECK-NEXT: %[[UPPER:.*]] = dim %[[ARG_0]], %[[DIM_INDEX]] : memref<?x?xf32>
+// CHECK-NEXT: %[[UPPER:.*]] = memref.dim %[[ARG_0]], %[[DIM_INDEX]] : memref<?x?xf32>
// CHECK-NEXT: %[[LOWER_1:.*]] = constant 0 : index
// CHECK-NEXT: %[[LOWER_2:.*]] = constant 0 : index
// CHECK-NEXT: %[[STEP_1:.*]] = constant 1 : index
// CHECK-NEXT: %[[STEP_2:.*]] = constant 1 : index
// CHECK-NEXT: scf.parallel (%[[I:.*]], %[[J:.*]]) = (%[[LOWER_1]], %[[LOWER_2]]) to (%[[UPPER]], %[[UPPER]]) step (%[[STEP_1]], %[[STEP_2]]) {
-// CHECK-NEXT: %[[VAL_1:.*]] = load
-// CHECK-NEXT: %[[VAL_2:.*]] = load
+// CHECK-NEXT: %[[VAL_1:.*]] = memref.load
+// CHECK-NEXT: %[[VAL_2:.*]] = memref.load
// CHECK-NEXT: %[[PRODUCT:.*]] = mulf
// CHECK-NEXT: store
// CHECK-NEXT: scf.yield
@@ -781,8 +781,8 @@ func @affine_parallel_with_reductions(%arg0: memref<3x3xf32>, %arg1: memref<3x3x
// CHECK-NEXT: %[[INIT_1:.*]] = constant 0.000000e+00 : f32
// CHECK-NEXT: %[[INIT_2:.*]] = constant 1.000000e+00 : f32
// CHECK-NEXT: %[[RES:.*]] = scf.parallel (%[[I:.*]], %[[J:.*]]) = (%[[LOWER_1]], %[[LOWER_2]]) to (%[[UPPER_1]], %[[UPPER_2]]) step (%[[STEP_1]], %[[STEP_2]]) init (%[[INIT_1]], %[[INIT_2]]) -> (f32, f32) {
-// CHECK-NEXT: %[[VAL_1:.*]] = load
-// CHECK-NEXT: %[[VAL_2:.*]] = load
+// CHECK-NEXT: %[[VAL_1:.*]] = memref.load
+// CHECK-NEXT: %[[VAL_2:.*]] = memref.load
// CHECK-NEXT: %[[PRODUCT:.*]] = mulf
// CHECK-NEXT: %[[SUM:.*]] = addf
// CHECK-NEXT: scf.reduce(%[[PRODUCT]]) : f32 {
diff --git a/mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir b/mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir
index 3d444a5c2853..ba2a3914145c 100644
--- a/mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir
+++ b/mlir/test/Conversion/AsyncToLLVM/convert-to-llvm.mlir
@@ -20,7 +20,7 @@ func @execute_no_async_args(%arg0: f32, %arg1: memref<1xf32>) {
// CHECK: %[[TOKEN:.*]] = call @async_execute_fn(%arg0, %arg1)
%token = async.execute {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
// CHECK: call @mlirAsyncRuntimeAwaitToken(%[[TOKEN]])
@@ -51,7 +51,7 @@ func @execute_no_async_args(%arg0: f32, %arg1: memref<1xf32>) {
// Resume coroutine after suspension.
// CHECK: ^[[RESUME]]:
-// CHECK: store %arg0, %arg1[%c0] : memref<1xf32>
+// CHECK: memref.store %arg0, %arg1[%c0] : memref<1xf32>
// CHECK: call @mlirAsyncRuntimeEmplaceToken(%[[RET]])
// Delete coroutine.
@@ -74,12 +74,12 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
%token1 = async.execute {
%c1 = constant 1: index
- store %arg0, %arg2[%c0] : memref<1xf32>
+ memref.store %arg0, %arg2[%c0] : memref<1xf32>
async.yield
}
async.await %token1 : !async.token
- store %arg1, %arg2[%c0] : memref<1xf32>
+ memref.store %arg1, %arg2[%c0] : memref<1xf32>
async.yield
}
// CHECK: call @mlirAsyncRuntimeAwaitToken(%[[TOKEN]])
@@ -95,7 +95,7 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
// CHECK: %[[HDL_0:.*]] = llvm.intr.coro.begin
// CHECK: call @mlirAsyncRuntimeExecute
// CHECK: llvm.intr.coro.suspend
-// CHECK: store %arg0, %arg1[%arg2] : memref<1xf32>
+// CHECK: memref.store %arg0, %arg1[%arg2] : memref<1xf32>
// CHECK: call @mlirAsyncRuntimeEmplaceToken(%[[RET_0]])
// Function outlined from the outer async.execute operation.
@@ -115,7 +115,7 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
// CHECK: llvm.intr.coro.suspend
// Emplace result token after second resumption.
-// CHECK: store %arg2, %arg1[%c0] : memref<1xf32>
+// CHECK: memref.store %arg2, %arg1[%c0] : memref<1xf32>
// CHECK: call @mlirAsyncRuntimeEmplaceToken(%[[RET_1]])
// -----
@@ -125,13 +125,13 @@ func @async_execute_token_dependency(%arg0: f32, %arg1: memref<1xf32>) {
// CHECK: %0 = call @async_execute_fn(%arg0, %arg1)
%token = async.execute {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
// CHECK: %1 = call @async_execute_fn_0(%0, %arg0, %arg1)
%token_0 = async.execute [%token] {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
return
@@ -144,7 +144,7 @@ func @async_execute_token_dependency(%arg0: f32, %arg1: memref<1xf32>) {
// CHECK: %[[HDL_0:.*]] = llvm.intr.coro.begin
// CHECK: call @mlirAsyncRuntimeExecute
// CHECK: llvm.intr.coro.suspend
-// CHECK: store %arg0, %arg1[%c0] : memref<1xf32>
+// CHECK: memref.store %arg0, %arg1[%c0] : memref<1xf32>
// CHECK: call @mlirAsyncRuntimeEmplaceToken(%[[RET_0]])
// Function outlined from the second async.execute operation with dependency.
@@ -163,7 +163,7 @@ func @async_execute_token_dependency(%arg0: f32, %arg1: memref<1xf32>) {
// CHECK: llvm.intr.coro.suspend
// Emplace result token after second resumption.
-// CHECK: store %arg1, %arg2[%c0] : memref<1xf32>
+// CHECK: memref.store %arg1, %arg2[%c0] : memref<1xf32>
// CHECK: call @mlirAsyncRuntimeEmplaceToken(%[[RET_1]])
// -----
diff --git a/mlir/test/Conversion/GPUCommon/memory-attrbution.mlir b/mlir/test/Conversion/GPUCommon/memory-attrbution.mlir
index 7f9b8b5fd76d..ff8d16b8c6fd 100644
--- a/mlir/test/Conversion/GPUCommon/memory-attrbution.mlir
+++ b/mlir/test/Conversion/GPUCommon/memory-attrbution.mlir
@@ -42,7 +42,7 @@ gpu.module @kernel {
// ROCDL: llvm.getelementptr
// ROCDL: llvm.store
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<4xf32, 5>
+ memref.store %arg0, %arg1[%c0] : memref<4xf32, 5>
"terminator"() : () -> ()
}
@@ -108,7 +108,7 @@ gpu.module @kernel {
// ROCDL: llvm.getelementptr
// ROCDL: llvm.store
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<4xf32, 3>
+ memref.store %arg0, %arg1[%c0] : memref<4xf32, 3>
"terminator"() : () -> ()
}
@@ -178,7 +178,7 @@ gpu.module @kernel {
// ROCDL: %[[descr10:.*]] = llvm.insertvalue %[[c1]], %[[descr9]][4, 2]
%c0 = constant 0 : index
- store %arg0, %arg1[%c0,%c0,%c0] : memref<4x2x6xf32, 3>
+ memref.store %arg0, %arg1[%c0,%c0,%c0] : memref<4x2x6xf32, 3>
"terminator"() : () -> ()
}
}
@@ -222,10 +222,10 @@ gpu.module @kernel {
// ROCDL: llvm.alloca %[[c4]] x f32 : (i64) -> !llvm.ptr<f32, 5>
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32, 3>
- store %arg0, %arg2[%c0] : memref<2xf32, 3>
- store %arg0, %arg3[%c0] : memref<3xf32, 5>
- store %arg0, %arg4[%c0] : memref<4xf32, 5>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32, 3>
+ memref.store %arg0, %arg2[%c0] : memref<2xf32, 3>
+ memref.store %arg0, %arg3[%c0] : memref<3xf32, 5>
+ memref.store %arg0, %arg4[%c0] : memref<4xf32, 5>
"terminator"() : () -> ()
}
}
diff --git a/mlir/test/Conversion/GPUToSPIRV/load-store.mlir b/mlir/test/Conversion/GPUToSPIRV/load-store.mlir
index 94543eb73027..ae1649d28120 100644
--- a/mlir/test/Conversion/GPUToSPIRV/load-store.mlir
+++ b/mlir/test/Conversion/GPUToSPIRV/load-store.mlir
@@ -17,7 +17,7 @@ module attributes {
%c1_2 = constant 1 : index
gpu.launch_func @kernels::@load_store_kernel
blocks in (%0, %c1_2, %c1_2) threads in (%1, %c1_2, %c1_2)
- args(%arg0 : memref<12x4xf32>, %arg1 : memref<12x4xf32>, %arg2 : memref<12x4xf32>,
+ args(%arg0 : memref<12x4xf32>, %arg1 : memref<12x4xf32>, %arg2 : memref<12x4xf32>,
%c0 : index, %c0_0 : index, %c1 : index, %c1_1 : index)
return
}
@@ -69,15 +69,15 @@ module attributes {
// CHECK: %[[OFFSET1_2:.*]] = spv.IAdd %[[OFFSET1_1]], %[[UPDATE1_2]] : i32
// CHECK: %[[PTR1:.*]] = spv.AccessChain %[[ARG0]]{{\[}}%[[ZERO]], %[[OFFSET1_2]]{{\]}}
// CHECK-NEXT: %[[VAL1:.*]] = spv.Load "StorageBuffer" %[[PTR1]]
- %14 = load %arg0[%12, %13] : memref<12x4xf32>
+ %14 = memref.load %arg0[%12, %13] : memref<12x4xf32>
// CHECK: %[[PTR2:.*]] = spv.AccessChain %[[ARG1]]{{\[}}{{%.*}}, {{%.*}}{{\]}}
// CHECK-NEXT: %[[VAL2:.*]] = spv.Load "StorageBuffer" %[[PTR2]]
- %15 = load %arg1[%12, %13] : memref<12x4xf32>
+ %15 = memref.load %arg1[%12, %13] : memref<12x4xf32>
// CHECK: %[[VAL3:.*]] = spv.FAdd %[[VAL1]], %[[VAL2]]
%16 = addf %14, %15 : f32
// CHECK: %[[PTR3:.*]] = spv.AccessChain %[[ARG2]]{{\[}}{{%.*}}, {{%.*}}{{\]}}
// CHECK-NEXT: spv.Store "StorageBuffer" %[[PTR3]], %[[VAL3]]
- store %16, %arg2[%12, %13] : memref<12x4xf32>
+ memref.store %16, %arg2[%12, %13] : memref<12x4xf32>
gpu.return
}
}
diff --git a/mlir/test/Conversion/GPUToVulkan/lower-gpu-launch-vulkan-launch.mlir b/mlir/test/Conversion/GPUToVulkan/lower-gpu-launch-vulkan-launch.mlir
index cb80864f20cc..d451c0a9ffe7 100644
--- a/mlir/test/Conversion/GPUToVulkan/lower-gpu-launch-vulkan-launch.mlir
+++ b/mlir/test/Conversion/GPUToVulkan/lower-gpu-launch-vulkan-launch.mlir
@@ -1,6 +1,6 @@
// RUN: mlir-opt %s -convert-gpu-launch-to-vulkan-launch | FileCheck %s
-// CHECK: %[[resource:.*]] = alloc() : memref<12xf32>
+// CHECK: %[[resource:.*]] = memref.alloc() : memref<12xf32>
// CHECK: %[[index:.*]] = constant 1 : index
// CHECK: call @vulkanLaunch(%[[index]], %[[index]], %[[index]], %[[resource]]) {spirv_blob = "{{.*}}", spirv_entry_point = "kernel"}
@@ -24,7 +24,7 @@ module attributes {gpu.container_module} {
}
}
func @foo() {
- %0 = alloc() : memref<12xf32>
+ %0 = memref.alloc() : memref<12xf32>
%c1 = constant 1 : index
gpu.launch_func @kernels::@kernel
blocks in(%c1, %c1, %c1)
diff --git a/mlir/test/Conversion/LinalgToVector/linalg-to-vector.mlir b/mlir/test/Conversion/LinalgToVector/linalg-to-vector.mlir
index 21aba6cdaf7b..27fbb29023e9 100644
--- a/mlir/test/Conversion/LinalgToVector/linalg-to-vector.mlir
+++ b/mlir/test/Conversion/LinalgToVector/linalg-to-vector.mlir
@@ -18,15 +18,15 @@ func @conv_1d(%arg0: memref<?xf32>, %arg1: memref<?xf32>, %arg2: memref<?xf32>)
// CHECK-DAG: %[[c3:.*]] = constant 3 : index
// CHECK-DAG: %[[c0:.*]] = constant 0 : index
// CHECK-DAG: %[[c1:.*]] = constant 1 : index
-// CHECK: %[[v0:.*]] = dim %[[arg1]], %[[c0]] : memref<?xf32>
-// CHECK: %[[v1:.*]] = dim %[[arg2]], %[[c0]] : memref<?xf32>
-// CHECK: %[[v2:.*]] = dim %[[arg0]], %[[c0]] : memref<?xf32>
-// CHECK: %[[v3:.*]] = alloc(%[[c12]]) : memref<?xi8>
-// CHECK: %[[v4:.*]] = alloc(%[[c12]]) : memref<?xi8>
-// CHECK: %[[v5:.*]] = alloc(%[[c4]]) : memref<?xi8>
-// CHECK: %[[v6:.*]] = std.view %[[v3]][%[[c0]]][] : memref<?xi8> to memref<3xf32>
-// CHECK: %[[v7:.*]] = std.view %[[v4]][%[[c0]]][] : memref<?xi8> to memref<3xf32>
-// CHECK: %[[v8:.*]] = std.view %[[v5]][%[[c0]]][] : memref<?xi8> to memref<1xf32>
+// CHECK: %[[v0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?xf32>
+// CHECK: %[[v1:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?xf32>
+// CHECK: %[[v2:.*]] = memref.dim %[[arg0]], %[[c0]] : memref<?xf32>
+// CHECK: %[[v3:.*]] = memref.alloc(%[[c12]]) : memref<?xi8>
+// CHECK: %[[v4:.*]] = memref.alloc(%[[c12]]) : memref<?xi8>
+// CHECK: %[[v5:.*]] = memref.alloc(%[[c4]]) : memref<?xi8>
+// CHECK: %[[v6:.*]] = memref.view %[[v3]][%[[c0]]][] : memref<?xi8> to memref<3xf32>
+// CHECK: %[[v7:.*]] = memref.view %[[v4]][%[[c0]]][] : memref<?xi8> to memref<3xf32>
+// CHECK: %[[v8:.*]] = memref.view %[[v5]][%[[c0]]][] : memref<?xi8> to memref<1xf32>
// CHECK: scf.for %[[arg3:.*]] = %[[c0]] to %[[v1]] step %[[c1]] {
// CHECK: %[[v9:.*]] = affine.min #[[$map0]](%[[arg3]])[%[[v1]]]
// CHECK: %[[v10:.*]] = subview %[[arg2]][%[[arg3]]] [%[[v9]]] [1] : memref<?xf32> to memref<?xf32, #[[$map1]]>
diff --git a/mlir/test/Conversion/SCFToGPU/no_blocks_no_threads.mlir b/mlir/test/Conversion/SCFToGPU/no_blocks_no_threads.mlir
index 451fcbe173da..3638fe50a268 100644
--- a/mlir/test/Conversion/SCFToGPU/no_blocks_no_threads.mlir
+++ b/mlir/test/Conversion/SCFToGPU/no_blocks_no_threads.mlir
@@ -21,11 +21,11 @@ func @one_d_loop(%A : memref<?xf32>, %B : memref<?xf32>) {
// CHECK-BLOCKS-NEXT: gpu.launch blocks(%[[B0:.*]], %[[B1:.*]], %[[B2:.*]]) in (%{{.*}} = %[[BOUND]], %{{.*}} = %[[ONE]], %{{.*}}0 = %[[ONE]]) threads(%[[T0:.*]], %[[T1:.*]], %[[T2:.*]]) in (%{{.*}} = %[[ONE]], %{{.*}} = %[[ONE]], %{{.*}} = %[[ONE]])
affine.for %i = 0 to 42 {
// CHECK-THREADS-NEXT: %[[INDEX:.*]] = addi %{{.*}}, %[[T0]]
- // CHECK-THREADS-NEXT: load %{{.*}}[%[[INDEX]]]
+ // CHECK-THREADS-NEXT: memref.load %{{.*}}[%[[INDEX]]]
// CHECK-BLOCKS-NEXT: %[[INDEX:.*]] = addi %{{.*}}, %[[B0]]
- // CHECK-BLOCKS-NEXT: load %{{.*}}[%[[INDEX]]]
- %0 = load %A[%i] : memref<?xf32>
- store %0, %B[%i] : memref<?xf32>
+ // CHECK-BLOCKS-NEXT: memref.load %{{.*}}[%[[INDEX]]]
+ %0 = memref.load %A[%i] : memref<?xf32>
+ memref.store %0, %B[%i] : memref<?xf32>
// CHECK-THREADS: gpu.terminator
// CHECK-BLOCKS: gpu.terminator
}
diff --git a/mlir/test/Conversion/SCFToGPU/parallel_loop.mlir b/mlir/test/Conversion/SCFToGPU/parallel_loop.mlir
index e72aabe3bbdb..b4851a4c7c2b 100644
--- a/mlir/test/Conversion/SCFToGPU/parallel_loop.mlir
+++ b/mlir/test/Conversion/SCFToGPU/parallel_loop.mlir
@@ -9,8 +9,8 @@ func @parallel_loop_bidy_bidx(%arg0 : index, %arg1 : index, %arg2 : index,
%step = constant 2 : index
scf.parallel (%i0, %i1) = (%arg0, %arg1) to (%arg2, %arg3)
step (%arg4, %step) {
- %val = load %buf[%i0, %i1] : memref<?x?xf32>
- store %val, %res[%i1, %i0] : memref<?x?xf32>
+ %val = memref.load %buf[%i0, %i1] : memref<?x?xf32>
+ memref.store %val, %res[%i1, %i0] : memref<?x?xf32>
} { mapping = [{processor = 1, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}, {processor = 0, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}] }
return
}
@@ -28,8 +28,8 @@ func @parallel_loop_bidy_bidx(%arg0 : index, %arg1 : index, %arg2 : index,
// CHECK: gpu.launch blocks([[VAL_11:%.*]], [[VAL_12:%.*]], [[VAL_13:%.*]]) in ([[VAL_14:%.*]] = [[VAL_10]], [[VAL_15:%.*]] = [[VAL_9]], [[VAL_16:%.*]] = [[VAL_8]]) threads([[VAL_17:%.*]], [[VAL_18:%.*]], [[VAL_19:%.*]]) in ([[VAL_20:%.*]] = [[VAL_8]], [[VAL_21:%.*]] = [[VAL_8]], [[VAL_22:%.*]] = [[VAL_8]]) {
// CHECK: [[VAL_23:%.*]] = affine.apply #[[$MAP1]]([[VAL_12]]){{\[}}[[VAL_4]], [[VAL_0]]]
// CHECK: [[VAL_24:%.*]] = affine.apply #[[$MAP1]]([[VAL_11]]){{\[}}[[VAL_7]], [[VAL_1]]]
-// CHECK: [[VAL_25:%.*]] = load [[VAL_5]]{{\[}}[[VAL_23]], [[VAL_24]]] : memref<?x?xf32>
-// CHECK: store [[VAL_25]], [[VAL_6]]{{\[}}[[VAL_24]], [[VAL_23]]] : memref<?x?xf32>
+// CHECK: [[VAL_25:%.*]] = memref.load [[VAL_5]]{{\[}}[[VAL_23]], [[VAL_24]]] : memref<?x?xf32>
+// CHECK: memref.store [[VAL_25]], [[VAL_6]]{{\[}}[[VAL_24]], [[VAL_23]]] : memref<?x?xf32>
// CHECK: gpu.terminator
// CHECK: }
// CHECK: return
@@ -53,8 +53,8 @@ func @parallel_loop_tiled(%arg0 : index, %arg1 : index, %arg2 : index,
step (%one, %one) {
%idx0 = addi %i0, %si0 : index
%idx1 = addi %i1, %si1 : index
- %val = load %buf[%idx0, %idx1] : memref<?x?xf32>
- store %val, %res[%idx1, %idx0] : memref<?x?xf32>
+ %val = memref.load %buf[%idx0, %idx1] : memref<?x?xf32>
+ memref.store %val, %res[%idx1, %idx0] : memref<?x?xf32>
} { mapping = [
{processor = 4, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>},
{processor = 3, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}
@@ -87,8 +87,8 @@ func @parallel_loop_tiled(%arg0 : index, %arg1 : index, %arg2 : index,
// CHECK: [[VAL_55:%.*]] = affine.apply #[[$MAP1]]([[VAL_46]]){{\[}}[[VAL_33]], [[VAL_32]]]
// CHECK: [[VAL_56:%.*]] = addi [[VAL_52]], [[VAL_54]] : index
// CHECK: [[VAL_57:%.*]] = addi [[VAL_53]], [[VAL_55]] : index
-// CHECK: [[VAL_58:%.*]] = load [[VAL_30]]{{\[}}[[VAL_56]], [[VAL_57]]] : memref<?x?xf32>
-// CHECK: store [[VAL_58]], [[VAL_31]]{{\[}}[[VAL_57]], [[VAL_56]]] : memref<?x?xf32>
+// CHECK: [[VAL_58:%.*]] = memref.load [[VAL_30]]{{\[}}[[VAL_56]], [[VAL_57]]] : memref<?x?xf32>
+// CHECK: memref.store [[VAL_58]], [[VAL_31]]{{\[}}[[VAL_57]], [[VAL_56]]] : memref<?x?xf32>
// CHECK: gpu.terminator
// CHECK: }
// CHECK: return
@@ -106,8 +106,8 @@ func @parallel_loop_bidy_seq(%arg0 : index, %arg1 : index, %arg2 : index,
%step = constant 2 : index
scf.parallel (%i0, %i1) = (%arg0, %arg1) to (%arg2, %arg3)
step (%arg4, %step) {
- %val = load %buf[%i0, %i1] : memref<?x?xf32>
- store %val, %res[%i1, %i0] : memref<?x?xf32>
+ %val = memref.load %buf[%i0, %i1] : memref<?x?xf32>
+ memref.store %val, %res[%i1, %i0] : memref<?x?xf32>
} { mapping = [
{processor = 1, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>},
{processor = 6, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}
@@ -127,8 +127,8 @@ func @parallel_loop_bidy_seq(%arg0 : index, %arg1 : index, %arg2 : index,
// CHECK: gpu.launch blocks([[VAL_69:%.*]], [[VAL_70:%.*]], [[VAL_71:%.*]]) in ([[VAL_72:%.*]] = [[VAL_67]], [[VAL_73:%.*]] = [[VAL_68]], [[VAL_74:%.*]] = [[VAL_67]]) threads([[VAL_75:%.*]], [[VAL_76:%.*]], [[VAL_77:%.*]]) in ([[VAL_78:%.*]] = [[VAL_67]], [[VAL_79:%.*]] = [[VAL_67]], [[VAL_80:%.*]] = [[VAL_67]]) {
// CHECK: [[VAL_81:%.*]] = affine.apply #[[$MAP1]]([[VAL_70]]){{\[}}[[VAL_63]], [[VAL_59]]]
// CHECK: scf.for [[VAL_82:%.*]] = [[VAL_60]] to [[VAL_62]] step [[VAL_66]] {
-// CHECK: [[VAL_83:%.*]] = load [[VAL_64]]{{\[}}[[VAL_81]], [[VAL_82]]] : memref<?x?xf32>
-// CHECK: store [[VAL_83]], [[VAL_65]]{{\[}}[[VAL_82]], [[VAL_81]]] : memref<?x?xf32>
+// CHECK: [[VAL_83:%.*]] = memref.load [[VAL_64]]{{\[}}[[VAL_81]], [[VAL_82]]] : memref<?x?xf32>
+// CHECK: memref.store [[VAL_83]], [[VAL_65]]{{\[}}[[VAL_82]], [[VAL_81]]] : memref<?x?xf32>
// CHECK: }
// CHECK: gpu.terminator
// CHECK: }
@@ -153,8 +153,8 @@ func @parallel_loop_tiled_seq(%arg0 : index, %arg1 : index, %arg2 : index,
step (%one, %one) {
%idx0 = addi %i0, %si0 : index
%idx1 = addi %i1, %si1 : index
- %val = load %buf[%idx0, %idx1] : memref<?x?xf32>
- store %val, %res[%idx1, %idx0] : memref<?x?xf32>
+ %val = memref.load %buf[%idx0, %idx1] : memref<?x?xf32>
+ memref.store %val, %res[%idx1, %idx0] : memref<?x?xf32>
} { mapping = [
{processor = 4, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>},
{processor = 6, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}
@@ -185,8 +185,8 @@ func @parallel_loop_tiled_seq(%arg0 : index, %arg1 : index, %arg2 : index,
// CHECK: scf.for [[VAL_111:%.*]] = [[VAL_90]] to [[VAL_92]] step [[VAL_91]] {
// CHECK: [[VAL_112:%.*]] = addi [[VAL_108]], [[VAL_110]] : index
// CHECK: [[VAL_113:%.*]] = addi [[VAL_109]], [[VAL_111]] : index
-// CHECK: [[VAL_114:%.*]] = load [[VAL_88]]{{\[}}[[VAL_112]], [[VAL_113]]] : memref<?x?xf32>
-// CHECK: store [[VAL_114]], [[VAL_89]]{{\[}}[[VAL_113]], [[VAL_112]]] : memref<?x?xf32>
+// CHECK: [[VAL_114:%.*]] = memref.load [[VAL_88]]{{\[}}[[VAL_112]], [[VAL_113]]] : memref<?x?xf32>
+// CHECK: memref.store [[VAL_114]], [[VAL_89]]{{\[}}[[VAL_113]], [[VAL_112]]] : memref<?x?xf32>
// CHECK: }
// CHECK: }
// CHECK: gpu.terminator
@@ -208,31 +208,31 @@ module {
%c0 = constant 0 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32, #map0>
- %1 = dim %arg0, %c1 : memref<?x?xf32, #map0>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32, #map0>
+ %1 = memref.dim %arg0, %c1 : memref<?x?xf32, #map0>
scf.parallel (%arg3, %arg4) = (%c0, %c0) to (%0, %1) step (%c2, %c3) {
- %2 = dim %arg0, %c0 : memref<?x?xf32, #map0>
+ %2 = memref.dim %arg0, %c0 : memref<?x?xf32, #map0>
%3 = affine.min #map1(%arg3)[%2]
%squared_min = muli %3, %3 : index
- %4 = dim %arg0, %c1 : memref<?x?xf32, #map0>
+ %4 = memref.dim %arg0, %c1 : memref<?x?xf32, #map0>
%5 = affine.min #map2(%arg4)[%4]
- %6 = std.subview %arg0[%arg3, %arg4][%squared_min, %5][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
- %7 = dim %arg1, %c0 : memref<?x?xf32, #map0>
+ %6 = memref.subview %arg0[%arg3, %arg4][%squared_min, %5][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
+ %7 = memref.dim %arg1, %c0 : memref<?x?xf32, #map0>
%8 = affine.min #map1(%arg3)[%7]
- %9 = dim %arg1, %c1 : memref<?x?xf32, #map0>
+ %9 = memref.dim %arg1, %c1 : memref<?x?xf32, #map0>
%10 = affine.min #map2(%arg4)[%9]
- %11 = std.subview %arg1[%arg3, %arg4][%8, %10][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
- %12 = dim %arg2, %c0 : memref<?x?xf32, #map0>
+ %11 = memref.subview %arg1[%arg3, %arg4][%8, %10][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
+ %12 = memref.dim %arg2, %c0 : memref<?x?xf32, #map0>
%13 = affine.min #map1(%arg3)[%12]
- %14 = dim %arg2, %c1 : memref<?x?xf32, #map0>
+ %14 = memref.dim %arg2, %c1 : memref<?x?xf32, #map0>
%15 = affine.min #map2(%arg4)[%14]
- %16 = std.subview %arg2[%arg3, %arg4][%13, %15][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
+ %16 = memref.subview %arg2[%arg3, %arg4][%13, %15][%c1, %c1] : memref<?x?xf32, #map0> to memref<?x?xf32, #map3>
scf.parallel (%arg5, %arg6) = (%c0, %c0) to (%squared_min, %5) step (%c1, %c1) {
- %17 = load %6[%arg5, %arg6] : memref<?x?xf32, #map3>
- %18 = load %11[%arg5, %arg6] : memref<?x?xf32, #map3>
- %19 = load %16[%arg5, %arg6] : memref<?x?xf32, #map3>
+ %17 = memref.load %6[%arg5, %arg6] : memref<?x?xf32, #map3>
+ %18 = memref.load %11[%arg5, %arg6] : memref<?x?xf32, #map3>
+ %19 = memref.load %16[%arg5, %arg6] : memref<?x?xf32, #map3>
%20 = addf %17, %18 : f32
- store %20, %16[%arg5, %arg6] : memref<?x?xf32, #map3>
+ memref.store %20, %16[%arg5, %arg6] : memref<?x?xf32, #map3>
scf.yield
} {mapping = [{bound = affine_map<(d0) -> (d0)>, map = affine_map<(d0) -> (d0)>, processor = 3 : i64}, {bound = affine_map<(d0) -> (d0)>, map = affine_map<(d0) -> (d0)>, processor = 4 : i64}]}
scf.yield
@@ -255,8 +255,8 @@ module {
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C3:.*]] = constant 3 : index
// CHECK: %[[C2:.*]] = constant 2 : index
-// CHECK: [[VAL_7:%.*]] = dim [[VAL_0]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
-// CHECK: [[VAL_8:%.*]] = dim [[VAL_0]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_7:%.*]] = memref.dim [[VAL_0]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_8:%.*]] = memref.dim [[VAL_0]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_9:%.*]] = constant 1 : index
// CHECK: [[VAL_10:%.*]] = affine.apply #[[$MAP1]]([[VAL_7]]){{\[}}%[[C0]], %[[C2]]]
// CHECK: [[VAL_11:%.*]] = affine.apply #[[$MAP1]]([[VAL_8]]){{\[}}%[[C0]], %[[C3]]]
@@ -267,33 +267,33 @@ module {
// CHECK: gpu.launch blocks([[VAL_16:%.*]], [[VAL_17:%.*]], [[VAL_18:%.*]]) in ([[VAL_19:%.*]] = [[VAL_10]], [[VAL_20:%.*]] = [[VAL_11]], [[VAL_21:%.*]] = [[VAL_9]]) threads([[VAL_22:%.*]], [[VAL_23:%.*]], [[VAL_24:%.*]]) in ([[VAL_25:%.*]] = [[VAL_13]], [[VAL_26:%.*]] = [[VAL_15]], [[VAL_27:%.*]] = [[VAL_9]]) {
// CHECK: [[VAL_28:%.*]] = affine.apply #[[$MAP2]]([[VAL_16]]){{\[}}%[[C2]], %[[C0]]]
// CHECK: [[VAL_29:%.*]] = affine.apply #[[$MAP2]]([[VAL_17]]){{\[}}%[[C3]], %[[C0]]]
-// CHECK: [[VAL_30:%.*]] = dim [[VAL_0]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_30:%.*]] = memref.dim [[VAL_0]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_31:%.*]] = affine.min #[[$MAP3]]([[VAL_28]]){{\[}}[[VAL_30]]]
// CHECK: [[VAL_31_SQUARED:%.*]] = muli [[VAL_31]], [[VAL_31]] : index
-// CHECK: [[VAL_32:%.*]] = dim [[VAL_0]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_32:%.*]] = memref.dim [[VAL_0]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_33:%.*]] = affine.min #[[$MAP4]]([[VAL_29]]){{\[}}[[VAL_32]]]
-// CHECK: [[VAL_34:%.*]] = subview [[VAL_0]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_31_SQUARED]], [[VAL_33]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
-// CHECK: [[VAL_35:%.*]] = dim [[VAL_1]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_34:%.*]] = memref.subview [[VAL_0]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_31_SQUARED]], [[VAL_33]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_35:%.*]] = memref.dim [[VAL_1]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_36:%.*]] = affine.min #[[$MAP3]]([[VAL_28]]){{\[}}[[VAL_35]]]
-// CHECK: [[VAL_37:%.*]] = dim [[VAL_1]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_37:%.*]] = memref.dim [[VAL_1]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_38:%.*]] = affine.min #[[$MAP4]]([[VAL_29]]){{\[}}[[VAL_37]]]
-// CHECK: [[VAL_39:%.*]] = subview [[VAL_1]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_36]], [[VAL_38]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
-// CHECK: [[VAL_40:%.*]] = dim [[VAL_2]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_39:%.*]] = memref.subview [[VAL_1]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_36]], [[VAL_38]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_40:%.*]] = memref.dim [[VAL_2]], %[[C0]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_41:%.*]] = affine.min #[[$MAP3]]([[VAL_28]]){{\[}}[[VAL_40]]]
-// CHECK: [[VAL_42:%.*]] = dim [[VAL_2]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
+// CHECK: [[VAL_42:%.*]] = memref.dim [[VAL_2]], %[[C1]] : memref<?x?xf32, #[[$MAP0]]>
// CHECK: [[VAL_43:%.*]] = affine.min #[[$MAP4]]([[VAL_29]]){{\[}}[[VAL_42]]]
-// CHECK: [[VAL_44:%.*]] = subview [[VAL_2]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_41]], [[VAL_43]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_44:%.*]] = memref.subview [[VAL_2]]{{\[}}[[VAL_28]], [[VAL_29]]] {{\[}}[[VAL_41]], [[VAL_43]]] {{\[}}%[[C1]], %[[C1]]] : memref<?x?xf32, #[[$MAP0]]> to memref<?x?xf32, #[[$MAP5]]>
// CHECK: [[VAL_45:%.*]] = affine.apply #[[$MAP2]]([[VAL_22]]){{\[}}%[[C1]], %[[C0]]]
// CHECK: [[VAL_46:%.*]] = cmpi slt, [[VAL_45]], [[VAL_31_SQUARED]] : index
// CHECK: scf.if [[VAL_46]] {
// CHECK: [[VAL_47:%.*]] = affine.apply #[[$MAP2]]([[VAL_23]]){{\[}}%[[C1]], %[[C0]]]
// CHECK: [[VAL_48:%.*]] = cmpi slt, [[VAL_47]], [[VAL_33]] : index
// CHECK: scf.if [[VAL_48]] {
-// CHECK: [[VAL_49:%.*]] = load [[VAL_34]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
-// CHECK: [[VAL_50:%.*]] = load [[VAL_39]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
-// CHECK: [[VAL_51:%.*]] = load [[VAL_44]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_49:%.*]] = memref.load [[VAL_34]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_50:%.*]] = memref.load [[VAL_39]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
+// CHECK: [[VAL_51:%.*]] = memref.load [[VAL_44]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
// CHECK: [[VAL_52:%.*]] = addf [[VAL_49]], [[VAL_50]] : f32
-// CHECK: store [[VAL_52]], [[VAL_44]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
+// CHECK: memref.store [[VAL_52]], [[VAL_44]]{{\[}}[[VAL_45]], [[VAL_47]]] : memref<?x?xf32, #[[$MAP5]]>
// CHECK: }
// CHECK: }
// CHECK: gpu.terminator
@@ -353,8 +353,8 @@ func @parallel_loop_loop_variant_bound(%arg0 : index, %arg1 : index, %arg2 : ind
step (%one, %one) {
%idx0 = addi %i0, %si0 : index
%idx1 = addi %i1, %si1 : index
- %val = load %buf[%idx0, %idx1] : memref<?x?xf32>
- store %val, %res[%idx1, %idx0] : memref<?x?xf32>
+ %val = memref.load %buf[%idx0, %idx1] : memref<?x?xf32>
+ memref.store %val, %res[%idx1, %idx0] : memref<?x?xf32>
} { mapping = [
{processor = 4, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>},
{processor = 6, map = affine_map<(d0) -> (d0)>, bound = affine_map<(d0) -> (d0)>}
diff --git a/mlir/test/Conversion/SCFToGPU/step_one.mlir b/mlir/test/Conversion/SCFToGPU/step_one.mlir
index 631886a95403..e7d71aa1c1fe 100644
--- a/mlir/test/Conversion/SCFToGPU/step_one.mlir
+++ b/mlir/test/Conversion/SCFToGPU/step_one.mlir
@@ -64,12 +64,12 @@ func @step_1(%A : memref<?x?x?x?xf32>, %B : memref<?x?x?x?xf32>) {
// CHECK-22-NEXT: %[[jj:.*]] = addi %{{.*}}, %{{.*}} : index
// Using remapped values instead of loop iterators.
- // CHECK-11: {{.*}} = load %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
- // CHECK-22: {{.*}} = load %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
- %0 = load %A[%i, %j, %ii, %jj] : memref<?x?x?x?xf32>
- // CHECK-11-NEXT: store {{.*}}, %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
- // CHECK-22-NEXT: store {{.*}}, %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
- store %0, %B[%i, %j, %ii, %jj] : memref<?x?x?x?xf32>
+ // CHECK-11: {{.*}} = memref.load %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
+ // CHECK-22: {{.*}} = memref.load %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
+ %0 = memref.load %A[%i, %j, %ii, %jj] : memref<?x?x?x?xf32>
+ // CHECK-11-NEXT: memref.store {{.*}}, %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
+ // CHECK-22-NEXT: memref.store {{.*}}, %{{.*}}[%[[i]], %[[j]], %[[ii]], %[[jj]]] : memref<?x?x?x?xf32>
+ memref.store %0, %B[%i, %j, %ii, %jj] : memref<?x?x?x?xf32>
// CHECK-11: gpu.terminator
// CHECK-22: gpu.terminator
diff --git a/mlir/test/Conversion/SCFToGPU/step_positive.mlir b/mlir/test/Conversion/SCFToGPU/step_positive.mlir
index 850f87037351..8d486e9598ef 100644
--- a/mlir/test/Conversion/SCFToGPU/step_positive.mlir
+++ b/mlir/test/Conversion/SCFToGPU/step_positive.mlir
@@ -18,10 +18,10 @@ func @step_var(%A : memref<?x?xf32>, %B : memref<?x?xf32>) {
// CHECK-NEXT: %[[prod_j:.*]] = muli %{{.*}}, %{{.*}} : index
// CHECK-NEXT: %[[j:.*]] = addi %{{.*}}, %[[prod_j]] : index
- // CHECK: {{.*}} = load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32>
- %0 = load %A[%i, %j] : memref<?x?xf32>
- // CHECK: store {{.*}}, %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32>
- store %0, %B[%i, %j] : memref<?x?xf32>
+ // CHECK: {{.*}} = memref.load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32>
+ %0 = memref.load %A[%i, %j] : memref<?x?xf32>
+ // CHECK: memref.store {{.*}}, %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32>
+ memref.store %0, %B[%i, %j] : memref<?x?xf32>
}
}
return
diff --git a/mlir/test/Conversion/SCFToSPIRV/for.mlir b/mlir/test/Conversion/SCFToSPIRV/for.mlir
index 10c1557ce5dc..1b0aed94d912 100644
--- a/mlir/test/Conversion/SCFToSPIRV/for.mlir
+++ b/mlir/test/Conversion/SCFToSPIRV/for.mlir
@@ -36,8 +36,8 @@ func @loop_kernel(%arg2 : memref<10xf32>, %arg3 : memref<10xf32>) {
// CHECK: spv.mlir.merge
// CHECK: }
scf.for %arg4 = %lb to %ub step %step {
- %1 = load %arg2[%arg4] : memref<10xf32>
- store %1, %arg3[%arg4] : memref<10xf32>
+ %1 = memref.load %arg2[%arg4] : memref<10xf32>
+ memref.store %1, %arg3[%arg4] : memref<10xf32>
}
return
}
@@ -78,8 +78,8 @@ func @loop_yield(%arg2 : memref<10xf32>, %arg3 : memref<10xf32>) {
// CHECK-DAG: %[[OUT2:.*]] = spv.Load "Function" %[[VAR2]] : f32
// CHECK: spv.Store "StorageBuffer" {{%.*}}, %[[OUT1]] : f32
// CHECK: spv.Store "StorageBuffer" {{%.*}}, %[[OUT2]] : f32
- store %result#0, %arg3[%lb] : memref<10xf32>
- store %result#1, %arg3[%ub] : memref<10xf32>
+ memref.store %result#0, %arg3[%lb] : memref<10xf32>
+ memref.store %result#1, %arg3[%ub] : memref<10xf32>
return
}
diff --git a/mlir/test/Conversion/SCFToSPIRV/if.mlir b/mlir/test/Conversion/SCFToSPIRV/if.mlir
index 45dd7b599666..40bae1734c25 100644
--- a/mlir/test/Conversion/SCFToSPIRV/if.mlir
+++ b/mlir/test/Conversion/SCFToSPIRV/if.mlir
@@ -20,7 +20,7 @@ func @kernel_simple_selection(%arg2 : memref<10xf32>, %arg3 : i1) {
// CHECK-NEXT: spv.Return
scf.if %arg3 {
- store %value, %arg2[%i] : memref<10xf32>
+ memref.store %value, %arg2[%i] : memref<10xf32>
}
return
}
@@ -61,19 +61,19 @@ func @kernel_nested_selection(%arg3 : memref<10xf32>, %arg4 : memref<10xf32>, %a
scf.if %arg5 {
scf.if %arg6 {
- %value = load %arg3[%i] : memref<10xf32>
- store %value, %arg4[%i] : memref<10xf32>
+ %value = memref.load %arg3[%i] : memref<10xf32>
+ memref.store %value, %arg4[%i] : memref<10xf32>
} else {
- %value = load %arg4[%i] : memref<10xf32>
- store %value, %arg3[%i] : memref<10xf32>
+ %value = memref.load %arg4[%i] : memref<10xf32>
+ memref.store %value, %arg3[%i] : memref<10xf32>
}
} else {
scf.if %arg6 {
- %value = load %arg3[%j] : memref<10xf32>
- store %value, %arg4[%j] : memref<10xf32>
+ %value = memref.load %arg3[%j] : memref<10xf32>
+ memref.store %value, %arg4[%j] : memref<10xf32>
} else {
- %value = load %arg4[%j] : memref<10xf32>
- store %value, %arg3[%j] : memref<10xf32>
+ %value = memref.load %arg4[%j] : memref<10xf32>
+ memref.store %value, %arg3[%j] : memref<10xf32>
}
}
return
@@ -116,8 +116,8 @@ func @simple_if_yield(%arg2 : memref<10xf32>, %arg3 : i1) {
}
%i = constant 0 : index
%j = constant 1 : index
- store %0#0, %arg2[%i] : memref<10xf32>
- store %0#1, %arg2[%j] : memref<10xf32>
+ memref.store %0#0, %arg2[%i] : memref<10xf32>
+ memref.store %0#1, %arg2[%j] : memref<10xf32>
return
}
@@ -149,7 +149,7 @@ func @simple_if_yield_type_change(%arg2 : memref<10xf32>, %arg3 : memref<10xf32>
} else {
scf.yield %arg3 : memref<10xf32>
}
- store %value, %0[%i] : memref<10xf32>
+ memref.store %value, %0[%i] : memref<10xf32>
return
}
diff --git a/mlir/test/Conversion/SPIRVToLLVM/lower-host-to-llvm-calls.mlir b/mlir/test/Conversion/SPIRVToLLVM/lower-host-to-llvm-calls.mlir
index 5c5506151478..bc850d065661 100644
--- a/mlir/test/Conversion/SPIRVToLLVM/lower-host-to-llvm-calls.mlir
+++ b/mlir/test/Conversion/SPIRVToLLVM/lower-host-to-llvm-calls.mlir
@@ -1,5 +1,5 @@
// RUN: mlir-opt --lower-host-to-llvm %s | FileCheck %s
-
+
module attributes {gpu.container_module, spv.target_env = #spv.target_env<#spv.vce<v1.0, [Shader], [SPV_KHR_variable_pointers]>, {max_compute_workgroup_invocations = 128 : i32, max_compute_workgroup_size = dense<[128, 128, 64]> : vector<3xi32>}>} {
// CHECK: llvm.mlir.global linkonce @__spv__foo_bar_arg_0_descriptor_set0_binding0() : !llvm.struct<(array<6 x i32>)>
@@ -8,7 +8,7 @@ module attributes {gpu.container_module, spv.target_env = #spv.target_env<#spv.v
// CHECK: spv.module @__spv__foo
// CHECK: spv.GlobalVariable @bar_arg_0 bind(0, 0) : !spv.ptr<!spv.struct<(!spv.array<6 x i32, stride=4> [0])>, StorageBuffer>
// CHECK: spv.func @__spv__foo_bar
-
+
// CHECK: spv.EntryPoint "GLCompute" @__spv__foo_bar
// CHECK: spv.ExecutionMode @__spv__foo_bar "LocalSize", 1, 1, 1
@@ -38,7 +38,7 @@ module attributes {gpu.container_module, spv.target_env = #spv.target_env<#spv.v
}
func @main() {
- %buffer = alloc() : memref<6xi32>
+ %buffer = memref.alloc() : memref<6xi32>
%one = constant 1 : index
gpu.launch_func @foo::@bar blocks in (%one, %one, %one)
threads in (%one, %one, %one) args(%buffer : memref<6xi32>)
diff --git a/mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir b/mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir
index a4a0f7ece9b4..751f5002703b 100644
--- a/mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir
+++ b/mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir
@@ -31,7 +31,7 @@ func @binary_ops_on_size(%lhs : !shape.size, %rhs : !shape.size) {
// CHECK-SAME: (%[[SHAPE:.*]]: tensor<?xindex>) -> index
func @rank(%shape : tensor<?xindex>) -> index {
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[RESULT:.*]] = dim %[[SHAPE]], %[[C0]]
+ // CHECK: %[[RESULT:.*]] = memref.dim %[[SHAPE]], %[[C0]]
// CHECK: return %[[RESULT]] : index
%rank = shape.rank %shape : tensor<?xindex> -> index
return %rank : index
@@ -60,12 +60,12 @@ func @rank(%shape : !shape.shape) {
// -----
-// Express `get_extent` as `std.dim` when it relies directly on the outcome of a
+// Express `get_extent` as `memref.dim` when it relies directly on the outcome of a
// `shape_of` operation.
// CHECK-LABEL: @get_extent_shape_of
// CHECK-SAME: (%[[ARG:.*]]: tensor<2x3xf32>, %[[IDX:.*]]: index) -> index
func @get_extent_shape_of(%arg : tensor<2x3xf32>, %idx : index) -> index {
- // CHECK: %[[RESULT:.*]] = dim %[[ARG]], %[[IDX]] : tensor<2x3xf32>
+ // CHECK: %[[RESULT:.*]] = memref.dim %[[ARG]], %[[IDX]] : tensor<2x3xf32>
// CHECK: return %[[RESULT]] : index
%shape = shape.shape_of %arg : tensor<2x3xf32> -> tensor<?xindex>
%result = shape.get_extent %shape, %idx : tensor<?xindex>, index -> index
@@ -178,7 +178,7 @@ func @shape_reduce(%shape : tensor<?xindex>) -> index {
// CHECK-NEXT: %[[INIT:.*]] = constant 1 : index
// CHECK-NEXT: %[[C0:.*]] = constant 0 : index
// CHECK-NEXT: %[[C1:.*]] = constant 1 : index
-// CHECK-NEXT: %[[RANK:.*]] = dim %[[SHAPE]], %[[C0]] : tensor<?xindex>
+// CHECK-NEXT: %[[RANK:.*]] = memref.dim %[[SHAPE]], %[[C0]] : tensor<?xindex>
// CHECK-NEXT: %[[RESULT:.*]] = scf.for %[[I:.*]] = %[[C0]] to %[[RANK]] step %[[C1]] iter_args(%[[ACC:.*]] = %[[INIT]]) -> (index)
// CHECK-NEXT: %[[EXTENT:.*]] = tensor.extract %[[SHAPE]][%[[I]]]
// CHECK-NEXT: %[[NEW_ACC:.*]] = muli %[[ACC]], %[[EXTENT]] : index
@@ -206,7 +206,7 @@ func @shape_of_unranked(%arg : tensor<*xf32>) {
// CHECK: %[[RANK:.*]] = rank %[[ARG]] : tensor<*xf32>
// CHECK: %[[SHAPE:.*]] = tensor.generate %[[RANK]] {
// CHECK: ^bb0(%[[I:.*]]: index):
- // CHECK: %[[EXTENT:.*]] = dim %[[ARG]], %[[I]] : tensor<*xf32>
+ // CHECK: %[[EXTENT:.*]] = memref.dim %[[ARG]], %[[I]] : tensor<*xf32>
// CHECK: yield %[[EXTENT]] : index
// CHECK: } : tensor<?xindex>
%shape = shape.shape_of %arg : tensor<*xf32> -> tensor<?xindex>
@@ -258,7 +258,7 @@ func @shape_of_dyn(%arg : tensor<1x5x?xf32>) {
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C5:.*]] = constant 5 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
- // CHECK-DAG: %[[DYN_DIM:.*]] = dim %[[ARG]], %[[C2]] : tensor<1x5x?xf32>
+ // CHECK-DAG: %[[DYN_DIM:.*]] = memref.dim %[[ARG]], %[[C2]] : tensor<1x5x?xf32>
// CHECK-DAG: %[[SHAPE_UNCASTED:.*]] = tensor.from_elements %[[C1]], %[[C5]], %[[DYN_DIM]] : tensor<3xindex>
%shape = shape.shape_of %arg : tensor<1x5x?xf32> -> tensor<?xindex>
return
@@ -270,8 +270,8 @@ func @shape_of_dyn(%arg : tensor<1x5x?xf32>) {
// CHECK-SAME: (%[[A:.*]]: tensor<?xindex>, %[[B:.*]]: tensor<?xindex>) -> i1
func @shape_eq(%a : tensor<?xindex>, %b : tensor<?xindex>) -> i1 {
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[RANK_A:.*]] = dim %[[A]], %[[C0]] : tensor<?xindex>
- // CHECK: %[[RANK_B:.*]] = dim %[[B]], %[[C0]] : tensor<?xindex>
+ // CHECK: %[[RANK_A:.*]] = memref.dim %[[A]], %[[C0]] : tensor<?xindex>
+ // CHECK: %[[RANK_B:.*]] = memref.dim %[[B]], %[[C0]] : tensor<?xindex>
// CHECK: %[[RANK_EQ:.*]] = cmpi eq, %[[RANK_A]], %[[RANK_B]]
// CHECK: %[[SHAPE_EQ:.*]] = scf.if %[[RANK_EQ]] -> (i1) {
// CHECK: %[[C1:.*]] = constant 1 : index
@@ -299,8 +299,8 @@ func @shape_eq(%a : tensor<?xindex>, %b : tensor<?xindex>) -> i1 {
// CHECK-SAME: (%[[A:.*]]: tensor<?xindex>, %[[B:.*]]: tensor<?xindex>, %[[C:.*]]: tensor<?xindex>) -> i1
func @shape_eq(%a : tensor<?xindex>, %b : tensor<?xindex>, %c : tensor<?xindex>) -> i1 {
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[RANK_A:.*]] = dim %[[A]], %[[C0]] : tensor<?xindex>
- // CHECK: %[[RANK_B:.*]] = dim %[[B]], %[[C0]] : tensor<?xindex>
+ // CHECK: %[[RANK_A:.*]] = memref.dim %[[A]], %[[C0]] : tensor<?xindex>
+ // CHECK: %[[RANK_B:.*]] = memref.dim %[[B]], %[[C0]] : tensor<?xindex>
// CHECK: %[[RANK_EQ:.*]] = cmpi eq, %[[RANK_A]], %[[RANK_B]]
// CHECK: %[[SHAPE_EQ:.*]] = scf.if %[[RANK_EQ]] -> (i1) {
// CHECK: %[[C1:.*]] = constant 1 : index
@@ -317,7 +317,7 @@ func @shape_eq(%a : tensor<?xindex>, %b : tensor<?xindex>, %c : tensor<?xindex>)
// CHECK: %[[SHAPE_EQ_INNER:.*]] = constant false
// CHECK: scf.yield %[[SHAPE_EQ_INNER]] : i1
// CHECK: }
- // CHECK: %[[RANK_C:.*]] = dim %[[C]], %[[C0]] : tensor<?xindex>
+ // CHECK: %[[RANK_C:.*]] = memref.dim %[[C]], %[[C0]] : tensor<?xindex>
// CHECK: %[[RANK_EQ:.*]] = cmpi eq, %[[RANK_A]], %[[RANK_C]]
// CHECK: %[[SHAPE_EQ2:.*]] = scf.if %[[RANK_EQ]] -> (i1) {
// CHECK: %[[C1:.*]] = constant 1 : index
@@ -362,9 +362,9 @@ func @try_is_broadcastable (%a : tensor<2xindex>, %b : tensor<3xindex>, %c : ten
// CHECK-SAME: %[[ARG2:.*]]: tensor<2xindex>)
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[RANK0:.*]] = dim %[[ARG0]], %[[C0]] : tensor<2xindex>
-// CHECK: %[[RANK1:.*]] = dim %[[ARG1]], %[[C0]] : tensor<3xindex>
-// CHECK: %[[RANK2:.*]] = dim %[[ARG2]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK0:.*]] = memref.dim %[[ARG0]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK1:.*]] = memref.dim %[[ARG1]], %[[C0]] : tensor<3xindex>
+// CHECK: %[[RANK2:.*]] = memref.dim %[[ARG2]], %[[C0]] : tensor<2xindex>
// CHECK: %[[CMP0:.*]] = cmpi ugt, %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[LARGER_DIM:.*]] = select %[[CMP0]], %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[CMP1:.*]] = cmpi ugt, %[[RANK2]], %[[LARGER_DIM]] : index
@@ -452,9 +452,9 @@ func @broadcast(%a : tensor<2xindex>, %b : tensor<3xindex>, %c : tensor<2xindex>
// CHECK-SAME: %[[ARG2:.*]]: tensor<2xindex>)
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[RANK0:.*]] = dim %[[ARG0]], %[[C0]] : tensor<2xindex>
-// CHECK: %[[RANK1:.*]] = dim %[[ARG1]], %[[C0]] : tensor<3xindex>
-// CHECK: %[[RANK2:.*]] = dim %[[ARG2]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK0:.*]] = memref.dim %[[ARG0]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK1:.*]] = memref.dim %[[ARG1]], %[[C0]] : tensor<3xindex>
+// CHECK: %[[RANK2:.*]] = memref.dim %[[ARG2]], %[[C0]] : tensor<2xindex>
// CHECK: %[[CMP0:.*]] = cmpi ugt, %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[LARGER_DIM:.*]] = select %[[CMP0]], %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[CMP1:.*]] = cmpi ugt, %[[RANK2]], %[[LARGER_DIM]] : index
@@ -544,9 +544,9 @@ func @broadcast_3_shapes_
diff erent_extents(%a : tensor<2xindex>,
// CHECK-SAME: %[[ARG1:.*]]: tensor<3xindex>,
// CHECK-SAME: %[[ARG2:.*]]: tensor<2xindex>) {
// CHECK: %[[C0:.*]] = constant 0 : index
-// CHECK: %[[RANK0:.*]] = dim %[[ARG0]], %[[C0]] : tensor<2xindex>
-// CHECK: %[[RANK1:.*]] = dim %[[ARG1]], %[[C0]] : tensor<3xindex>
-// CHECK: %[[RANK2:.*]] = dim %[[ARG2]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK0:.*]] = memref.dim %[[ARG0]], %[[C0]] : tensor<2xindex>
+// CHECK: %[[RANK1:.*]] = memref.dim %[[ARG1]], %[[C0]] : tensor<3xindex>
+// CHECK: %[[RANK2:.*]] = memref.dim %[[ARG2]], %[[C0]] : tensor<2xindex>
// CHECK: %[[CMP0:.*]] = cmpi ugt, %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[LARGER_DIM:.*]] = select %[[CMP0]], %[[RANK1]], %[[RANK0]] : index
// CHECK: %[[CMP1:.*]] = cmpi ugt, %[[RANK2]], %[[LARGER_DIM]] : index
@@ -600,7 +600,7 @@ func @broadcast_3_shapes_
diff erent_extents(%a : tensor<2xindex>,
// CHECK-SAME: %[[SHAPE:.*]]: tensor<?xindex>, %[[INDEX:.*]]: index
func @split_at(%shape: tensor<?xindex>, %index: index) -> (tensor<?xindex>, tensor<?xindex>) {
// CHECK-NEXT: %[[C0:.*]] = constant 0 : index
- // CHECK-NEXT: %[[RANK:.*]] = dim %[[SHAPE]], %[[C0]] : tensor<?xindex>
+ // CHECK-NEXT: %[[RANK:.*]] = memref.dim %[[SHAPE]], %[[C0]] : tensor<?xindex>
// CHECK-NEXT: %[[POSINDEX:.*]] = addi %[[INDEX]], %[[RANK]] : index
// CHECK-NEXT: %[[ISNEG:.*]] = cmpi slt, %[[INDEX]], %[[C0]] : index
// CHECK-NEXT: %[[SELECT:.*]] = select %[[ISNEG]], %[[POSINDEX]], %[[INDEX]] : index
diff --git a/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir b/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir
index ae2ead8d95f8..6db4106d6e69 100644
--- a/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/calling-convention.mlir
@@ -75,7 +75,7 @@ func @caller() {
// CHECK-LABEL: @callee
// EMIT_C_ATTRIBUTE-LABEL: @callee
func @callee(%arg0: memref<?xf32>, %arg1: index) {
- %0 = load %arg0[%arg1] : memref<?xf32>
+ %0 = memref.load %arg0[%arg1] : memref<?xf32>
return
}
@@ -100,7 +100,7 @@ func @callee(%arg0: memref<?xf32>, %arg1: index) {
// CHECK-LABEL: @other_callee
// EMIT_C_ATTRIBUTE-LABEL: @other_callee
func @other_callee(%arg0: memref<?xf32>, %arg1: index) attributes { llvm.emit_c_interface } {
- %0 = load %arg0[%arg1] : memref<?xf32>
+ %0 = memref.load %arg0[%arg1] : memref<?xf32>
return
}
@@ -151,7 +151,7 @@ func @return_var_memref(%arg0: memref<4x3xf32>) -> memref<*xf32> {
// CHECK: %[[DESC_0:.*]] = llvm.mlir.undef : !llvm.struct<(i64, ptr<i8>)>
// CHECK: %[[DESC_1:.*]] = llvm.insertvalue %{{.*}}, %[[DESC_0]][0]
// CHECK: %[[DESC_2:.*]] = llvm.insertvalue %[[MEMORY]], %[[DESC_1]][1]
- %0 = memref_cast %arg0: memref<4x3xf32> to memref<*xf32>
+ %0 = memref.cast %arg0: memref<4x3xf32> to memref<*xf32>
// CHECK: %[[ONE:.*]] = llvm.mlir.constant(1 : index)
// CHECK: %[[TWO:.*]] = llvm.mlir.constant(2 : index)
@@ -213,7 +213,7 @@ func @return_two_var_memref(%arg0: memref<4x3xf32>) -> (memref<*xf32>, memref<*x
// CHECK: %[[DESC_0:.*]] = llvm.mlir.undef : !llvm.struct<(i64, ptr<i8>)>
// CHECK: %[[DESC_1:.*]] = llvm.insertvalue %{{.*}}, %[[DESC_0]][0]
// CHECK: %[[DESC_2:.*]] = llvm.insertvalue %[[MEMORY]], %[[DESC_1]][1]
- %0 = memref_cast %arg0 : memref<4x3xf32> to memref<*xf32>
+ %0 = memref.cast %arg0 : memref<4x3xf32> to memref<*xf32>
// Only check that we allocate the memory for each operand of the "return"
// separately, even if both operands are the same value. The calling
diff --git a/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir b/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir
index 41959a79d58f..49a5a2f2f871 100644
--- a/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/convert-argattrs.mlir
@@ -5,7 +5,7 @@
// 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 = load %static[%c0, %c0]: memref<10x20xf32>
+ %0 = memref.load %static[%c0, %c0]: memref<10x20xf32>
return
}
diff --git a/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir b/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir
index 9ea40718e260..ac189f467748 100644
--- a/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/convert-dynamic-memref-ops.mlir
@@ -48,7 +48,7 @@ func @mixed_alloc(%arg0: index, %arg1: index) -> memref<?x42x?xf32> {
// CHECK-NEXT: llvm.insertvalue %[[st0]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[4, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[one]], %{{.*}}[4, 2] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
- %0 = alloc(%arg0, %arg1) : memref<?x42x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x42x?xf32>
// CHECK-NEXT: llvm.return %{{.*}} : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
return %0 : memref<?x42x?xf32>
}
@@ -58,7 +58,7 @@ func @mixed_dealloc(%arg0: memref<?x42x?xf32>) {
// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
// CHECK-NEXT: %[[ptri8:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr<f32> to !llvm.ptr<i8>
// CHECK-NEXT: llvm.call @free(%[[ptri8]]) : (!llvm.ptr<i8>) -> ()
- dealloc %arg0 : memref<?x42x?xf32>
+ memref.dealloc %arg0 : memref<?x42x?xf32>
// CHECK-NEXT: llvm.return
return
}
@@ -82,7 +82,7 @@ func @dynamic_alloc(%arg0: index, %arg1: index) -> memref<?x?xf32> {
// CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[3, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[one]], %{{.*}}[4, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
// CHECK-NEXT: llvm.return %{{.*}} : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
return %0 : memref<?x?xf32>
}
@@ -107,7 +107,7 @@ func @dynamic_alloca(%arg0: index, %arg1: index) -> memref<?x?xf32> {
// CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[3, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[N]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK-NEXT: llvm.insertvalue %[[st1]], %{{.*}}[4, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %0 = alloca(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloca(%arg0, %arg1) : memref<?x?xf32>
// Test with explicitly specified alignment. llvm.alloca takes care of the
// alignment. The same pointer is thus used for allocation and aligned
@@ -116,7 +116,7 @@ func @dynamic_alloca(%arg0: index, %arg1: index) -> memref<?x?xf32> {
// CHECK: %[[desc:.*]] = llvm.mlir.undef : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: %[[desc1:.*]] = llvm.insertvalue %[[alloca_aligned]], %[[desc]][0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.insertvalue %[[alloca_aligned]], %[[desc1]][1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- alloca(%arg0, %arg1) {alignment = 32} : memref<?x?xf32>
+ memref.alloca(%arg0, %arg1) {alignment = 32} : memref<?x?xf32>
return %0 : memref<?x?xf32>
}
@@ -125,7 +125,7 @@ func @dynamic_dealloc(%arg0: memref<?x?xf32>) {
// CHECK: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK-NEXT: %[[ptri8:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr<f32> to !llvm.ptr<i8>
// CHECK-NEXT: llvm.call @free(%[[ptri8]]) : (!llvm.ptr<i8>) -> ()
- dealloc %arg0 : memref<?x?xf32>
+ memref.dealloc %arg0 : memref<?x?xf32>
return
}
@@ -142,23 +142,23 @@ func @stdlib_aligned_alloc(%N : index) -> memref<32x18xf32> {
// ALIGNED-ALLOC-NEXT: %[[alignment:.*]] = llvm.mlir.constant(32 : index) : i64
// ALIGNED-ALLOC-NEXT: %[[allocated:.*]] = llvm.call @aligned_alloc(%[[alignment]], %[[bytes]]) : (i64, i64) -> !llvm.ptr<i8>
// ALIGNED-ALLOC-NEXT: llvm.bitcast %[[allocated]] : !llvm.ptr<i8> to !llvm.ptr<f32>
- %0 = alloc() {alignment = 32} : memref<32x18xf32>
+ %0 = memref.alloc() {alignment = 32} : memref<32x18xf32>
// Do another alloc just to test that we have a unique declaration for
// aligned_alloc.
// ALIGNED-ALLOC: llvm.call @aligned_alloc
- %1 = alloc() {alignment = 64} : memref<4096xf32>
+ %1 = memref.alloc() {alignment = 64} : memref<4096xf32>
// Alignment is to element type boundaries (minimum 16 bytes).
// ALIGNED-ALLOC: %[[c32:.*]] = llvm.mlir.constant(32 : index) : i64
// ALIGNED-ALLOC-NEXT: llvm.call @aligned_alloc(%[[c32]]
- %2 = alloc() : memref<4096xvector<8xf32>>
+ %2 = memref.alloc() : memref<4096xvector<8xf32>>
// The minimum alignment is 16 bytes unless explicitly specified.
// ALIGNED-ALLOC: %[[c16:.*]] = llvm.mlir.constant(16 : index) : i64
// ALIGNED-ALLOC-NEXT: llvm.call @aligned_alloc(%[[c16]],
- %3 = alloc() : memref<4096xvector<2xf32>>
+ %3 = memref.alloc() : memref<4096xvector<2xf32>>
// ALIGNED-ALLOC: %[[c8:.*]] = llvm.mlir.constant(8 : index) : i64
// ALIGNED-ALLOC-NEXT: llvm.call @aligned_alloc(%[[c8]],
- %4 = alloc() {alignment = 8} : memref<1024xvector<4xf32>>
+ %4 = memref.alloc() {alignment = 8} : memref<1024xvector<4xf32>>
// Bump the memref allocation size if its size is not a multiple of alignment.
// ALIGNED-ALLOC: %[[c32:.*]] = llvm.mlir.constant(32 : index) : i64
// ALIGNED-ALLOC-NEXT: llvm.mlir.constant(1 : index) : i64
@@ -167,11 +167,11 @@ func @stdlib_aligned_alloc(%N : index) -> memref<32x18xf32> {
// ALIGNED-ALLOC-NEXT: llvm.urem
// ALIGNED-ALLOC-NEXT: %[[SIZE_ALIGNED:.*]] = llvm.sub
// ALIGNED-ALLOC-NEXT: llvm.call @aligned_alloc(%[[c32]], %[[SIZE_ALIGNED]])
- %5 = alloc() {alignment = 32} : memref<100xf32>
+ %5 = memref.alloc() {alignment = 32} : memref<100xf32>
// Bump alignment to the next power of two if it isn't.
// ALIGNED-ALLOC: %[[c128:.*]] = llvm.mlir.constant(128 : index) : i64
// ALIGNED-ALLOC: llvm.call @aligned_alloc(%[[c128]]
- %6 = alloc(%N) : memref<?xvector<18xf32>>
+ %6 = memref.alloc(%N) : memref<?xvector<18xf32>>
return %0 : memref<32x18xf32>
}
@@ -187,7 +187,7 @@ func @mixed_load(%mixed : memref<42x?xf32>, %i : index, %j : index) {
// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// CHECK-NEXT: llvm.load %[[addr]] : !llvm.ptr<f32>
- %0 = load %mixed[%i, %j] : memref<42x?xf32>
+ %0 = memref.load %mixed[%i, %j] : memref<42x?xf32>
return
}
@@ -208,7 +208,7 @@ func @dynamic_load(%dynamic : memref<?x?xf32>, %i : index, %j : index) {
// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// CHECK-NEXT: llvm.load %[[addr]] : !llvm.ptr<f32>
- %0 = load %dynamic[%i, %j] : memref<?x?xf32>
+ %0 = memref.load %dynamic[%i, %j] : memref<?x?xf32>
return
}
@@ -232,17 +232,17 @@ func @prefetch(%A : memref<?x?xf32>, %i : index, %j : index) {
// CHECK-NEXT: [[C3:%.*]] = llvm.mlir.constant(3 : i32) : i32
// CHECK-NEXT: [[C1_1:%.*]] = llvm.mlir.constant(1 : i32) : i32
// CHECK-NEXT: "llvm.intr.prefetch"(%[[addr]], [[C1]], [[C3]], [[C1_1]]) : (!llvm.ptr<f32>, i32, i32, i32) -> ()
- prefetch %A[%i, %j], write, locality<3>, data : memref<?x?xf32>
+ memref.prefetch %A[%i, %j], write, locality<3>, data : memref<?x?xf32>
// CHECK: [[C0:%.*]] = llvm.mlir.constant(0 : i32) : i32
// CHECK: [[C0_1:%.*]] = llvm.mlir.constant(0 : i32) : i32
// CHECK: [[C1_2:%.*]] = llvm.mlir.constant(1 : i32) : i32
// CHECK: "llvm.intr.prefetch"(%{{.*}}, [[C0]], [[C0_1]], [[C1_2]]) : (!llvm.ptr<f32>, i32, i32, i32) -> ()
- prefetch %A[%i, %j], read, locality<0>, data : memref<?x?xf32>
+ memref.prefetch %A[%i, %j], read, locality<0>, data : memref<?x?xf32>
// CHECK: [[C0_2:%.*]] = llvm.mlir.constant(0 : i32) : i32
// CHECK: [[C2:%.*]] = llvm.mlir.constant(2 : i32) : i32
// CHECK: [[C0_3:%.*]] = llvm.mlir.constant(0 : i32) : i32
// CHECK: "llvm.intr.prefetch"(%{{.*}}, [[C0_2]], [[C2]], [[C0_3]]) : (!llvm.ptr<f32>, i32, i32, i32) -> ()
- prefetch %A[%i, %j], read, locality<2>, instr : memref<?x?xf32>
+ memref.prefetch %A[%i, %j], read, locality<2>, instr : memref<?x?xf32>
return
}
@@ -263,7 +263,7 @@ func @dynamic_store(%dynamic : memref<?x?xf32>, %i : index, %j : index, %val : f
// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// CHECK-NEXT: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr<f32>
- store %val, %dynamic[%i, %j] : memref<?x?xf32>
+ memref.store %val, %dynamic[%i, %j] : memref<?x?xf32>
return
}
@@ -284,56 +284,56 @@ func @mixed_store(%mixed : memref<42x?xf32>, %i : index, %j : index, %val : f32)
// CHECK-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// CHECK-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// CHECK-NEXT: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr<f32>
- store %val, %mixed[%i, %j] : memref<42x?xf32>
+ memref.store %val, %mixed[%i, %j] : memref<42x?xf32>
return
}
// CHECK-LABEL: func @memref_cast_static_to_dynamic
func @memref_cast_static_to_dynamic(%static : memref<10x42xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %static : memref<10x42xf32> to memref<?x?xf32>
+ %0 = memref.cast %static : memref<10x42xf32> to memref<?x?xf32>
return
}
// CHECK-LABEL: func @memref_cast_static_to_mixed
func @memref_cast_static_to_mixed(%static : memref<10x42xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %static : memref<10x42xf32> to memref<?x42xf32>
+ %0 = memref.cast %static : memref<10x42xf32> to memref<?x42xf32>
return
}
// CHECK-LABEL: func @memref_cast_dynamic_to_static
func @memref_cast_dynamic_to_static(%dynamic : memref<?x?xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %dynamic : memref<?x?xf32> to memref<10x12xf32>
+ %0 = memref.cast %dynamic : memref<?x?xf32> to memref<10x12xf32>
return
}
// CHECK-LABEL: func @memref_cast_dynamic_to_mixed
func @memref_cast_dynamic_to_mixed(%dynamic : memref<?x?xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %dynamic : memref<?x?xf32> to memref<?x12xf32>
+ %0 = memref.cast %dynamic : memref<?x?xf32> to memref<?x12xf32>
return
}
// CHECK-LABEL: func @memref_cast_mixed_to_dynamic
func @memref_cast_mixed_to_dynamic(%mixed : memref<42x?xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %mixed : memref<42x?xf32> to memref<?x?xf32>
+ %0 = memref.cast %mixed : memref<42x?xf32> to memref<?x?xf32>
return
}
// CHECK-LABEL: func @memref_cast_mixed_to_static
func @memref_cast_mixed_to_static(%mixed : memref<42x?xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %mixed : memref<42x?xf32> to memref<42x1xf32>
+ %0 = memref.cast %mixed : memref<42x?xf32> to memref<42x1xf32>
return
}
// CHECK-LABEL: func @memref_cast_mixed_to_mixed
func @memref_cast_mixed_to_mixed(%mixed : memref<42x?xf32>) {
// CHECK-NOT: llvm.bitcast
- %0 = memref_cast %mixed : memref<42x?xf32> to memref<?x1xf32>
+ %0 = memref.cast %mixed : memref<42x?xf32> to memref<?x1xf32>
return
}
@@ -347,7 +347,7 @@ func @memref_cast_ranked_to_unranked(%arg : memref<42x2x?xf32>) {
// CHECK : llvm.mlir.undef : !llvm.struct<(i64, ptr<i8>)>
// CHECK-DAG: llvm.insertvalue %[[r]], %{{.*}}[0] : !llvm.struct<(i64, ptr<i8>)>
// CHECK-DAG: llvm.insertvalue %[[p2]], %{{.*}}[1] : !llvm.struct<(i64, ptr<i8>)>
- %0 = memref_cast %arg : memref<42x2x?xf32> to memref<*xf32>
+ %0 = memref.cast %arg : memref<42x2x?xf32> to memref<*xf32>
return
}
@@ -355,7 +355,7 @@ func @memref_cast_ranked_to_unranked(%arg : memref<42x2x?xf32>) {
func @memref_cast_unranked_to_ranked(%arg : memref<*xf32>) {
// CHECK: %[[p:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(i64, ptr<i8>)>
// CHECK-NEXT: llvm.bitcast %[[p]] : !llvm.ptr<i8> to !llvm.ptr<struct<(ptr<f32>, ptr<f32>, i64, array<4 x i64>, array<4 x i64>)>>
- %0 = memref_cast %arg : memref<*xf32> to memref<?x?x10x2xf32>
+ %0 = memref.cast %arg : memref<*xf32> to memref<?x?x10x2xf32>
return
}
@@ -363,19 +363,19 @@ func @memref_cast_unranked_to_ranked(%arg : memref<*xf32>) {
func @mixed_memref_dim(%mixed : memref<42x?x?x13x?xf32>) {
// CHECK: llvm.mlir.constant(42 : index) : i64
%c0 = constant 0 : index
- %0 = dim %mixed, %c0 : memref<42x?x?x13x?xf32>
+ %0 = memref.dim %mixed, %c0 : memref<42x?x?x13x?xf32>
// CHECK: llvm.extractvalue %[[ld:.*]][3, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<5 x i64>, array<5 x i64>)>
%c1 = constant 1 : index
- %1 = dim %mixed, %c1 : memref<42x?x?x13x?xf32>
+ %1 = memref.dim %mixed, %c1 : memref<42x?x?x13x?xf32>
// CHECK: llvm.extractvalue %[[ld]][3, 2] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<5 x i64>, array<5 x i64>)>
%c2 = constant 2 : index
- %2 = dim %mixed, %c2 : memref<42x?x?x13x?xf32>
+ %2 = memref.dim %mixed, %c2 : memref<42x?x?x13x?xf32>
// CHECK: llvm.mlir.constant(13 : index) : i64
%c3 = constant 3 : index
- %3 = dim %mixed, %c3 : memref<42x?x?x13x?xf32>
+ %3 = memref.dim %mixed, %c3 : memref<42x?x?x13x?xf32>
// CHECK: llvm.extractvalue %[[ld]][3, 4] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<5 x i64>, array<5 x i64>)>
%c4 = constant 4 : index
- %4 = dim %mixed, %c4 : memref<42x?x?x13x?xf32>
+ %4 = memref.dim %mixed, %c4 : memref<42x?x?x13x?xf32>
return
}
@@ -398,13 +398,13 @@ func @memref_dim_with_dyn_index(%arg : memref<3x?xf32>, %idx : index) -> index {
// CHECK-DAG: %[[RESULT_PTR:.*]] = llvm.getelementptr %[[SIZES_PTR]][%[[C0]], %[[IDX]]] : (!llvm.ptr<array<2 x i64>>, i64, i64) -> !llvm.ptr<i64>
// CHECK-DAG: %[[RESULT:.*]] = llvm.load %[[RESULT_PTR]] : !llvm.ptr<i64>
// CHECK-DAG: llvm.return %[[RESULT]] : i64
- %result = dim %arg, %idx : memref<3x?xf32>
+ %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
+ %output = memref.reinterpret_cast %input to
offset: [0], sizes: [6, 1], strides: [1, 1]
: memref<2x3xf32> to memref<6x1xf32>
return
@@ -433,7 +433,7 @@ func @memref_reinterpret_cast_unranked_to_dynamic_shape(%offset: index,
%stride_0 : index,
%stride_1 : index,
%input : memref<*xf32>) {
- %output = memref_reinterpret_cast %input to
+ %output = memref.reinterpret_cast %input to
offset: [%offset], sizes: [%size_0, %size_1],
strides: [%stride_0, %stride_1]
: memref<*xf32> to memref<?x?xf32, offset: ?, strides: [?, ?]>
@@ -462,7 +462,7 @@ func @memref_reinterpret_cast_unranked_to_dynamic_shape(%offset: index,
// CHECK-LABEL: @memref_reshape
func @memref_reshape(%input : memref<2x3xf32>, %shape : memref<?xindex>) {
- %output = memref_reshape %input(%shape)
+ %output = memref.reshape %input(%shape)
: (memref<2x3xf32>, memref<?xindex>) -> memref<*xf32>
return
}
diff --git a/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir b/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir
index f13da7ae48c5..e09707c76697 100644
--- a/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/convert-static-memref-ops.mlir
@@ -93,7 +93,7 @@ func @zero_d_alloc() -> memref<f32> {
// BAREPTR-NEXT: llvm.insertvalue %[[ptr]], %{{.*}}[1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
// BAREPTR-NEXT: %[[c0:.*]] = llvm.mlir.constant(0 : index) : i64
// BAREPTR-NEXT: llvm.insertvalue %[[c0]], %{{.*}}[2] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
- %0 = alloc() : memref<f32>
+ %0 = memref.alloc() : memref<f32>
return %0 : memref<f32>
}
@@ -109,7 +109,7 @@ func @zero_d_dealloc(%arg0: memref<f32>) {
// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
// BAREPTR-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr<f32> to !llvm.ptr<i8>
// BAREPTR-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr<i8>) -> ()
- dealloc %arg0 : memref<f32>
+ memref.dealloc %arg0 : memref<f32>
return
}
@@ -161,7 +161,7 @@ func @aligned_1d_alloc() -> memref<42xf32> {
// BAREPTR-NEXT: llvm.insertvalue %[[alignedBitCast]], %{{.*}}[1] : !llvm.struct<(ptr<f32>, ptr<f32>, 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<f32>, ptr<f32>, i64, array<1 x i64>, array<1 x i64>)>
- %0 = alloc() {alignment = 8} : memref<42xf32>
+ %0 = memref.alloc() {alignment = 8} : memref<42xf32>
return %0 : memref<42xf32>
}
@@ -183,7 +183,7 @@ func @static_alloc() -> memref<32x18xf32> {
// BAREPTR-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr<f32> to i64
// BAREPTR-NEXT: %[[allocated:.*]] = llvm.call @malloc(%[[size_bytes]]) : (i64) -> !llvm.ptr<i8>
// BAREPTR-NEXT: llvm.bitcast %[[allocated]] : !llvm.ptr<i8> to !llvm.ptr<f32>
- %0 = alloc() : memref<32x18xf32>
+ %0 = memref.alloc() : memref<32x18xf32>
return %0 : memref<32x18xf32>
}
@@ -199,7 +199,7 @@ func @static_alloca() -> memref<32x18xf32> {
// CHECK-NEXT: %[[gep:.*]] = llvm.getelementptr %[[null]][%[[num_elems]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// CHECK-NEXT: %[[size_bytes:.*]] = llvm.ptrtoint %[[gep]] : !llvm.ptr<f32> to i64
// CHECK-NEXT: %[[allocated:.*]] = llvm.alloca %[[size_bytes]] x f32 : (i64) -> !llvm.ptr<f32>
- %0 = alloca() : memref<32x18xf32>
+ %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
@@ -208,7 +208,7 @@ func @static_alloca() -> memref<32x18xf32> {
// CHECK: %[[desc:.*]] = llvm.mlir.undef : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: %[[desc1:.*]] = llvm.insertvalue %[[alloca_aligned]], %[[desc]][0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.insertvalue %[[alloca_aligned]], %[[desc1]][1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- alloca() {alignment = 32} : memref<32x18xf32>
+ memref.alloca() {alignment = 32} : memref<32x18xf32>
return %0 : memref<32x18xf32>
}
@@ -224,7 +224,7 @@ func @static_dealloc(%static: memref<10x8xf32>) {
// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// BAREPTR-NEXT: %[[bc:.*]] = llvm.bitcast %[[ptr]] : !llvm.ptr<f32> to !llvm.ptr<i8>
// BAREPTR-NEXT: llvm.call @free(%[[bc]]) : (!llvm.ptr<i8>) -> ()
- dealloc %static : memref<10x8xf32>
+ memref.dealloc %static : memref<10x8xf32>
return
}
@@ -238,7 +238,7 @@ func @zero_d_load(%arg0: memref<f32>) -> f32 {
// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
// BAREPTR-NEXT: llvm.load %[[ptr:.*]] : !llvm.ptr<f32>
- %0 = load %arg0[] : memref<f32>
+ %0 = memref.load %arg0[] : memref<f32>
return %0 : f32
}
@@ -265,7 +265,7 @@ func @static_load(%static : memref<10x42xf32>, %i : index, %j : index) {
// BAREPTR-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// BAREPTR-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// BAREPTR-NEXT: llvm.load %[[addr]] : !llvm.ptr<f32>
- %0 = load %static[%i, %j] : memref<10x42xf32>
+ %0 = memref.load %static[%i, %j] : memref<10x42xf32>
return
}
@@ -280,7 +280,7 @@ func @zero_d_store(%arg0: memref<f32>, %arg1: f32) {
// BAREPTR: %[[ptr:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
// BAREPTR-NEXT: llvm.store %[[val]], %[[ptr]] : !llvm.ptr<f32>
- store %arg1, %arg0[] : memref<f32>
+ memref.store %arg1, %arg0[] : memref<f32>
return
}
@@ -314,7 +314,7 @@ func @static_store(%static : memref<10x42xf32>, %i : index, %j : index, %val : f
// BAREPTR-NEXT: %[[off1:.*]] = llvm.add %[[offI]], %[[J]] : i64
// BAREPTR-NEXT: %[[addr:.*]] = llvm.getelementptr %[[ptr]][%[[off1]]] : (!llvm.ptr<f32>, i64) -> !llvm.ptr<f32>
// BAREPTR-NEXT: llvm.store %{{.*}}, %[[addr]] : !llvm.ptr<f32>
- store %val, %static[%i, %j] : memref<10x42xf32>
+ memref.store %val, %static[%i, %j] : memref<10x42xf32>
return
}
@@ -327,23 +327,23 @@ func @static_memref_dim(%static : memref<42x32x15x13x27xf32>) {
// BAREPTR: llvm.insertvalue %{{.*}}, %{{.*}}[4, 4] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<5 x i64>, array<5 x i64>)>
// BAREPTR: llvm.mlir.constant(42 : index) : i64
%c0 = constant 0 : index
- %0 = dim %static, %c0 : memref<42x32x15x13x27xf32>
+ %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 = dim %static, %c1 : memref<42x32x15x13x27xf32>
+ %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 = dim %static, %c2 : memref<42x32x15x13x27xf32>
+ %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 = dim %static, %c3 : memref<42x32x15x13x27xf32>
+ %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 = dim %static, %c4 : memref<42x32x15x13x27xf32>
+ %4 = memref.dim %static, %c4 : memref<42x32x15x13x27xf32>
return
}
diff --git a/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir b/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir
index 2b38def9e6f2..2e8a599ad162 100644
--- a/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir
@@ -786,7 +786,7 @@ func @splat(%a: vector<4xf32>, %b: f32) -> vector<4xf32> {
func @view(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK: llvm.mlir.constant(2048 : index) : i64
// CHECK: llvm.mlir.undef : !llvm.struct<(ptr<i8>, ptr<i8>, i64, array<1 x i64>, array<1 x i64>)>
- %0 = alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
// Test two dynamic sizes.
// CHECK: llvm.mlir.undef : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
@@ -802,7 +802,7 @@ func @view(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK: llvm.insertvalue %[[ARG0]], %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.mul %{{.*}}, %[[ARG1]]
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %1 = view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
+ %1 = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
// Test one dynamic size.
// CHECK: llvm.mlir.undef : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
@@ -819,7 +819,7 @@ func @view(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.mul %{{.*}}, %[[ARG1]]
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %3 = view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
+ %3 = memref.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
// Test static sizes.
// CHECK: llvm.mlir.undef : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
@@ -837,12 +837,12 @@ func @view(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.mlir.constant(4 : index) : i64
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %5 = view %0[%arg2][] : memref<2048xi8> to memref<64x4xf32>
+ %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<i8, 4>, ptr<i8, 4>, i64, array<1 x i64>, array<1 x i64>)>
- %6 = alloc() : memref<2048xi8, 4>
+ %6 = memref.alloc() : memref<2048xi8, 4>
// CHECK: llvm.mlir.undef : !llvm.struct<(ptr<f32, 4>, ptr<f32, 4>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: %[[BASE_PTR_4:.*]] = llvm.extractvalue %{{.*}}[1] : !llvm.struct<(ptr<i8, 4>, ptr<i8, 4>, i64, array<1 x i64>, array<1 x i64>)>
@@ -859,7 +859,7 @@ func @view(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[3, 0] : !llvm.struct<(ptr<f32, 4>, ptr<f32, 4>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.mlir.constant(4 : index) : i64
// CHECK: llvm.insertvalue %{{.*}}, %{{.*}}[4, 0] : !llvm.struct<(ptr<f32, 4>, ptr<f32, 4>, i64, array<2 x i64>, array<2 x i64>)>
- %7 = view %6[%arg2][] : memref<2048xi8, 4> to memref<64x4xf32, 4>
+ %7 = memref.view %6[%arg2][] : memref<2048xi8, 4> to memref<64x4xf32, 4>
return
}
@@ -919,7 +919,7 @@ func @subview(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg0 : index,
// CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
+ %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
memref<64x4xf32, offset: 0, strides: [4, 1]>
to memref<?x?xf32, offset: ?, strides: [?, ?]>
return
@@ -980,7 +980,7 @@ func @subview_non_zero_addrspace(%0 : memref<64x4xf32, offset: 0, strides: [4, 1
// CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG0]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32, 3>, ptr<f32, 3>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
+ %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
memref<64x4xf32, offset: 0, strides: [4, 1], 3>
to memref<?x?xf32, offset: ?, strides: [?, ?], 3>
return
@@ -1055,7 +1055,7 @@ func @subview_const_size(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %arg
// CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG7]], %[[STRIDE0]] : i32
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST4]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
// CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[%arg0, %arg1][4, 2][%arg0, %arg1] :
+ %1 = memref.subview %0[%arg0, %arg1][4, 2][%arg0, %arg1] :
memref<64x4xf32, offset: 0, strides: [4, 1]>
to memref<4x2xf32, offset: ?, strides: [?, ?]>
return
@@ -1126,7 +1126,7 @@ func @subview_const_stride(%0 : memref<64x4xf32, offset: 0, strides: [4, 1]>, %a
// CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64)
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[ARG7]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
// CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[%arg0, %arg1][%arg0, %arg1][1, 2] :
+ %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][1, 2] :
memref<64x4xf32, offset: 0, strides: [4, 1]>
to memref<?x?xf32, offset: ?, strides: [4, 2]>
return
@@ -1156,7 +1156,7 @@ func @subview_const_stride_and_offset(%0 : memref<64x4xf32, offset: 0, strides:
// CHECK32: %[[CST4:.*]] = llvm.mlir.constant(4 : i64)
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
// CHECK32: llvm.insertvalue %[[CST4]], %[[DESC5]][4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[0, 8][62, 3][1, 1] :
+ %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
@@ -1201,7 +1201,7 @@ func @subview_mixed_static_dynamic(%0 : memref<64x4xf32, offset: 0, strides: [4,
// CHECK32: %[[DESCSTRIDE0:.*]] = llvm.mul %[[ARG0]], %[[STRIDE0]] : i32
// CHECK32: %[[DESC5:.*]] = llvm.insertvalue %[[CST62]], %[[DESC4]][3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
// CHECK32: llvm.insertvalue %[[DESCSTRIDE0]], %[[DESC5]][4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i32, array<2 x i32>, array<2 x i32>)>
- %1 = subview %0[%arg1, 8][62, %arg2][%arg0, 1] :
+ %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
@@ -1227,7 +1227,7 @@ func @subview_leading_operands(%0 : memref<5x3xf32>, %1: memref<5x?xf32>) {
// CHECK: %[[C3_3:.*]] = llvm.mlir.constant(3 : i64) : i64
// CHECK: llvm.insertvalue %[[C3_2]], %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.insertvalue %[[C3_3]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %2 = subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
+ %2 = memref.subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
return
}
@@ -1259,7 +1259,7 @@ func @subview_leading_operands_dynamic(%0 : memref<5x?xf32>) {
// CHECK: %[[MUL:.*]] = llvm.mul %[[C1_2]], %[[ST0]] : i64
// CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.insertvalue %[[MUL]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %1 = subview %0[2][3][1]: memref<5x?xf32> to memref<3x?xf32, offset: ?, strides: [?, 1]>
+ %1 = memref.subview %0[2][3][1]: memref<5x?xf32> to memref<3x?xf32, offset: ?, strides: [?, 1]>
return
}
@@ -1281,7 +1281,7 @@ func @subview_rank_reducing_leading_operands(%0 : memref<5x3xf32>) {
// CHECK: %[[C1:.*]] = llvm.mlir.constant(1 : i64) : i64
// CHECK: llvm.insertvalue %[[C3]], %{{.*}}[3, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<1 x i64>, array<1 x i64>)>
// CHECK: llvm.insertvalue %[[C1]], %{{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<1 x i64>, array<1 x i64>)>
- %1 = subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
+ %1 = memref.subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
return
}
@@ -1346,7 +1346,7 @@ func @assume_alignment(%0 : memref<4x4xf16>) {
// 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) -> ()
- assume_alignment %0, 16 : memref<4x4xf16>
+ memref.assume_alignment %0, 16 : memref<4x4xf16>
return
}
@@ -1393,7 +1393,7 @@ func @rank_of_ranked(%ranked: memref<?xi32>) {
// CHECK32-LABEL: func @dim_of_unranked
func @dim_of_unranked(%unranked: memref<*xi32>) -> index {
%c0 = constant 0 : index
- %dim = dim %unranked, %c0 : memref<*xi32>
+ %dim = memref.dim %unranked, %c0 : memref<*xi32>
return %dim : index
}
// CHECK-NEXT: llvm.mlir.undef : !llvm.struct<(i64, ptr<i8>)>
diff --git a/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir b/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir
index f4eba23c9d38..fcb5b1c8a5a2 100644
--- a/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir
+++ b/mlir/test/Conversion/StandardToLLVM/standard-to-llvm.mlir
@@ -3,10 +3,10 @@
// CHECK-LABEL: func @address_space(
// CHECK-SAME: !llvm.ptr<f32, 7>
func @address_space(%arg0 : memref<32xf32, affine_map<(d0) -> (d0)>, 7>) {
- %0 = alloc() : memref<32xf32, affine_map<(d0) -> (d0)>, 5>
+ %0 = memref.alloc() : memref<32xf32, affine_map<(d0) -> (d0)>, 5>
%1 = constant 7 : index
// CHECK: llvm.load %{{.*}} : !llvm.ptr<f32, 5>
- %2 = load %0[%1] : memref<32xf32, affine_map<(d0) -> (d0)>, 5>
+ %2 = memref.load %0[%1] : memref<32xf32, affine_map<(d0) -> (d0)>, 5>
std.return
}
@@ -122,25 +122,25 @@ func @assert_test_function(%arg : i1) {
// CHECK: llvm.extractvalue {{.*}}[3, 2] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
// CHECK: llvm.insertvalue {{.*}}[3, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<3 x i64>, array<3 x i64>)>
func @transpose(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>) {
- %0 = transpose %arg0 (i, j, k) -> (k, i, j) : memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]> to memref<?x?x?xf32, affine_map<(d0, d1, d2)[s0, s1, s2] -> (d2 * s1 + s0 + d0 * s2 + d1)>>
+ %0 = memref.transpose %arg0 (i, j, k) -> (k, i, j) : memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]> to memref<?x?x?xf32, affine_map<(d0, d1, d2)[s0, s1, s2] -> (d2 * s1 + s0 + d0 * s2 + d1)>>
return
}
// -----
// CHECK: llvm.mlir.global external @gv0() : !llvm.array<2 x f32>
-global_memref @gv0 : memref<2xf32> = uninitialized
+memref.global @gv0 : memref<2xf32> = uninitialized
// CHECK: llvm.mlir.global private @gv1() : !llvm.array<2 x f32>
-global_memref "private" @gv1 : memref<2xf32>
+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>>
-global_memref @gv2 : memref<2x3xf32> = dense<[[0.0, 1.0, 2.0], [3.0, 4.0, 5.0]]>
+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 = get_global_memref @gv0 : memref<2xf32>
+ %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<array<2 x f32>>
@@ -179,13 +179,13 @@ func @get_gv2_memref() {
// CHECK: llvm.insertvalue %[[DIM1]], {{.*}}[4, 0] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
// CHECK: llvm.insertvalue %[[STRIDE1]], {{.*}}[4, 1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64, array<2 x i64>, array<2 x i64>)>
- %0 = get_global_memref @gv2 : memref<2x3xf32>
+ %0 = memref.get_global @gv2 : memref<2x3xf32>
return
}
// Test scalar memref.
// CHECK: llvm.mlir.global external @gv3(1.000000e+00 : f32) : f32
-global_memref @gv3 : memref<f32> = dense<1.0>
+memref.global @gv3 : memref<f32> = dense<1.0>
// CHECK-LABEL: func @get_gv3_memref
func @get_gv3_memref() {
@@ -199,7 +199,7 @@ func @get_gv3_memref() {
// CHECK: llvm.insertvalue %[[GEP]], {{.*}}[1] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
// CHECK: %[[OFFSET:.*]] = llvm.mlir.constant(0 : index) : i64
// CHECK: llvm.insertvalue %[[OFFSET]], {{.*}}[2] : !llvm.struct<(ptr<f32>, ptr<f32>, i64)>
- %0 = get_global_memref @gv3 : memref<f32>
+ %0 = memref.get_global @gv3 : memref<f32>
return
}
diff --git a/mlir/test/Conversion/StandardToSPIRV/alloc.mlir b/mlir/test/Conversion/StandardToSPIRV/alloc.mlir
index fcfdbc345fa6..2d9dcf472fcf 100644
--- a/mlir/test/Conversion/StandardToSPIRV/alloc.mlir
+++ b/mlir/test/Conversion/StandardToSPIRV/alloc.mlir
@@ -1,7 +1,7 @@
// RUN: mlir-opt -allow-unregistered-dialect -split-input-file -convert-std-to-spirv -canonicalize -verify-diagnostics %s -o - | FileCheck %s
//===----------------------------------------------------------------------===//
-// std allocation/deallocation ops
+// memref allocation/deallocation ops
//===----------------------------------------------------------------------===//
module attributes {
@@ -10,22 +10,22 @@ module attributes {
}
{
func @alloc_dealloc_workgroup_mem(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<4x5xf32, 3>
- %1 = load %0[%arg0, %arg1] : memref<4x5xf32, 3>
- store %1, %0[%arg0, %arg1] : memref<4x5xf32, 3>
- dealloc %0 : memref<4x5xf32, 3>
+ %0 = memref.alloc() : memref<4x5xf32, 3>
+ %1 = memref.load %0[%arg0, %arg1] : memref<4x5xf32, 3>
+ memref.store %1, %0[%arg0, %arg1] : memref<4x5xf32, 3>
+ memref.dealloc %0 : memref<4x5xf32, 3>
return
}
}
// CHECK: spv.GlobalVariable @[[VAR:.+]] : !spv.ptr<!spv.struct<(!spv.array<20 x f32, stride=4>)>, Workgroup>
// CHECK: func @alloc_dealloc_workgroup_mem
-// CHECK-NOT: alloc
+// CHECK-NOT: memref.alloc
// CHECK: %[[PTR:.+]] = spv.mlir.addressof @[[VAR]]
// CHECK: %[[LOADPTR:.+]] = spv.AccessChain %[[PTR]]
// CHECK: %[[VAL:.+]] = spv.Load "Workgroup" %[[LOADPTR]] : f32
// CHECK: %[[STOREPTR:.+]] = spv.AccessChain %[[PTR]]
// CHECK: spv.Store "Workgroup" %[[STOREPTR]], %[[VAL]] : f32
-// CHECK-NOT: dealloc
+// CHECK-NOT: memref.dealloc
// CHECK: spv.Return
// -----
@@ -36,10 +36,10 @@ module attributes {
}
{
func @alloc_dealloc_workgroup_mem(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<4x5xi16, 3>
- %1 = load %0[%arg0, %arg1] : memref<4x5xi16, 3>
- store %1, %0[%arg0, %arg1] : memref<4x5xi16, 3>
- dealloc %0 : memref<4x5xi16, 3>
+ %0 = memref.alloc() : memref<4x5xi16, 3>
+ %1 = memref.load %0[%arg0, %arg1] : memref<4x5xi16, 3>
+ memref.store %1, %0[%arg0, %arg1] : memref<4x5xi16, 3>
+ memref.dealloc %0 : memref<4x5xi16, 3>
return
}
}
@@ -65,8 +65,8 @@ module attributes {
}
{
func @two_allocs() {
- %0 = alloc() : memref<4x5xf32, 3>
- %1 = alloc() : memref<2x3xi32, 3>
+ %0 = memref.alloc() : memref<4x5xf32, 3>
+ %1 = memref.alloc() : memref<2x3xi32, 3>
return
}
}
@@ -86,8 +86,8 @@ module attributes {
}
{
func @two_allocs_vector() {
- %0 = alloc() : memref<4xvector<4xf32>, 3>
- %1 = alloc() : memref<2xvector<2xi32>, 3>
+ %0 = memref.alloc() : memref<4xvector<4xf32>, 3>
+ %1 = memref.alloc() : memref<2xvector<2xi32>, 3>
return
}
}
@@ -109,8 +109,8 @@ module attributes {
{
func @alloc_dealloc_dynamic_workgroup_mem(%arg0 : index) {
// expected-error @+2 {{unhandled allocation type}}
- // expected-error @+1 {{'std.alloc' op operand #0 must be index}}
- %0 = alloc(%arg0) : memref<4x?xf32, 3>
+ // expected-error @+1 {{'memref.alloc' op operand #0 must be index}}
+ %0 = memref.alloc(%arg0) : memref<4x?xf32, 3>
return
}
}
@@ -124,7 +124,7 @@ module attributes {
{
func @alloc_dealloc_mem() {
// expected-error @+1 {{unhandled allocation type}}
- %0 = alloc() : memref<4x5xf32>
+ %0 = memref.alloc() : memref<4x5xf32>
return
}
}
@@ -139,8 +139,8 @@ module attributes {
{
func @alloc_dealloc_dynamic_workgroup_mem(%arg0 : memref<4x?xf32, 3>) {
// expected-error @+2 {{unhandled deallocation type}}
- // expected-error @+1 {{'std.dealloc' op operand #0 must be memref of any type values}}
- dealloc %arg0 : memref<4x?xf32, 3>
+ // expected-error @+1 {{'memref.dealloc' op operand #0 must be memref of any type values}}
+ memref.dealloc %arg0 : memref<4x?xf32, 3>
return
}
}
@@ -155,7 +155,7 @@ module attributes {
func @alloc_dealloc_mem(%arg0 : memref<4x5xf32>) {
// expected-error @+2 {{unhandled deallocation type}}
// expected-error @+1 {{op operand #0 must be memref of any type values}}
- dealloc %arg0 : memref<4x5xf32>
+ memref.dealloc %arg0 : memref<4x5xf32>
return
}
}
diff --git a/mlir/test/Conversion/StandardToSPIRV/legalization.mlir b/mlir/test/Conversion/StandardToSPIRV/legalization.mlir
index c5c59613b56e..98b5d930eee1 100644
--- a/mlir/test/Conversion/StandardToSPIRV/legalization.mlir
+++ b/mlir/test/Conversion/StandardToSPIRV/legalization.mlir
@@ -3,70 +3,70 @@
// CHECK-LABEL: @fold_static_stride_subview_with_load
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index
func @fold_static_stride_subview_with_load(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index) -> f32 {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C3:%.*]] = constant 3 : index
// CHECK: [[STRIDE1:%.*]] = muli [[ARG3]], [[C2]] : index
// CHECK: [[INDEX1:%.*]] = addi [[ARG1]], [[STRIDE1]] : index
// CHECK: [[STRIDE2:%.*]] = muli [[ARG4]], [[C3]] : index
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
- // CHECK: load [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
- %0 = subview %arg0[%arg1, %arg2][4, 4][2, 3] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
- %1 = load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [64, 3]>
+ // CHECK: memref.load [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][2, 3] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
+ %1 = memref.load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [64, 3]>
return %1 : f32
}
// CHECK-LABEL: @fold_dynamic_stride_subview_with_load
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index, [[ARG5:%.*]]: index, [[ARG6:%.*]]: index
func @fold_dynamic_stride_subview_with_load(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index, %arg5 : index, %arg6 : index) -> f32 {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[STRIDE1:%.*]] = muli [[ARG3]], [[ARG5]] : index
// CHECK: [[INDEX1:%.*]] = addi [[ARG1]], [[STRIDE1]] : index
// CHECK: [[STRIDE2:%.*]] = muli [[ARG4]], [[ARG6]] : index
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
- // CHECK: load [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
- %0 = subview %arg0[%arg1, %arg2][4, 4][%arg5, %arg6] :
+ // CHECK: memref.load [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][%arg5, %arg6] :
memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [?, ?]>
- %1 = load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [?, ?]>
+ %1 = memref.load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [?, ?]>
return %1 : f32
}
// CHECK-LABEL: @fold_static_stride_subview_with_store
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index, [[ARG5:%.*]]: f32
func @fold_static_stride_subview_with_store(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index, %arg5 : f32) {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C3:%.*]] = constant 3 : index
// CHECK: [[STRIDE1:%.*]] = muli [[ARG3]], [[C2]] : index
// CHECK: [[INDEX1:%.*]] = addi [[ARG1]], [[STRIDE1]] : index
// CHECK: [[STRIDE2:%.*]] = muli [[ARG4]], [[C3]] : index
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
- // CHECK: store [[ARG5]], [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
- %0 = subview %arg0[%arg1, %arg2][4, 4][2, 3] :
+ // CHECK: memref.store [[ARG5]], [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][2, 3] :
memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
- store %arg5, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [64, 3]>
+ memref.store %arg5, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [64, 3]>
return
}
// CHECK-LABEL: @fold_dynamic_stride_subview_with_store
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index, [[ARG5:%.*]]: index, [[ARG6:%.*]]: index, [[ARG7:%.*]]: f32
func @fold_dynamic_stride_subview_with_store(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index, %arg5 : index, %arg6 : index, %arg7 : f32) {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[STRIDE1:%.*]] = muli [[ARG3]], [[ARG5]] : index
// CHECK: [[INDEX1:%.*]] = addi [[ARG1]], [[STRIDE1]] : index
// CHECK: [[STRIDE2:%.*]] = muli [[ARG4]], [[ARG6]] : index
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
- // CHECK: store [[ARG7]], [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
- %0 = subview %arg0[%arg1, %arg2][4, 4][%arg5, %arg6] :
+ // CHECK: memref.store [[ARG7]], [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][%arg5, %arg6] :
memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [?, ?]>
- store %arg7, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [?, ?]>
+ memref.store %arg7, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [?, ?]>
return
}
// CHECK-LABEL: @fold_static_stride_subview_with_transfer_read
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index
func @fold_static_stride_subview_with_transfer_read(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index) -> vector<4xf32> {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[F1:%.*]] = constant 1.000000e+00 : f32
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C3:%.*]] = constant 3 : index
@@ -76,7 +76,7 @@ func @fold_static_stride_subview_with_transfer_read(%arg0 : memref<12x32xf32>, %
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
// CHECK: vector.transfer_read [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}}, [[F1]] {masked = [false]}
%f1 = constant 1.0 : f32
- %0 = subview %arg0[%arg1, %arg2][4, 4][2, 3] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][2, 3] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
%1 = vector.transfer_read %0[%arg3, %arg4], %f1 {masked = [false]} : memref<4x4xf32, offset:?, strides: [64, 3]>, vector<4xf32>
return %1 : vector<4xf32>
}
@@ -84,7 +84,7 @@ func @fold_static_stride_subview_with_transfer_read(%arg0 : memref<12x32xf32>, %
// CHECK-LABEL: @fold_static_stride_subview_with_transfer_write
// CHECK-SAME: [[ARG0:%.*]]: memref<12x32xf32>, [[ARG1:%.*]]: index, [[ARG2:%.*]]: index, [[ARG3:%.*]]: index, [[ARG4:%.*]]: index, [[ARG5:%.*]]: vector<4xf32>
func @fold_static_stride_subview_with_transfer_write(%arg0 : memref<12x32xf32>, %arg1 : index, %arg2 : index, %arg3 : index, %arg4 : index, %arg5 : vector<4xf32>) {
- // CHECK-NOT: subview
+ // CHECK-NOT: memref.subview
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C3:%.*]] = constant 3 : index
// CHECK: [[STRIDE1:%.*]] = muli [[ARG3]], [[C2]] : index
@@ -92,7 +92,7 @@ func @fold_static_stride_subview_with_transfer_write(%arg0 : memref<12x32xf32>,
// CHECK: [[STRIDE2:%.*]] = muli [[ARG4]], [[C3]] : index
// CHECK: [[INDEX2:%.*]] = addi [[ARG2]], [[STRIDE2]] : index
// CHECK: vector.transfer_write [[ARG5]], [[ARG0]]{{\[}}[[INDEX1]], [[INDEX2]]{{\]}} {masked = [false]}
- %0 = subview %arg0[%arg1, %arg2][4, 4][2, 3] :
+ %0 = memref.subview %arg0[%arg1, %arg2][4, 4][2, 3] :
memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [64, 3]>
vector.transfer_write %arg5, %0[%arg3, %arg4] {masked = [false]} : vector<4xf32>, memref<4x4xf32, offset:?, strides: [64, 3]>
return
diff --git a/mlir/test/Conversion/StandardToSPIRV/std-ops-to-spirv.mlir b/mlir/test/Conversion/StandardToSPIRV/std-ops-to-spirv.mlir
index 7a11228f26d7..d074969febff 100644
--- a/mlir/test/Conversion/StandardToSPIRV/std-ops-to-spirv.mlir
+++ b/mlir/test/Conversion/StandardToSPIRV/std-ops-to-spirv.mlir
@@ -864,7 +864,7 @@ func @select(%arg0 : i32, %arg1 : i32) {
}
//===----------------------------------------------------------------------===//
-// std load/store ops
+// memref load/store ops
//===----------------------------------------------------------------------===//
// CHECK-LABEL: @load_store_zero_rank_float
@@ -876,13 +876,13 @@ func @load_store_zero_rank_float(%arg0: memref<f32>, %arg1: memref<f32>) {
// CHECK-SAME: [[ZERO1]], [[ZERO1]]
// CHECK-SAME: ] :
// CHECK: spv.Load "StorageBuffer" %{{.*}} : f32
- %0 = load %arg0[] : memref<f32>
+ %0 = memref.load %arg0[] : memref<f32>
// CHECK: [[ZERO2:%.*]] = spv.Constant 0 : i32
// CHECK: spv.AccessChain [[ARG1]][
// CHECK-SAME: [[ZERO2]], [[ZERO2]]
// CHECK-SAME: ] :
// CHECK: spv.Store "StorageBuffer" %{{.*}} : f32
- store %0, %arg1[] : memref<f32>
+ memref.store %0, %arg1[] : memref<f32>
return
}
@@ -895,13 +895,13 @@ func @load_store_zero_rank_int(%arg0: memref<i32>, %arg1: memref<i32>) {
// CHECK-SAME: [[ZERO1]], [[ZERO1]]
// CHECK-SAME: ] :
// CHECK: spv.Load "StorageBuffer" %{{.*}} : i32
- %0 = load %arg0[] : memref<i32>
+ %0 = memref.load %arg0[] : memref<i32>
// CHECK: [[ZERO2:%.*]] = spv.Constant 0 : i32
// CHECK: spv.AccessChain [[ARG1]][
// CHECK-SAME: [[ZERO2]], [[ZERO2]]
// CHECK-SAME: ] :
// CHECK: spv.Store "StorageBuffer" %{{.*}} : i32
- store %0, %arg1[] : memref<i32>
+ memref.store %0, %arg1[] : memref<i32>
return
}
@@ -934,7 +934,7 @@ func @load_i8(%arg0: memref<i8>) {
// CHECK: %[[T2:.+]] = spv.Constant 24 : i32
// CHECK: %[[T3:.+]] = spv.ShiftLeftLogical %[[T1]], %[[T2]] : i32, i32
// CHECK: spv.ShiftRightArithmetic %[[T3]], %[[T2]] : i32, i32
- %0 = load %arg0[] : memref<i8>
+ %0 = memref.load %arg0[] : memref<i8>
return
}
@@ -960,7 +960,7 @@ func @load_i16(%arg0: memref<10xi16>, %index : index) {
// CHECK: %[[T2:.+]] = spv.Constant 16 : i32
// CHECK: %[[T3:.+]] = spv.ShiftLeftLogical %[[T1]], %[[T2]] : i32, i32
// CHECK: spv.ShiftRightArithmetic %[[T3]], %[[T2]] : i32, i32
- %0 = load %arg0[%index] : memref<10xi16>
+ %0 = memref.load %arg0[%index] : memref<10xi16>
return
}
@@ -969,7 +969,7 @@ func @load_i32(%arg0: memref<i32>) {
// CHECK-NOT: spv.SDiv
// CHECK: spv.Load
// CHECK-NOT: spv.ShiftRightArithmetic
- %0 = load %arg0[] : memref<i32>
+ %0 = memref.load %arg0[] : memref<i32>
return
}
@@ -978,7 +978,7 @@ func @load_f32(%arg0: memref<f32>) {
// CHECK-NOT: spv.SDiv
// CHECK: spv.Load
// CHECK-NOT: spv.ShiftRightArithmetic
- %0 = load %arg0[] : memref<f32>
+ %0 = memref.load %arg0[] : memref<f32>
return
}
@@ -1000,7 +1000,7 @@ func @store_i8(%arg0: memref<i8>, %value: i8) {
// CHECK: %[[PTR:.+]] = spv.AccessChain %[[ARG0]][%[[ZERO]], %[[ACCESS_IDX]]]
// CHECK: spv.AtomicAnd "Device" "AcquireRelease" %[[PTR]], %[[MASK]]
// CHECK: spv.AtomicOr "Device" "AcquireRelease" %[[PTR]], %[[STORE_VAL]]
- store %value, %arg0[] : memref<i8>
+ memref.store %value, %arg0[] : memref<i8>
return
}
@@ -1026,7 +1026,7 @@ func @store_i16(%arg0: memref<10xi16>, %index: index, %value: i16) {
// CHECK: %[[PTR:.+]] = spv.AccessChain %[[ARG0]][%[[ZERO]], %[[ACCESS_IDX]]]
// CHECK: spv.AtomicAnd "Device" "AcquireRelease" %[[PTR]], %[[MASK]]
// CHECK: spv.AtomicOr "Device" "AcquireRelease" %[[PTR]], %[[STORE_VAL]]
- store %value, %arg0[%index] : memref<10xi16>
+ memref.store %value, %arg0[%index] : memref<10xi16>
return
}
@@ -1035,7 +1035,7 @@ func @store_i32(%arg0: memref<i32>, %value: i32) {
// CHECK: spv.Store
// CHECK-NOT: spv.AtomicAnd
// CHECK-NOT: spv.AtomicOr
- store %value, %arg0[] : memref<i32>
+ memref.store %value, %arg0[] : memref<i32>
return
}
@@ -1044,7 +1044,7 @@ func @store_f32(%arg0: memref<f32>, %value: f32) {
// CHECK: spv.Store
// CHECK-NOT: spv.AtomicAnd
// CHECK-NOT: spv.AtomicOr
- store %value, %arg0[] : memref<f32>
+ memref.store %value, %arg0[] : memref<f32>
return
}
@@ -1077,7 +1077,7 @@ func @load_i8(%arg0: memref<i8>) {
// CHECK: %[[T2:.+]] = spv.Constant 24 : i32
// CHECK: %[[T3:.+]] = spv.ShiftLeftLogical %[[T1]], %[[T2]] : i32, i32
// CHECK: spv.ShiftRightArithmetic %[[T3]], %[[T2]] : i32, i32
- %0 = load %arg0[] : memref<i8>
+ %0 = memref.load %arg0[] : memref<i8>
return
}
@@ -1086,7 +1086,7 @@ func @load_i16(%arg0: memref<i16>) {
// CHECK-NOT: spv.SDiv
// CHECK: spv.Load
// CHECK-NOT: spv.ShiftRightArithmetic
- %0 = load %arg0[] : memref<i16>
+ %0 = memref.load %arg0[] : memref<i16>
return
}
@@ -1108,7 +1108,7 @@ func @store_i8(%arg0: memref<i8>, %value: i8) {
// CHECK: %[[PTR:.+]] = spv.AccessChain %[[ARG0]][%[[ZERO]], %[[ACCESS_IDX]]]
// CHECK: spv.AtomicAnd "Device" "AcquireRelease" %[[PTR]], %[[MASK]]
// CHECK: spv.AtomicOr "Device" "AcquireRelease" %[[PTR]], %[[STORE_VAL]]
- store %value, %arg0[] : memref<i8>
+ memref.store %value, %arg0[] : memref<i8>
return
}
@@ -1117,7 +1117,7 @@ func @store_i16(%arg0: memref<10xi16>, %index: index, %value: i16) {
// CHECK: spv.Store
// CHECK-NOT: spv.AtomicAnd
// CHECK-NOT: spv.AtomicOr
- store %value, %arg0[%index] : memref<10xi16>
+ memref.store %value, %arg0[%index] : memref<10xi16>
return
}
diff --git a/mlir/test/Conversion/StandardToSPIRV/subview-to-spirv.mlir b/mlir/test/Conversion/StandardToSPIRV/subview-to-spirv.mlir
index 44ccda4396aa..942f827275f6 100644
--- a/mlir/test/Conversion/StandardToSPIRV/subview-to-spirv.mlir
+++ b/mlir/test/Conversion/StandardToSPIRV/subview-to-spirv.mlir
@@ -3,7 +3,7 @@
module {
//===----------------------------------------------------------------------===//
-// std.subview
+// memref.subview
//===----------------------------------------------------------------------===//
// CHECK-LABEL: @fold_static_stride_subview
@@ -21,17 +21,17 @@ func @fold_static_stride_subview
// CHECK: %[[T1:.*]] = addi %[[ARG1]], %[[T0]]
// CHECK: %[[T2:.*]] = muli %[[ARG4]], %[[ARG2]]
// CHECK: %[[T3:.*]] = addi %[[T2]], %[[C2]]
- // CHECK: %[[LOADVAL:.*]] = load %[[ARG0]][%[[T1]], %[[T3]]]
+ // CHECK: %[[LOADVAL:.*]] = memref.load %[[ARG0]][%[[T1]], %[[T3]]]
// CHECK: %[[STOREVAL:.*]] = math.sqrt %[[LOADVAL]]
// CHECK: %[[T6:.*]] = muli %[[ARG3]], %[[C3]]
// CHECK: %[[T7:.*]] = addi %[[ARG1]], %[[T6]]
// CHECK: %[[T8:.*]] = muli %[[ARG4]], %[[ARG2]]
// CHECK: %[[T9:.*]] = addi %[[T8]], %[[C2]]
- // CHECK: store %[[STOREVAL]], %[[ARG0]][%[[T7]], %[[T9]]]
- %0 = subview %arg0[%arg1, 2][4, 4][3, %arg2] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [96, ?]>
- %1 = load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [96, ?]>
+ // CHECK: memref.store %[[STOREVAL]], %[[ARG0]][%[[T7]], %[[T9]]]
+ %0 = memref.subview %arg0[%arg1, 2][4, 4][3, %arg2] : memref<12x32xf32> to memref<4x4xf32, offset:?, strides: [96, ?]>
+ %1 = memref.load %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [96, ?]>
%2 = math.sqrt %1 : f32
- store %2, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [96, ?]>
+ memref.store %2, %0[%arg3, %arg4] : memref<4x4xf32, offset:?, strides: [96, ?]>
return
}
diff --git a/mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir b/mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir
index 39826a652092..bf528f611e8d 100644
--- a/mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir
+++ b/mlir/test/Conversion/VectorToLLVM/vector-to-llvm.mlir
@@ -1056,7 +1056,7 @@ func @transfer_read_1d(%A : memref<?xf32>, %base: index) -> vector<17xf32> {
// CHECK: %[[vecPtr:.*]] = llvm.bitcast %[[gep]] :
// CHECK-SAME: !llvm.ptr<f32> to !llvm.ptr<vector<17xf32>>
// CHECK: %[[C0:.*]] = constant 0 : index
-// CHECK: %[[DIM:.*]] = dim %{{.*}}, %[[C0]] : memref<?xf32>
+// CHECK: %[[DIM:.*]] = memref.dim %{{.*}}, %[[C0]] : memref<?xf32>
//
// 2. Create a vector with linear indices [ 0 .. vector_length - 1 ].
// CHECK: %[[linearIndex:.*]] = constant dense
@@ -1118,7 +1118,7 @@ func @transfer_read_2d_to_1d(%A : memref<?x?xf32>, %base0: index, %base1: index)
// CHECK-LABEL: func @transfer_read_2d_to_1d
// CHECK-SAME: %[[BASE_0:[a-zA-Z0-9]*]]: index, %[[BASE_1:[a-zA-Z0-9]*]]: index) -> vector<17xf32>
// CHECK: %[[c1:.*]] = constant 1 : index
-// CHECK: %[[DIM:.*]] = dim %{{.*}}, %[[c1]] : memref<?x?xf32>
+// CHECK: %[[DIM:.*]] = memref.dim %{{.*}}, %[[c1]] : memref<?x?xf32>
//
// Create offsetVector = [ offset + 0 .. offset + vector_length - 1 ].
// CHECK: %[[trunc:.*]] = index_cast %[[BASE_1]] : index to i32
@@ -1152,7 +1152,7 @@ func @transfer_read_1d_non_zero_addrspace(%A : memref<?xf32, 3>, %base: index) -
//
// 2. Check address space of the memref is correct.
// CHECK: %[[c0:.*]] = constant 0 : index
-// CHECK: %[[DIM:.*]] = dim %{{.*}}, %[[c0]] : memref<?xf32, 3>
+// CHECK: %[[DIM:.*]] = memref.dim %{{.*}}, %[[c0]] : memref<?xf32, 3>
//
// 3. Check address space for GEP is correct.
// CHECK: %[[gep_b:.*]] = llvm.getelementptr {{.*}} :
diff --git a/mlir/test/Conversion/VectorToSCF/vector-to-loops.mlir b/mlir/test/Conversion/VectorToSCF/vector-to-loops.mlir
index 7f69638d749b..97e4f4c37dc3 100644
--- a/mlir/test/Conversion/VectorToSCF/vector-to-loops.mlir
+++ b/mlir/test/Conversion/VectorToSCF/vector-to-loops.mlir
@@ -4,7 +4,7 @@
// CHECK-LABEL: func @materialize_read_1d() {
func @materialize_read_1d() {
%f0 = constant 0.0: f32
- %A = alloc () : memref<7x42xf32>
+ %A = memref.alloc () : memref<7x42xf32>
affine.for %i0 = 0 to 7 step 4 {
affine.for %i1 = 0 to 42 step 4 {
%f1 = vector.transfer_read %A[%i0, %i1], %f0 {permutation_map = affine_map<(d0, d1) -> (d0)>} : memref<7x42xf32>, vector<4xf32>
@@ -16,7 +16,7 @@ func @materialize_read_1d() {
%f4 = vector.transfer_read %A[%i0, %ip3], %f0 {permutation_map = affine_map<(d0, d1) -> (d0)>} : memref<7x42xf32>, vector<4xf32>
// Both accesses in the load must be clipped otherwise %i1 + 2 and %i1 + 3 will go out of bounds.
// CHECK: scf.if
- // CHECK-NEXT: load
+ // CHECK-NEXT: memref.load
// CHECK-NEXT: vector.insertelement
// CHECK-NEXT: store
// CHECK-NEXT: else
@@ -35,7 +35,7 @@ func @materialize_read_1d() {
// CHECK-LABEL: func @materialize_read_1d_partially_specialized
func @materialize_read_1d_partially_specialized(%dyn1 : index, %dyn2 : index, %dyn4 : index) {
%f0 = constant 0.0: f32
- %A = alloc (%dyn1, %dyn2, %dyn4) : memref<7x?x?x42x?xf32>
+ %A = memref.alloc (%dyn1, %dyn2, %dyn4) : memref<7x?x?x42x?xf32>
affine.for %i0 = 0 to 7 {
affine.for %i1 = 0 to %dyn1 {
affine.for %i2 = 0 to %dyn2 {
@@ -52,9 +52,9 @@ func @materialize_read_1d_partially_specialized(%dyn1 : index, %dyn2 : index, %d
}
}
}
- // CHECK: %[[tensor:[0-9]+]] = alloc
- // CHECK-NOT: {{.*}} dim %[[tensor]], %c0
- // CHECK-NOT: {{.*}} dim %[[tensor]], %c3
+ // CHECK: %[[tensor:[0-9]+]] = memref.alloc
+ // CHECK-NOT: {{.*}} memref.dim %[[tensor]], %c0
+ // CHECK-NOT: {{.*}} memref.dim %[[tensor]], %c3
return
}
@@ -65,13 +65,13 @@ func @materialize_read_1d_partially_specialized(%dyn1 : index, %dyn2 : index, %d
// CHECK-LABEL: func @materialize_read(%{{.*}}: index, %{{.*}}: index, %{{.*}}: index, %{{.*}}: index) {
func @materialize_read(%M: index, %N: index, %O: index, %P: index) {
%f0 = constant 0.0: f32
- // CHECK-DAG: %[[ALLOC:.*]] = alloca() : memref<5x4xvector<3xf32>>
+ // CHECK-DAG: %[[ALLOC:.*]] = memref.alloca() : memref<5x4xvector<3xf32>>
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C3:.*]] = constant 3 : index
// CHECK-DAG: %[[C4:.*]] = constant 4 : index
// CHECK-DAG: %[[C5:.*]] = constant 5 : index
- // CHECK: %{{.*}} = alloc(%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) : memref<?x?x?x?xf32>
+ // CHECK: %{{.*}} = memref.alloc(%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) : memref<?x?x?x?xf32>
// CHECK-NEXT: affine.for %[[I0:.*]] = 0 to %{{.*}} step 3 {
// CHECK-NEXT: affine.for %[[I1:.*]] = 0 to %{{.*}} {
// CHECK-NEXT: affine.for %[[I2:.*]] = 0 to %{{.*}} {
@@ -80,11 +80,11 @@ func @materialize_read(%M: index, %N: index, %O: index, %P: index) {
// CHECK-NEXT: scf.for %[[I5:.*]] = %[[C0]] to %[[C4]] step %[[C1]] {
// CHECK-NEXT: scf.for %[[I6:.*]] = %[[C0]] to %[[C5]] step %[[C1]] {
// CHECK: %[[VIDX:.*]] = index_cast %[[I4]]
- // CHECK: %[[VEC:.*]] = load %[[ALLOC]][%[[I6]], %[[I5]]] : memref<5x4xvector<3xf32>>
+ // CHECK: %[[VEC:.*]] = memref.load %[[ALLOC]][%[[I6]], %[[I5]]] : memref<5x4xvector<3xf32>>
// CHECK: %[[L0:.*]] = affine.apply #[[$ADD]](%[[I0]], %[[I4]])
// CHECK: %[[L3:.*]] = affine.apply #[[$ADD]](%[[I3]], %[[I6]])
// CHECK-NEXT: scf.if
- // CHECK-NEXT: %[[SCAL:.*]] = load %{{.*}}[%[[L0]], %[[I1]], %[[I2]], %[[L3]]] : memref<?x?x?x?xf32>
+ // CHECK-NEXT: %[[SCAL:.*]] = memref.load %{{.*}}[%[[L0]], %[[I1]], %[[I2]], %[[L3]]] : memref<?x?x?x?xf32>
// CHECK-NEXT: %[[RVEC:.*]] = vector.insertelement %[[SCAL]], %[[VEC]][%[[VIDX]] : i32] : vector<3xf32>
// CHECK-NEXT: store %[[RVEC]], %[[ALLOC]][%[[I6]], %[[I5]]] : memref<5x4xvector<3xf32>>
// CHECK-NEXT: } else {
@@ -95,7 +95,7 @@ func @materialize_read(%M: index, %N: index, %O: index, %P: index) {
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: %[[ALLOC_CAST:.*]] = vector.type_cast %[[ALLOC]] : memref<5x4xvector<3xf32>> to memref<vector<5x4x3xf32>>
- // CHECK-NEXT: %[[LD:.*]] = load %[[ALLOC_CAST]][] : memref<vector<5x4x3xf32>>
+ // CHECK-NEXT: %[[LD:.*]] = memref.load %[[ALLOC_CAST]][] : memref<vector<5x4x3xf32>>
// CHECK-NEXT: "dummy_use"(%[[LD]]) : (vector<5x4x3xf32>) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: }
@@ -107,7 +107,7 @@ func @materialize_read(%M: index, %N: index, %O: index, %P: index) {
// Check that I0 + I4 (of size 3) read from first index load(L0, ...) and write into last index store(..., I4)
// Check that I3 + I6 (of size 5) read from last index load(..., L3) and write into first index store(I6, ...)
// Other dimensions are just accessed with I1, I2 resp.
- %A = alloc (%M, %N, %O, %P) : memref<?x?x?x?xf32, 0>
+ %A = memref.alloc (%M, %N, %O, %P) : memref<?x?x?x?xf32, 0>
affine.for %i0 = 0 to %M step 3 {
affine.for %i1 = 0 to %N {
affine.for %i2 = 0 to %O {
@@ -129,14 +129,14 @@ func @materialize_read(%M: index, %N: index, %O: index, %P: index) {
// CHECK-LABEL:func @materialize_write(%{{.*}}: index, %{{.*}}: index, %{{.*}}: index, %{{.*}}: index) {
func @materialize_write(%M: index, %N: index, %O: index, %P: index) {
- // CHECK-DAG: %[[ALLOC:.*]] = alloca() : memref<5x4xvector<3xf32>>
+ // CHECK-DAG: %[[ALLOC:.*]] = memref.alloca() : memref<5x4xvector<3xf32>>
// CHECK-DAG: %{{.*}} = constant dense<1.000000e+00> : vector<5x4x3xf32>
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C3:.*]] = constant 3 : index
// CHECK-DAG: %[[C4:.*]] = constant 4 : index
// CHECK-DAG: %[[C5:.*]] = constant 5 : index
- // CHECK: %{{.*}} = alloc(%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) : memref<?x?x?x?xf32>
+ // CHECK: %{{.*}} = memref.alloc(%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) : memref<?x?x?x?xf32>
// CHECK-NEXT: affine.for %[[I0:.*]] = 0 to %{{.*}} step 3 {
// CHECK-NEXT: affine.for %[[I1:.*]] = 0 to %{{.*}} step 4 {
// CHECK-NEXT: affine.for %[[I2:.*]] = 0 to %{{.*}} {
@@ -151,7 +151,7 @@ func @materialize_write(%M: index, %N: index, %O: index, %P: index) {
// CHECK: %[[S1:.*]] = affine.apply #[[$ADD]](%[[I1]], %[[I5]])
// CHECK: %[[S3:.*]] = affine.apply #[[$ADD]](%[[I3]], %[[I6]])
// CHECK-NEXT: scf.if
- // CHECK-NEXT: %[[VEC:.*]] = load {{.*}}[%[[I6]], %[[I5]]] : memref<5x4xvector<3xf32>>
+ // CHECK-NEXT: %[[VEC:.*]] = memref.load {{.*}}[%[[I6]], %[[I5]]] : memref<5x4xvector<3xf32>>
// CHECK-NEXT: %[[SCAL:.*]] = vector.extractelement %[[VEC]][%[[VIDX]] : i32] : vector<3xf32>
// CHECK: store %[[SCAL]], {{.*}}[%[[S0]], %[[S1]], %[[I2]], %[[S3]]] : memref<?x?x?x?xf32>
// CHECK-NEXT: }
@@ -169,7 +169,7 @@ func @materialize_write(%M: index, %N: index, %O: index, %P: index) {
// Check that I1 + I5 (of size 4) read from second index load(..., I5, ...) and write into second index store(..., S1, ...)
// Check that I3 + I6 (of size 5) read from first index load(I6, ...) and write into last index store(..., S3)
// Other dimension is just accessed with I2.
- %A = alloc (%M, %N, %O, %P) : memref<?x?x?x?xf32, 0>
+ %A = memref.alloc (%M, %N, %O, %P) : memref<?x?x?x?xf32, 0>
%f1 = constant dense<1.000000e+00> : vector<5x4x3xf32>
affine.for %i0 = 0 to %M step 3 {
affine.for %i1 = 0 to %N step 4 {
@@ -204,9 +204,9 @@ func @transfer_read_progressive(%A : memref<?x?xf32>, %base: index) -> vector<3x
%f7 = constant 7.0: f32
// CHECK-DAG: %[[splat:.*]] = constant dense<7.000000e+00> : vector<15xf32>
- // CHECK-DAG: %[[alloc:.*]] = alloca() : memref<3xvector<15xf32>>
+ // CHECK-DAG: %[[alloc:.*]] = memref.alloca() : memref<3xvector<15xf32>>
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
- // CHECK-DAG: %[[dim:.*]] = dim %[[A]], %[[C0]] : memref<?x?xf32>
+ // CHECK-DAG: %[[dim:.*]] = memref.dim %[[A]], %[[C0]] : memref<?x?xf32>
// CHECK: affine.for %[[I:.*]] = 0 to 3 {
// CHECK: %[[add:.*]] = affine.apply #[[$MAP0]](%[[I]])[%[[base]]]
// CHECK: %[[cond1:.*]] = cmpi slt, %[[add]], %[[dim]] : index
@@ -217,13 +217,13 @@ func @transfer_read_progressive(%A : memref<?x?xf32>, %base: index) -> vector<3x
// CHECK: store %[[splat]], %[[alloc]][%[[I]]] : memref<3xvector<15xf32>>
// CHECK: }
// CHECK: %[[vmemref:.*]] = vector.type_cast %[[alloc]] : memref<3xvector<15xf32>> to memref<vector<3x15xf32>>
- // CHECK: %[[cst:.*]] = load %[[vmemref]][] : memref<vector<3x15xf32>>
+ // CHECK: %[[cst:.*]] = memref.load %[[vmemref]][] : memref<vector<3x15xf32>>
// FULL-UNROLL: %[[pad:.*]] = constant 7.000000e+00 : f32
// FULL-UNROLL: %[[VEC0:.*]] = constant dense<7.000000e+00> : vector<3x15xf32>
// FULL-UNROLL: %[[C0:.*]] = constant 0 : index
// FULL-UNROLL: %[[SPLAT:.*]] = constant dense<7.000000e+00> : vector<15xf32>
- // FULL-UNROLL: %[[DIM:.*]] = dim %[[A]], %[[C0]] : memref<?x?xf32>
+ // FULL-UNROLL: %[[DIM:.*]] = memref.dim %[[A]], %[[C0]] : memref<?x?xf32>
// FULL-UNROLL: cmpi slt, %[[base]], %[[DIM]] : index
// FULL-UNROLL: %[[VEC1:.*]] = scf.if %{{.*}} -> (vector<3x15xf32>) {
// FULL-UNROLL: vector.transfer_read %[[A]][%[[base]], %[[base]]], %[[pad]] : memref<?x?xf32>, vector<15xf32>
@@ -277,20 +277,20 @@ func @transfer_read_progressive(%A : memref<?x?xf32>, %base: index) -> vector<3x
// FULL-UNROLL-SAME: %[[vec:[a-zA-Z0-9]+]]: vector<3x15xf32>
func @transfer_write_progressive(%A : memref<?x?xf32>, %base: index, %vec: vector<3x15xf32>) {
// CHECK: %[[C0:.*]] = constant 0 : index
- // CHECK: %[[alloc:.*]] = alloca() : memref<3xvector<15xf32>>
+ // CHECK: %[[alloc:.*]] = memref.alloca() : memref<3xvector<15xf32>>
// CHECK: %[[vmemref:.*]] = vector.type_cast %[[alloc]] : memref<3xvector<15xf32>> to memref<vector<3x15xf32>>
// CHECK: store %[[vec]], %[[vmemref]][] : memref<vector<3x15xf32>>
- // CHECK: %[[dim:.*]] = dim %[[A]], %[[C0]] : memref<?x?xf32>
+ // CHECK: %[[dim:.*]] = memref.dim %[[A]], %[[C0]] : memref<?x?xf32>
// CHECK: affine.for %[[I:.*]] = 0 to 3 {
// CHECK: %[[add:.*]] = affine.apply #[[$MAP0]](%[[I]])[%[[base]]]
// CHECK: %[[cmp:.*]] = cmpi slt, %[[add]], %[[dim]] : index
// CHECK: scf.if %[[cmp]] {
- // CHECK: %[[vec_1d:.*]] = load %0[%[[I]]] : memref<3xvector<15xf32>>
+ // CHECK: %[[vec_1d:.*]] = memref.load %0[%[[I]]] : memref<3xvector<15xf32>>
// CHECK: vector.transfer_write %[[vec_1d]], %[[A]][%[[add]], %[[base]]] : vector<15xf32>, memref<?x?xf32>
// CHECK: }
// FULL-UNROLL: %[[C0:.*]] = constant 0 : index
- // FULL-UNROLL: %[[DIM:.*]] = dim %[[A]], %[[C0]] : memref<?x?xf32>
+ // FULL-UNROLL: %[[DIM:.*]] = memref.dim %[[A]], %[[C0]] : memref<?x?xf32>
// FULL-UNROLL: %[[CMP0:.*]] = cmpi slt, %[[base]], %[[DIM]] : index
// FULL-UNROLL: scf.if %[[CMP0]] {
// FULL-UNROLL: %[[V0:.*]] = vector.extract %[[vec]][0] : vector<3x15xf32>
@@ -331,12 +331,12 @@ func @transfer_write_progressive(%A : memref<?x?xf32>, %base: index, %vec: vecto
// FULL-UNROLL-SAME: %[[vec:[a-zA-Z0-9]+]]: vector<3x15xf32>
func @transfer_write_progressive_unmasked(%A : memref<?x?xf32>, %base: index, %vec: vector<3x15xf32>) {
// CHECK-NOT: scf.if
- // CHECK-NEXT: %[[alloc:.*]] = alloca() : memref<3xvector<15xf32>>
+ // CHECK-NEXT: %[[alloc:.*]] = memref.alloca() : memref<3xvector<15xf32>>
// CHECK-NEXT: %[[vmemref:.*]] = vector.type_cast %[[alloc]] : memref<3xvector<15xf32>> to memref<vector<3x15xf32>>
// CHECK-NEXT: store %[[vec]], %[[vmemref]][] : memref<vector<3x15xf32>>
// CHECK-NEXT: affine.for %[[I:.*]] = 0 to 3 {
// CHECK-NEXT: %[[add:.*]] = affine.apply #[[$MAP0]](%[[I]])[%[[base]]]
- // CHECK-NEXT: %[[vec_1d:.*]] = load %0[%[[I]]] : memref<3xvector<15xf32>>
+ // CHECK-NEXT: %[[vec_1d:.*]] = memref.load %0[%[[I]]] : memref<3xvector<15xf32>>
// CHECK-NEXT: vector.transfer_write %[[vec_1d]], %[[A]][%[[add]], %[[base]]] {masked = [false]} : vector<15xf32>, memref<?x?xf32>
// FULL-UNROLL: %[[VEC0:.*]] = vector.extract %[[vec]][0] : vector<3x15xf32>
@@ -384,8 +384,8 @@ func @transfer_read_minor_identity(%A : memref<?x?x?x?xf32>) -> vector<3x3xf32>
// CHECK: %[[cst:.*]] = constant 0.000000e+00 : f32
// CHECK: %[[c2:.*]] = constant 2 : index
// CHECK: %[[cst0:.*]] = constant dense<0.000000e+00> : vector<3xf32>
-// CHECK: %[[m:.*]] = alloca() : memref<3xvector<3xf32>>
-// CHECK: %[[d:.*]] = dim %[[A]], %[[c2]] : memref<?x?x?x?xf32>
+// CHECK: %[[m:.*]] = memref.alloca() : memref<3xvector<3xf32>>
+// CHECK: %[[d:.*]] = memref.dim %[[A]], %[[c2]] : memref<?x?x?x?xf32>
// CHECK: affine.for %[[arg1:.*]] = 0 to 3 {
// CHECK: %[[cmp:.*]] = cmpi slt, %[[arg1]], %[[d]] : index
// CHECK: scf.if %[[cmp]] {
@@ -396,7 +396,7 @@ func @transfer_read_minor_identity(%A : memref<?x?x?x?xf32>) -> vector<3x3xf32>
// CHECK: }
// CHECK: }
// CHECK: %[[cast:.*]] = vector.type_cast %[[m]] : memref<3xvector<3xf32>> to memref<vector<3x3xf32>>
-// CHECK: %[[ret:.*]] = load %[[cast]][] : memref<vector<3x3xf32>>
+// CHECK: %[[ret:.*]] = memref.load %[[cast]][] : memref<vector<3x3xf32>>
// CHECK: return %[[ret]] : vector<3x3xf32>
func @transfer_write_minor_identity(%A : vector<3x3xf32>, %B : memref<?x?x?x?xf32>) {
@@ -413,14 +413,14 @@ func @transfer_write_minor_identity(%A : vector<3x3xf32>, %B : memref<?x?x?x?xf3
// CHECK-SAME: %[[B:.*]]: memref<?x?x?x?xf32>)
// CHECK: %[[c0:.*]] = constant 0 : index
// CHECK: %[[c2:.*]] = constant 2 : index
-// CHECK: %[[m:.*]] = alloca() : memref<3xvector<3xf32>>
+// CHECK: %[[m:.*]] = memref.alloca() : memref<3xvector<3xf32>>
// CHECK: %[[cast:.*]] = vector.type_cast %[[m]] : memref<3xvector<3xf32>> to memref<vector<3x3xf32>>
// CHECK: store %[[A]], %[[cast]][] : memref<vector<3x3xf32>>
-// CHECK: %[[d:.*]] = dim %[[B]], %[[c2]] : memref<?x?x?x?xf32>
+// CHECK: %[[d:.*]] = memref.dim %[[B]], %[[c2]] : memref<?x?x?x?xf32>
// CHECK: affine.for %[[arg2:.*]] = 0 to 3 {
// CHECK: %[[cmp:.*]] = cmpi slt, %[[arg2]], %[[d]] : index
// CHECK: scf.if %[[cmp]] {
-// CHECK: %[[tmp:.*]] = load %[[m]][%[[arg2]]] : memref<3xvector<3xf32>>
+// CHECK: %[[tmp:.*]] = memref.load %[[m]][%[[arg2]]] : memref<3xvector<3xf32>>
// CHECK: vector.transfer_write %[[tmp]], %[[B]][%[[c0]], %[[c0]], %[[arg2]], %[[c0]]] : vector<3xf32>, memref<?x?x?x?xf32>
// CHECK: }
// CHECK: }
@@ -438,7 +438,7 @@ func @transfer_read_strided(%A : memref<8x4xf32, affine_map<(d0, d1) -> (d0 + d1
// CHECK-LABEL: transfer_read_strided(
// CHECK: scf.for
-// CHECK: load
+// CHECK: memref.load
func @transfer_write_strided(%A : vector<4xf32>, %B : memref<8x4xf32, affine_map<(d0, d1) -> (d0 + d1 * 8)>>) {
%c0 = constant 0 : index
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/uniform_divergent.mlir b/mlir/test/Dialect/Affine/SuperVectorize/uniform_divergent.mlir
index 9d9f06177e3b..2bda792937ae 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/uniform_divergent.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/uniform_divergent.mlir
@@ -42,7 +42,7 @@ func @multi_use_uniform_arg(%in : memref<512xf32>, %uniform : f32) {
// CHECK-LABEL: @uniform_load
func @uniform_load(%A : memref<?x?xf32>, %C : memref<?x?xf32>) {
%c0 = constant 0 : index
- %N = dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c0 : memref<?x?xf32>
affine.for %i = 0 to %N {
%uniform_ld = affine.load %A[%i, %i] : memref<?x?xf32>
affine.for %j = 0 to %N {
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vector_utils.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vector_utils.mlir
index ef8e2831a848..b0378b4d7508 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vector_utils.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vector_utils.mlir
@@ -4,12 +4,12 @@
func @vector_add_2d(%arg0: index, %arg1: index) -> f32 {
// Nothing should be matched in this first block.
- // CHECK-NOT:matched: {{.*}} = alloc{{.*}}
+ // CHECK-NOT:matched: {{.*}} = memref.alloc{{.*}}
// CHECK-NOT:matched: {{.*}} = constant 0{{.*}}
// CHECK-NOT:matched: {{.*}} = constant 1{{.*}}
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
- %1 = alloc(%arg0, %arg1) : memref<?x?xf32>
- %2 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %2 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
%c0 = constant 0 : index
%cst = constant 1.000000e+00 : f32
@@ -29,11 +29,11 @@ func @vector_add_2d(%arg0: index, %arg1: index) -> f32 {
// Nothing should be matched in this last block.
// CHECK-NOT:matched: {{.*}} = constant 7{{.*}}
// CHECK-NOT:matched: {{.*}} = constant 42{{.*}}
- // CHECK-NOT:matched: {{.*}} = load{{.*}}
+ // CHECK-NOT:matched: {{.*}} = memref.load{{.*}}
// CHECK-NOT:matched: return {{.*}}
%c7 = constant 7 : index
%c42 = constant 42 : index
- %9 = load %2[%c7, %c42] : memref<?x?xf32>
+ %9 = memref.load %2[%c7, %c42] : memref<?x?xf32>
return %9 : f32
}
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_1d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_1d.mlir
index ddfab3b66ed8..09ee6d513611 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_1d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_1d.mlir
@@ -8,15 +8,15 @@ func @vec1d_1(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: for {{.*}} step 128
// CHECK-NEXT: %{{.*}} = affine.apply #[[$map_id1]](%[[C0]])
@@ -36,15 +36,15 @@ func @vec1d_2(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK:for [[IV3:%[a-zA-Z0-9]+]] = 0 to [[ARG_M]] step 128
// CHECK-NEXT: %[[CST:.*]] = constant 0.0{{.*}}: f32
@@ -62,15 +62,15 @@ func @vec1d_3(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %arg0, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %arg0, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %arg1, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %arg0, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %arg0, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %arg1, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK:for [[IV8:%[arg0-9]+]] = 0 to [[ARG_M]] step 128
// CHECK-NEXT: for [[IV9:%[arg0-9]*]] = 0 to [[ARG_N]] {
@@ -90,9 +90,9 @@ func @vec1d_3(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-LABEL: func @vector_add_2d
func @vector_add_2d(%M : index, %N : index) -> f32 {
- %A = alloc (%M, %N) : memref<?x?xf32, 0>
- %B = alloc (%M, %N) : memref<?x?xf32, 0>
- %C = alloc (%M, %N) : memref<?x?xf32, 0>
+ %A = memref.alloc (%M, %N) : memref<?x?xf32, 0>
+ %B = memref.alloc (%M, %N) : memref<?x?xf32, 0>
+ %C = memref.alloc (%M, %N) : memref<?x?xf32, 0>
%f1 = constant 1.0 : f32
%f2 = constant 2.0 : f32
affine.for %i0 = 0 to %M {
@@ -147,15 +147,15 @@ func @vec_rejected_1(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK:for {{.*}} [[ARG_M]] {
affine.for %i1 = 0 to %M { // not vectorized
@@ -171,15 +171,15 @@ func @vec_rejected_2(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}}{{[0-9]*}} = 0 to [[ARG_M]] {
affine.for %i2 = 0 to %M { // not vectorized, would vectorize with --test-fastest-varying=1
@@ -195,15 +195,15 @@ func @vec_rejected_3(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK:for [[IV4:%[arg0-9]+]] = 0 to [[ARG_M]] step 128 {
// CHECK-NEXT: for [[IV5:%[arg0-9]*]] = 0 to [[ARG_N]] {
@@ -224,15 +224,15 @@ func @vec_rejected_4(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: for [[IV6:%[arg0-9]*]] = 0 to [[ARG_M]] {
// CHECK-NEXT: for [[IV7:%[arg0-9]*]] = 0 to [[ARG_N]] {
@@ -251,15 +251,15 @@ func @vec_rejected_5(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: for [[IV10:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
// CHECK: for [[IV11:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
@@ -279,15 +279,15 @@ func @vec_rejected_6(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: for [[IV12:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
// CHECK: for [[IV13:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
@@ -309,19 +309,19 @@ func @vec_rejected_7(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}}{{[0-9]*}} = 0 to %{{[0-9]*}} {
affine.for %i16 = 0 to %M { // not vectorized, can't vectorize a vector load
- %a16 = alloc(%M) : memref<?xvector<2xf32>>
+ %a16 = memref.alloc(%M) : memref<?xvector<2xf32>>
%l16 = affine.load %a16[%i16] : memref<?xvector<2xf32>>
}
return
@@ -337,15 +337,15 @@ func @vec_rejected_8(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}}{{[0-9]*}} = 0 to %{{[0-9]*}} {
// CHECK: for [[IV18:%[a-zA-Z0-9]+]] = 0 to [[ARG_M]] step 128
@@ -371,15 +371,15 @@ func @vec_rejected_9(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}}{{[0-9]*}} = 0 to %{{[0-9]*}} {
// CHECK: for [[IV18:%[a-zA-Z0-9]+]] = 0 to [[ARG_M]] step 128
@@ -404,15 +404,15 @@ func @vec_rejected_10(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
-// CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
-// CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+// CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+// CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}}{{[0-9]*}} = 0 to %{{[0-9]*}} {
affine.for %i15 = 0 to %M { // not vectorized due to condition below
@@ -430,15 +430,15 @@ func @vec_rejected_11(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
- // CHECK-DAG: [[ARG_M:%[0-9]+]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
- // CHECK-DAG: [[ARG_N:%[0-9]+]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
- // CHECK-DAG: [[ARG_P:%[0-9]+]] = dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
+ // CHECK-DAG: [[ARG_M:%[0-9]+]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+ // CHECK-DAG: [[ARG_N:%[0-9]+]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+ // CHECK-DAG: [[ARG_P:%[0-9]+]] = memref.dim %{{.*}}, %[[C2]] : memref<?x?x?xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?xf32>
- %N = dim %A, %c1 : memref<?x?xf32>
- %P = dim %B, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?xf32>
+ %P = memref.dim %B, %c2 : memref<?x?x?xf32>
// CHECK: for [[IV10:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
// CHECK: for [[IV11:%[arg0-9]*]] = 0 to %{{[0-9]*}} {
@@ -458,7 +458,7 @@ func @vec_rejected_11(%A : memref<?x?xf32>, %B : memref<?x?x?xf32>) {
// CHECK-LABEL: @vec_rejected_sequential
func @vec_rejected_sequential(%A : memref<?xf32>) {
%c0 = constant 0 : index
- %N = dim %A, %c0 : memref<?xf32>
+ %N = memref.dim %A, %c0 : memref<?xf32>
affine.for %i = 0 to %N {
// CHECK-NOT: vector
%a = affine.load %A[%i] : memref<?xf32>
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_2d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_2d.mlir
index 86208ff24caf..3d467405805f 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_2d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_2d.mlir
@@ -13,9 +13,9 @@ func @vec2d(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?x?xf32>
- %N = dim %A, %c1 : memref<?x?x?xf32>
- %P = dim %A, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %P = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: for {{.*}} = 0 to %{{.*}} {
// CHECK: for {{.*}} = 0 to %{{.*}} step 32
// CHECK: for {{.*}} = 0 to %{{.*}} step 256
@@ -47,9 +47,9 @@ func @vec2d(%A : memref<?x?x?xf32>) {
}
func @vector_add_2d(%M : index, %N : index) -> f32 {
- %A = alloc (%M, %N) : memref<?x?xf32, 0>
- %B = alloc (%M, %N) : memref<?x?xf32, 0>
- %C = alloc (%M, %N) : memref<?x?xf32, 0>
+ %A = memref.alloc (%M, %N) : memref<?x?xf32, 0>
+ %B = memref.alloc (%M, %N) : memref<?x?xf32, 0>
+ %C = memref.alloc (%M, %N) : memref<?x?xf32, 0>
%f1 = constant 1.0 : f32
%f2 = constant 2.0 : f32
affine.for %i0 = 0 to %M {
@@ -102,14 +102,14 @@ func @vector_add_2d(%M : index, %N : index) -> f32 {
func @vectorize_matmul(%arg0: memref<?x?xf32>, %arg1: memref<?x?xf32>, %arg2: memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %M = dim %arg0, %c0 : memref<?x?xf32>
- %K = dim %arg0, %c1 : memref<?x?xf32>
- %N = dim %arg2, %c1 : memref<?x?xf32>
+ %M = memref.dim %arg0, %c0 : memref<?x?xf32>
+ %K = memref.dim %arg0, %c1 : memref<?x?xf32>
+ %N = memref.dim %arg2, %c1 : memref<?x?xf32>
// VECT: %[[C0:.*]] = constant 0 : index
// VECT-NEXT: %[[C1:.*]] = constant 1 : index
- // VECT-NEXT: %[[M:.*]] = dim %{{.*}}, %[[C0]] : memref<?x?xf32>
- // VECT-NEXT: %[[K:.*]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
- // VECT-NEXT: %[[N:.*]] = dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+ // VECT-NEXT: %[[M:.*]] = memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32>
+ // VECT-NEXT: %[[K:.*]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
+ // VECT-NEXT: %[[N:.*]] = memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32>
// VECT: {{.*}} #[[$map_id1]](%[[M]]) step 4 {
// VECT-NEXT: {{.*}} #[[$map_id1]](%[[N]]) step 8 {
// VECT: %[[VC0:.*]] = constant dense<0.000000e+00> : vector<4x8xf32>
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_3d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_3d.mlir
index f96d73decd2c..1e054fa84c04 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_3d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_3d.mlir
@@ -4,9 +4,9 @@ func @vec3d(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %A, %c0 : memref<?x?x?xf32>
- %1 = dim %A, %c1 : memref<?x?x?xf32>
- %2 = dim %A, %c2 : memref<?x?x?xf32>
+ %0 = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %1 = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %2 = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 32 {
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_2d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_2d.mlir
index 19161e723753..39be5467e208 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_2d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_2d.mlir
@@ -7,9 +7,9 @@ func @vec2d(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?x?xf32>
- %N = dim %A, %c1 : memref<?x?x?xf32>
- %P = dim %A, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %P = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 32
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 256
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_transpose_2d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_transpose_2d.mlir
index ef2ca22eea50..4783ec1b8b93 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_transpose_2d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_outer_loop_transpose_2d.mlir
@@ -7,9 +7,9 @@ func @vec2d(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?x?xf32>
- %N = dim %A, %c1 : memref<?x?x?xf32>
- %P = dim %A, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %P = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: for {{.*}} = 0 to %{{.*}} {
// CHECK: for {{.*}} = 0 to %{{.*}} {
// CHECK: for {{.*}} = 0 to %{{.*}} {
@@ -40,9 +40,9 @@ func @vec2d_imperfectly_nested(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %A, %c0 : memref<?x?x?xf32>
- %1 = dim %A, %c1 : memref<?x?x?xf32>
- %2 = dim %A, %c2 : memref<?x?x?xf32>
+ %0 = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %1 = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %2 = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 32 {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 256 {
diff --git a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_transpose_2d.mlir b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_transpose_2d.mlir
index 925d4506d1d2..38a828e17d1a 100644
--- a/mlir/test/Dialect/Affine/SuperVectorize/vectorize_transpose_2d.mlir
+++ b/mlir/test/Dialect/Affine/SuperVectorize/vectorize_transpose_2d.mlir
@@ -7,9 +7,9 @@ func @vec2d(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %A, %c0 : memref<?x?x?xf32>
- %N = dim %A, %c1 : memref<?x?x?xf32>
- %P = dim %A, %c2 : memref<?x?x?xf32>
+ %M = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %N = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %P = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: for {{.*}} = 0 to %{{.*}} {
// CHECK: for {{.*}} = 0 to %{{.*}} {
// CHECK: for {{.*}} = 0 to %{{.*}} {
@@ -40,9 +40,9 @@ func @vec2d_imperfectly_nested(%A : memref<?x?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %A, %c0 : memref<?x?x?xf32>
- %1 = dim %A, %c1 : memref<?x?x?xf32>
- %2 = dim %A, %c2 : memref<?x?x?xf32>
+ %0 = memref.dim %A, %c0 : memref<?x?x?xf32>
+ %1 = memref.dim %A, %c1 : memref<?x?x?xf32>
+ %2 = memref.dim %A, %c2 : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 32 {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} step 256 {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
diff --git a/mlir/test/Dialect/Affine/affine-data-copy.mlir b/mlir/test/Dialect/Affine/affine-data-copy.mlir
index 697196561b15..11288784a4e9 100644
--- a/mlir/test/Dialect/Affine/affine-data-copy.mlir
+++ b/mlir/test/Dialect/Affine/affine-data-copy.mlir
@@ -49,7 +49,7 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// CHECK: affine.for %[[I:.*]] = 0 to 4096 step 128 {
// CHECK: affine.for %[[J:.*]] = 0 to 4096 step 128 {
-// CHECK: [[BUFC:%[0-9]+]] = alloc() : memref<128x128xf32>
+// CHECK: [[BUFC:%[0-9]+]] = memref.alloc() : memref<128x128xf32>
// The result matrix's copy gets hoisted out.
// Result matrix copy-in.
// CHECK: affine.for %[[II:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
@@ -61,7 +61,7 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// LHS matrix copy-in.
// CHECK: affine.for %[[K:.*]] = 0 to 4096 step 128 {
-// CHECK: [[BUFA:%[0-9]+]] = alloc() : memref<128x128xf32>
+// CHECK: [[BUFA:%[0-9]+]] = memref.alloc() : memref<128x128xf32>
// CHECK: affine.for %[[II:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %[[KK:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
@@ -70,7 +70,7 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// CHECK: }
// RHS matrix copy-in.
-// CHECK: [[BUFB:%[0-9]+]] = alloc() : memref<128x128xf32>
+// CHECK: [[BUFB:%[0-9]+]] = memref.alloc() : memref<128x128xf32>
// CHECK: affine.for %[[KK:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.for %[[JJ:.*]] = #[[$MAP_IDENTITY]](%{{.*}}) to #[[$MAP_PLUS_128]](%{{.*}}) {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
@@ -91,8 +91,8 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: dealloc [[BUFB]] : memref<128x128xf32>
-// CHECK: dealloc [[BUFA]] : memref<128x128xf32>
+// CHECK: memref.dealloc [[BUFB]] : memref<128x128xf32>
+// CHECK: memref.dealloc [[BUFA]] : memref<128x128xf32>
// CHECK: }
// Result matrix copy out.
@@ -102,15 +102,15 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// CHECK: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<4096x4096xf32>
// CHECK: }
// CHECK: }
-// CHECK: dealloc [[BUFC]] : memref<128x128xf32>
+// CHECK: memref.dealloc [[BUFC]] : memref<128x128xf32>
// CHECK: }
// CHECK: }
// Check that only one memref is copied when memref filter is used.
// FILTER: affine.for %{{.*}} = 0 to 4096 step 128 {
-// FILTER: alloc() : memref<128x4096xf32>
-// FILTER-NOT: alloc()
+// FILTER: memref.alloc() : memref<128x4096xf32>
+// FILTER-NOT: memref.alloc()
// FILTER: affine.for
// FILTER: affine.for %{{.*}} = 0 to 4096 {
// FILTER: affine.for %{{.*}} = 0 to 4096 step 128 {
@@ -118,8 +118,8 @@ func @matmul(%A: memref<4096x4096xf32>, %B: memref<4096x4096xf32>, %C: memref<40
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
// FILTER-NEXT: affine.for %{{.*}} = #map{{.*}}(%{{.*}}) to #map{{.*}}(%{{.*}}) {
-// FILTER: dealloc %{{.*}} : memref<128x4096xf32>
-// FILTER-NOT: dealloc %{{.*}} : memref<128x4096xf32>
+// FILTER: memref.dealloc %{{.*}} : memref<128x4096xf32>
+// FILTER-NOT: memref.dealloc %{{.*}} : memref<128x4096xf32>
// -----
@@ -145,43 +145,43 @@ func @single_elt_buffers(%arg0: memref<1024x1024xf32>, %arg1: memref<1024x1024xf
}
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
-// CHECK-SMALL: alloc() : memref<1x1xf32>
+// CHECK-SMALL: memref.alloc() : memref<1x1xf32>
// CHECK-SMALL: affine.load %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.for %arg{{.*}} = 0 to 1024 {
-// CHECK-SMALL: alloc() : memref<1x1xf32>
+// CHECK-SMALL: memref.alloc() : memref<1x1xf32>
// CHECK-SMALL: affine.load %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: addf %{{.*}}, %{{.*}} : f32
// CHECK-SMALL: affine.store %{{.*}}, %{{.*}}[0, 0] : memref<1x1xf32>
-// CHECK-SMALL: dealloc %{{.*}} : memref<1x1xf32>
+// CHECK-SMALL: memref.dealloc %{{.*}} : memref<1x1xf32>
// CHECK-SMALL: }
// CHECK-SMALL: affine.load %{{.*}}[0, 0] : memref<1x1xf32>
// CHECK-SMALL: affine.store %{{.*}}, %arg{{.*}}[%{{.*}}, %{{.*}}] : memref<1024x1024xf32>
-// CHECK-SMALL: dealloc %{{.*}} : memref<1x1xf32>
+// CHECK-SMALL: memref.dealloc %{{.*}} : memref<1x1xf32>
// CHECK-SMALL: }
// CHECK-SMALL: }
// CHECK-SMALL: return
// Check that only one memref is copied when memref filter is used.
-// FILTER: alloc() : memref<1024x1024xf32>
-// FILTER-NOT: alloc()
+// FILTER: memref.alloc() : memref<1024x1024xf32>
+// FILTER-NOT: memref.alloc()
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER: affine.for %{{.*}} = 0 to 1024 {
// FILTER-NEXT: affine.for %{{.*}} = 0 to 1024 {
// FILTER-NEXT: affine.for %{{.*}} = 0 to 1024 {
-// FILTER: dealloc %{{.*}} : memref<1024x1024xf32>
-// FILTER-NOT: dealloc
+// FILTER: memref.dealloc %{{.*}} : memref<1024x1024xf32>
+// FILTER-NOT: memref.dealloc
// FILTER: return
// CHeck that only one memref is copied, because for-memref-region is enabled
// (and the first ever encountered load is analyzed).
-// MEMREF_REGION: alloc() : memref<1024x1024xf32>
-// MEMREF_REGION-NOT: alloc()
+// MEMREF_REGION: memref.alloc() : memref<1024x1024xf32>
+// MEMREF_REGION-NOT: memref.alloc()
// MEMREF_REGION: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION: }
@@ -189,8 +189,8 @@ func @single_elt_buffers(%arg0: memref<1024x1024xf32>, %arg1: memref<1024x1024xf
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
// MEMREF_REGION-NEXT: affine.for %{{.*}} = 0 to 1024 {
-// MEMREF_REGION: dealloc %{{.*}} : memref<1024x1024xf32>
-// MEMREF_REGION-NOT: dealloc
+// MEMREF_REGION: memref.dealloc %{{.*}} : memref<1024x1024xf32>
+// MEMREF_REGION-NOT: memref.dealloc
// MEMREF_REGION-NEXT: return
// -----
@@ -216,7 +216,7 @@ func @min_upper_bound(%A: memref<4096xf32>) -> memref<4096xf32> {
return %A : memref<4096xf32>
}
// CHECK: affine.for %[[IV1:.*]] = 0 to 4096 step 100
-// CHECK: %[[BUF:.*]] = alloc() : memref<100xf32>
+// CHECK: %[[BUF:.*]] = memref.alloc() : memref<100xf32>
// CHECK-NEXT: affine.for %[[IV2:.*]] = #[[$MAP_IDENTITY]](%[[IV1]]) to min #[[$MAP_MIN_UB1]](%[[IV1]]) {
// CHECK-NEXT: affine.load %{{.*}}[%[[IV2]]] : memref<4096xf32>
// CHECK-NEXT: affine.store %{{.*}}, %[[BUF]][%[[IV2]] - %[[IV1]]] : memref<100xf32>
@@ -230,7 +230,7 @@ func @min_upper_bound(%A: memref<4096xf32>) -> memref<4096xf32> {
// CHECK-NEXT: affine.load %[[BUF]][%[[IV2]] - %[[IV1]]] : memref<100xf32>
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%[[IV2]]] : memref<4096xf32>
// CHECK-NEXT: }
-// CHECK-NEXT: dealloc %[[BUF]] : memref<100xf32>
+// CHECK-NEXT: memref.dealloc %[[BUF]] : memref<100xf32>
// CHECK-NEXT: }
// -----
@@ -257,7 +257,7 @@ func @max_lower_bound(%M: memref<2048x516xf64>, %i : index, %j : index) {
return
}
-// CHECK: %[[BUF:.*]] = alloc() : memref<2048x6xf64>
+// CHECK: %[[BUF:.*]] = memref.alloc() : memref<2048x6xf64>
// CHECK-NEXT: affine.for %[[ii:.*]] = 0 to 2048 {
// CHECK-NEXT: affine.for %[[jj:.*]] = max #[[$LB]]()[%[[i]], %[[j]]] to min #[[$UB]]()[%[[i]], %[[j]]] {
// CHECK-NEXT: affine.load %{{.*}}[%[[ii]], %[[jj]]] : memref<2048x516xf64>
@@ -269,4 +269,4 @@ func @max_lower_bound(%M: memref<2048x516xf64>, %i : index, %j : index) {
// CHECK-NEXT: affine.load %[[BUF]][%[[ii_]], %[[jj_]] - symbol(%[[j]]) * 6] : memref<2048x6xf64>
// CHECK-NEXT: }
// CHECK-NEXT: }
-// CHECK-NEXT: dealloc %[[BUF]] : memref<2048x6xf64>
+// CHECK-NEXT: memref.dealloc %[[BUF]] : memref<2048x6xf64>
diff --git a/mlir/test/Dialect/Affine/affine-loop-invariant-code-motion.mlir b/mlir/test/Dialect/Affine/affine-loop-invariant-code-motion.mlir
index 0c7a5fa06407..0b8c036bbc29 100644
--- a/mlir/test/Dialect/Affine/affine-loop-invariant-code-motion.mlir
+++ b/mlir/test/Dialect/Affine/affine-loop-invariant-code-motion.mlir
@@ -1,7 +1,7 @@
// RUN: mlir-opt %s -affine-loop-invariant-code-motion -split-input-file | FileCheck %s
func @nested_loops_both_having_invariant_code() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -12,7 +12,7 @@ func @nested_loops_both_having_invariant_code() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -29,14 +29,14 @@ func @nested_loops_both_having_invariant_code() {
// CHECK-LABEL: func @store_affine_apply
func @store_affine_apply() -> memref<10xf32> {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %arg0 = 0 to 10 {
%t0 = affine.apply affine_map<(d1) -> (d1 + 1)>(%arg0)
affine.store %cf7, %m[%t0] : memref<10xf32>
}
return %m : memref<10xf32>
// CHECK: %cst = constant 7.000000e+00 : f32
-// CHECK-NEXT: %0 = alloc() : memref<10xf32>
+// CHECK-NEXT: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply #map{{[0-9]*}}(%arg0)
// CHECK-NEXT: affine.store %cst, %0[%1] : memref<10xf32>
@@ -47,7 +47,7 @@ func @store_affine_apply() -> memref<10xf32> {
// -----
func @nested_loops_code_invariant_to_both() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -57,7 +57,7 @@ func @nested_loops_code_invariant_to_both() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -68,8 +68,8 @@ func @nested_loops_code_invariant_to_both() {
// -----
func @single_loop_nothing_invariant() {
- %m1 = alloc() : memref<10xf32>
- %m2 = alloc() : memref<10xf32>
+ %m1 = memref.alloc() : memref<10xf32>
+ %m2 = memref.alloc() : memref<10xf32>
affine.for %arg0 = 0 to 10 {
%v0 = affine.load %m1[%arg0] : memref<10xf32>
%v1 = affine.load %m2[%arg0] : memref<10xf32>
@@ -77,8 +77,8 @@ func @single_loop_nothing_invariant() {
affine.store %v2, %m1[%arg0] : memref<10xf32>
}
- // CHECK: %0 = alloc() : memref<10xf32>
- // CHECK-NEXT: %1 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
+ // CHECK-NEXT: %1 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: %2 = affine.load %0[%arg0] : memref<10xf32>
// CHECK-NEXT: %3 = affine.load %1[%arg0] : memref<10xf32>
@@ -91,7 +91,7 @@ func @single_loop_nothing_invariant() {
// -----
func @invariant_code_inside_affine_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
@@ -103,7 +103,7 @@ func @invariant_code_inside_affine_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply #map{{[0-9]*}}(%arg0)
@@ -119,7 +119,7 @@ func @invariant_code_inside_affine_if() {
// -----
func @dependent_stores() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -132,7 +132,7 @@ func @dependent_stores() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -149,7 +149,7 @@ func @dependent_stores() {
// -----
func @independent_stores() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -162,7 +162,7 @@ func @independent_stores() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -179,7 +179,7 @@ func @independent_stores() {
// -----
func @load_dependent_store() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -192,7 +192,7 @@ func @load_dependent_store() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -208,7 +208,7 @@ func @load_dependent_store() {
// -----
func @load_after_load() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -221,7 +221,7 @@ func @load_after_load() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -237,7 +237,7 @@ func @load_after_load() {
// -----
func @invariant_affine_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -249,7 +249,7 @@ func @invariant_affine_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.if #set(%arg0, %arg0) {
@@ -264,7 +264,7 @@ func @invariant_affine_if() {
// -----
func @invariant_affine_if2() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -276,7 +276,7 @@ func @invariant_affine_if2() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.for %arg1 = 0 to 10 {
@@ -293,7 +293,7 @@ func @invariant_affine_if2() {
// -----
func @invariant_affine_nested_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -307,7 +307,7 @@ func @invariant_affine_nested_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.for %arg1 = 0 to 10 {
@@ -327,7 +327,7 @@ func @invariant_affine_nested_if() {
// -----
func @invariant_affine_nested_if_else() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -343,7 +343,7 @@ func @invariant_affine_nested_if_else() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.for %arg1 = 0 to 10 {
@@ -365,8 +365,8 @@ func @invariant_affine_nested_if_else() {
// -----
func @invariant_affine_nested_if_else2() {
- %m = alloc() : memref<10xf32>
- %m2 = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
+ %m2 = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -382,8 +382,8 @@ func @invariant_affine_nested_if_else2() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
- // CHECK-NEXT: %1 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
+ // CHECK-NEXT: %1 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.if #set(%arg0, %arg0) {
@@ -403,7 +403,7 @@ func @invariant_affine_nested_if_else2() {
// -----
func @invariant_affine_nested_if2() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -417,7 +417,7 @@ func @invariant_affine_nested_if2() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.if #set(%arg0, %arg0) {
@@ -435,7 +435,7 @@ func @invariant_affine_nested_if2() {
// -----
func @invariant_affine_for_inside_affine_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -449,7 +449,7 @@ func @invariant_affine_for_inside_affine_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: affine.for %arg1 = 0 to 10 {
@@ -469,16 +469,16 @@ func @invariant_affine_for_inside_affine_if() {
// -----
func @invariant_constant_and_load() {
- %m = alloc() : memref<100xf32>
- %m2 = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
+ %m2 = memref.alloc() : memref<100xf32>
affine.for %arg0 = 0 to 5 {
%c0 = constant 0 : index
%v = affine.load %m2[%c0] : memref<100xf32>
affine.store %v, %m[%arg0] : memref<100xf32>
}
- // CHECK: %0 = alloc() : memref<100xf32>
- // CHECK-NEXT: %1 = alloc() : memref<100xf32>
+ // CHECK: %0 = memref.alloc() : memref<100xf32>
+ // CHECK-NEXT: %1 = memref.alloc() : memref<100xf32>
// CHECK-NEXT: %c0 = constant 0 : index
// CHECK-NEXT: %2 = affine.load %1[%c0] : memref<100xf32>
// CHECK-NEXT: affine.for %arg0 = 0 to 5 {
@@ -491,7 +491,7 @@ func @invariant_constant_and_load() {
// -----
func @nested_load_store_same_memref() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cst = constant 8.0 : f32
%c0 = constant 0 : index
affine.for %arg0 = 0 to 10 {
@@ -501,7 +501,7 @@ func @nested_load_store_same_memref() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: %c0 = constant 0 : index
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
@@ -516,7 +516,7 @@ func @nested_load_store_same_memref() {
// -----
func @nested_load_store_same_memref2() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cst = constant 8.0 : f32
%c0 = constant 0 : index
affine.for %arg0 = 0 to 10 {
@@ -526,7 +526,7 @@ func @nested_load_store_same_memref2() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: %c0 = constant 0 : index
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
@@ -541,7 +541,7 @@ func @nested_load_store_same_memref2() {
// CHECK-LABEL: func @do_not_hoist_dependent_side_effect_free_op
func @do_not_hoist_dependent_side_effect_free_op(%arg0: memref<10x512xf32>) {
- %0 = alloca() : memref<1xf32>
+ %0 = memref.alloca() : memref<1xf32>
%cst = constant 8.0 : f32
affine.for %i = 0 to 512 {
affine.for %j = 0 to 10 {
@@ -571,8 +571,8 @@ func @do_not_hoist_dependent_side_effect_free_op(%arg0: memref<10x512xf32>) {
// CHECK-LABEL: func @vector_loop_nothing_invariant
func @vector_loop_nothing_invariant() {
- %m1 = alloc() : memref<40xf32>
- %m2 = alloc() : memref<40xf32>
+ %m1 = memref.alloc() : memref<40xf32>
+ %m2 = memref.alloc() : memref<40xf32>
affine.for %arg0 = 0 to 10 {
%v0 = affine.vector_load %m1[%arg0*4] : memref<40xf32>, vector<4xf32>
%v1 = affine.vector_load %m2[%arg0*4] : memref<40xf32>, vector<4xf32>
@@ -593,9 +593,9 @@ func @vector_loop_nothing_invariant() {
// CHECK-LABEL: func @vector_loop_all_invariant
func @vector_loop_all_invariant() {
- %m1 = alloc() : memref<4xf32>
- %m2 = alloc() : memref<4xf32>
- %m3 = alloc() : memref<4xf32>
+ %m1 = memref.alloc() : memref<4xf32>
+ %m2 = memref.alloc() : memref<4xf32>
+ %m3 = memref.alloc() : memref<4xf32>
affine.for %arg0 = 0 to 10 {
%v0 = affine.vector_load %m1[0] : memref<4xf32>, vector<4xf32>
%v1 = affine.vector_load %m2[0] : memref<4xf32>, vector<4xf32>
@@ -605,9 +605,9 @@ func @vector_loop_all_invariant() {
return
}
-// CHECK: alloc()
-// CHECK-NEXT: alloc()
-// CHECK-NEXT: alloc()
+// CHECK: memref.alloc()
+// CHECK-NEXT: memref.alloc()
+// CHECK-NEXT: memref.alloc()
// CHECK-NEXT: affine.vector_load
// CHECK-NEXT: affine.vector_load
// CHECK-NEXT: addf
diff --git a/mlir/test/Dialect/Affine/affine-loop-normalize.mlir b/mlir/test/Dialect/Affine/affine-loop-normalize.mlir
index e8ad3ed6aae5..e86615786b07 100644
--- a/mlir/test/Dialect/Affine/affine-loop-normalize.mlir
+++ b/mlir/test/Dialect/Affine/affine-loop-normalize.mlir
@@ -9,7 +9,7 @@
// CHECK-LABEL: func @normalize_parallel()
func @normalize_parallel() {
%cst = constant 1.0 : f32
- %0 = alloc() : memref<2x4xf32>
+ %0 = memref.alloc() : memref<2x4xf32>
// CHECK: affine.parallel (%[[i0:.*]], %[[j0:.*]]) = (0, 0) to (4, 2)
affine.parallel (%i, %j) = (0, 1) to (10, 5) step (3, 2) {
// CHECK: %[[i1:.*]] = affine.apply [[$MAP0]](%[[i0]])
@@ -77,7 +77,7 @@ func @simple_loop_nest(){
// CHECK-LABEL: func @loop_with_unknown_upper_bound
// CHECK-SAME: (%[[ARG0:.*]]: memref<?x?xf32>, %[[ARG1:.*]]: index)
// CHECK-NEXT: %{{.*}} = constant 0 : index
-// CHECK-NEXT: %[[DIM:.*]] = dim %arg0, %c0 : memref<?x?xf32>
+// CHECK-NEXT: %[[DIM:.*]] = memref.dim %arg0, %c0 : memref<?x?xf32>
// CHECK-NEXT: affine.for %[[I:.*]] = 0 to [[$UB00]]()[%[[DIM]]] {
// CHECK-NEXT: %[[IIV:.*]] = affine.apply [[$IV00]](%[[I]])
// CHECK-NEXT: affine.for %[[II:.*]] = 0 to [[$UB11]]()[%[[ARG1]]] {
@@ -89,7 +89,7 @@ func @simple_loop_nest(){
// CHECK-NEXT: }
func @loop_with_unknown_upper_bound(%arg0: memref<?x?xf32>, %arg1: index) {
%c0 = constant 0 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32>
affine.for %i0 = 2 to %0 step 32 {
affine.for %i1 = 0 to %arg1 step 2 {
"test.foo"(%i0, %i1) : (index, index) -> ()
@@ -108,7 +108,7 @@ func @loop_with_unknown_upper_bound(%arg0: memref<?x?xf32>, %arg1: index) {
// CHECK-LABEL: func @loop_with_multiple_upper_bounds
// CHECK-SAME: (%[[ARG0:.*]]: memref<?x?xf32>, %[[ARG1:.*]]: index)
// CHECK-NEXT: %{{.*}} = constant 0 : index
-// CHECK-NEXT: %[[DIM:.*]] = dim %arg0, %c0 : memref<?x?xf32>
+// CHECK-NEXT: %[[DIM:.*]] = memref.dim %arg0, %c0 : memref<?x?xf32>
// CHECK-NEXT: affine.for %[[I:.*]] = 0 to [[$OUTERUB]]()[%[[DIM]]] {
// CHECK-NEXT: %[[IIV:.*]] = affine.apply [[$OUTERIV]](%[[I]])
// CHECK-NEXT: affine.for %[[II:.*]] = 0 to min [[$INNERUB]](%[[ARG1]]) {
@@ -120,7 +120,7 @@ func @loop_with_unknown_upper_bound(%arg0: memref<?x?xf32>, %arg1: index) {
// CHECK-NEXT: }
func @loop_with_multiple_upper_bounds(%arg0: memref<?x?xf32>, %arg1 : index) {
%c0 = constant 0 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32>
affine.for %i0 = 2 to %0 step 32{
affine.for %i1 = 2 to min affine_map<(d0)[] -> (d0, 512)>(%arg1) {
"test.foo"(%i0, %i1) : (index, index) -> ()
@@ -140,9 +140,9 @@ func @loop_with_multiple_upper_bounds(%arg0: memref<?x?xf32>, %arg1 : index) {
// CHECK-SAME: (%[[ARG0:.*]]: memref<1024x1024xf32>, %[[ARG1:.*]]: memref<1024x1024xf32>, %[[ARG2:.*]]: memref<1024x1024xf32>)
// CHECK-NEXT: %{{.*}} = constant 0 : index
// CHECK-NEXT: %{{.*}} = constant 1 : index
-// CHECK-NEXT: %[[DIM0:.*]] = dim %[[ARG0]], %{{.*}}
-// CHECK-NEXT: %[[DIM1:.*]] = dim %[[ARG1]], %{{.*}}
-// CHECK-NEXT: %[[DIM2:.*]] = dim %[[ARG0]], %{{.*}}
+// CHECK-NEXT: %[[DIM0:.*]] = memref.dim %[[ARG0]], %{{.*}}
+// CHECK-NEXT: %[[DIM1:.*]] = memref.dim %[[ARG1]], %{{.*}}
+// CHECK-NEXT: %[[DIM2:.*]] = memref.dim %[[ARG0]], %{{.*}}
// CHECK-NEXT: affine.for %[[I:.*]] = 0 to [[$INTERUB]]()[%[[DIM0]]] {
// CHECK-NEXT: %[[IIV:.*]] = affine.apply [[$INTERIV]](%[[I]])
// CHECK-NEXT: affine.for %[[J:.*]] = 0 to [[$INTERUB]]()[%[[DIM1]]] {
@@ -178,9 +178,9 @@ func @loop_with_multiple_upper_bounds(%arg0: memref<?x?xf32>, %arg1 : index) {
func @tiled_matmul(%0: memref<1024x1024xf32>, %1: memref<1024x1024xf32>, %2: memref<1024x1024xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %3 = dim %0, %c0 : memref<1024x1024xf32>
- %4 = dim %1, %c1 : memref<1024x1024xf32>
- %5 = dim %0, %c1 : memref<1024x1024xf32>
+ %3 = memref.dim %0, %c0 : memref<1024x1024xf32>
+ %4 = memref.dim %1, %c1 : memref<1024x1024xf32>
+ %5 = memref.dim %0, %c1 : memref<1024x1024xf32>
affine.for %arg0 = 0 to %3 step 32 {
affine.for %arg1 = 0 to %4 step 32 {
affine.for %arg2 = 0 to %5 step 32 {
diff --git a/mlir/test/Dialect/Affine/canonicalize.mlir b/mlir/test/Dialect/Affine/canonicalize.mlir
index 352066adbbd8..91547420015f 100644
--- a/mlir/test/Dialect/Affine/canonicalize.mlir
+++ b/mlir/test/Dialect/Affine/canonicalize.mlir
@@ -7,7 +7,7 @@
// CHECK-LABEL: func @compose_affine_maps_1dto2d_no_symbols() {
func @compose_affine_maps_1dto2d_no_symbols() {
- %0 = alloc() : memref<4x4xf32>
+ %0 = memref.alloc() : memref<4x4xf32>
affine.for %i0 = 0 to 15 {
// Test load[%x, %x]
@@ -17,8 +17,8 @@ func @compose_affine_maps_1dto2d_no_symbols() {
%x1_1 = affine.apply affine_map<(d0, d1) -> (d1)> (%x0, %x0)
// CHECK: %[[I0A:.*]] = affine.apply #[[$MAP0]](%{{.*}})
- // CHECK-NEXT: %[[V0:.*]] = load %0[%[[I0A]], %[[I0A]]]
- %v0 = load %0[%x1_0, %x1_1] : memref<4x4xf32>
+ // CHECK-NEXT: %[[V0:.*]] = memref.load %0[%[[I0A]], %[[I0A]]]
+ %v0 = memref.load %0[%x1_0, %x1_1] : memref<4x4xf32>
// Test store[%y, %y]
%y0 = affine.apply affine_map<(d0) -> (d0 + 1)> (%i0)
@@ -26,21 +26,21 @@ func @compose_affine_maps_1dto2d_no_symbols() {
%y1_1 = affine.apply affine_map<(d0, d1) -> (d1)> (%y0, %y0)
// CHECK-NEXT: %[[I1A:.*]] = affine.apply #[[$MAP1]](%{{.*}})
- // CHECK-NEXT: store %[[V0]], %0[%[[I1A]], %[[I1A]]]
- store %v0, %0[%y1_0, %y1_1] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %0[%[[I1A]], %[[I1A]]]
+ memref.store %v0, %0[%y1_0, %y1_1] : memref<4x4xf32>
// Test store[%x, %y]
%xy_0 = affine.apply affine_map<(d0, d1) -> (d0)> (%x0, %y0)
%xy_1 = affine.apply affine_map<(d0, d1) -> (d1)> (%x0, %y0)
- // CHECK-NEXT: store %[[V0]], %0[%[[I0A]], %[[I1A]]]
- store %v0, %0[%xy_0, %xy_1] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %0[%[[I0A]], %[[I1A]]]
+ memref.store %v0, %0[%xy_0, %xy_1] : memref<4x4xf32>
// Test store[%y, %x]
%yx_0 = affine.apply affine_map<(d0, d1) -> (d0)> (%y0, %x0)
%yx_1 = affine.apply affine_map<(d0, d1) -> (d1)> (%y0, %x0)
- // CHECK-NEXT: store %[[V0]], %0[%[[I1A]], %[[I0A]]]
- store %v0, %0[%yx_0, %yx_1] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %0[%[[I1A]], %[[I0A]]]
+ memref.store %v0, %0[%yx_0, %yx_1] : memref<4x4xf32>
}
return
}
@@ -53,7 +53,7 @@ func @compose_affine_maps_1dto2d_no_symbols() {
// CHECK-LABEL: func @compose_affine_maps_1dto2d_with_symbols() {
func @compose_affine_maps_1dto2d_with_symbols() {
- %0 = alloc() : memref<4x4xf32>
+ %0 = memref.alloc() : memref<4x4xf32>
affine.for %i0 = 0 to 15 {
// Test load[%x0, %x0] with symbol %c4
@@ -61,29 +61,29 @@ func @compose_affine_maps_1dto2d_with_symbols() {
%x0 = affine.apply affine_map<(d0)[s0] -> (d0 - s0)> (%i0)[%c4]
// CHECK: %[[I0:.*]] = affine.apply #[[$MAP4]](%{{.*}})
- // CHECK-NEXT: %[[V0:.*]] = load %{{.*}}[%[[I0]], %[[I0]]]
- %v0 = load %0[%x0, %x0] : memref<4x4xf32>
+ // CHECK-NEXT: %[[V0:.*]] = memref.load %{{.*}}[%[[I0]], %[[I0]]]
+ %v0 = memref.load %0[%x0, %x0] : memref<4x4xf32>
// Test load[%x0, %x1] with symbol %c4 captured by '%x0' map.
%x1 = affine.apply affine_map<(d0) -> (d0 + 1)> (%i0)
%y1 = affine.apply affine_map<(d0, d1) -> (d0+d1)> (%x0, %x1)
// CHECK-NEXT: %[[I1:.*]] = affine.apply #[[$MAP7]](%{{.*}})
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I1]], %[[I1]]]
- store %v0, %0[%y1, %y1] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I1]], %[[I1]]]
+ memref.store %v0, %0[%y1, %y1] : memref<4x4xf32>
// Test store[%x1, %x0] with symbol %c4 captured by '%x0' map.
%y2 = affine.apply affine_map<(d0, d1) -> (d0 + d1)> (%x1, %x0)
// CHECK-NEXT: %[[I2:.*]] = affine.apply #[[$MAP7]](%{{.*}})
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I2]], %[[I2]]]
- store %v0, %0[%y2, %y2] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I2]], %[[I2]]]
+ memref.store %v0, %0[%y2, %y2] : memref<4x4xf32>
// Test store[%x2, %x0] with symbol %c4 from '%x0' and %c5 from '%x2'
%c5 = constant 5 : index
%x2 = affine.apply affine_map<(d0)[s0] -> (d0 + s0)> (%i0)[%c5]
%y3 = affine.apply affine_map<(d0, d1) -> (d0 + d1)> (%x2, %x0)
// CHECK: %[[I3:.*]] = affine.apply #[[$MAP7a]](%{{.*}})
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I3]], %[[I3]]]
- store %v0, %0[%y3, %y3] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I3]], %[[I3]]]
+ memref.store %v0, %0[%y3, %y3] : memref<4x4xf32>
}
return
}
@@ -95,8 +95,8 @@ func @compose_affine_maps_1dto2d_with_symbols() {
// CHECK-LABEL: func @compose_affine_maps_2d_tile() {
func @compose_affine_maps_2d_tile() {
- %0 = alloc() : memref<16x32xf32>
- %1 = alloc() : memref<16x32xf32>
+ %0 = memref.alloc() : memref<16x32xf32>
+ %1 = memref.alloc() : memref<16x32xf32>
%c4 = constant 4 : index
%c8 = constant 8 : index
@@ -116,11 +116,11 @@ func @compose_affine_maps_2d_tile() {
((d1 * s1) + d3)> (%x0, %x1, %x2, %x3)[%c4, %c8]
// CHECK: %[[I0:.*]] = affine.apply #[[$MAP8]](%{{.*}}, %{{.*}})
// CHECK: %[[I1:.*]] = affine.apply #[[$MAP8a]](%{{.*}}, %{{.*}})
- // CHECK-NEXT: %[[L0:.*]] = load %{{.*}}[%[[I0]], %[[I1]]]
- %v0 = load %0[%x40, %x41] : memref<16x32xf32>
+ // CHECK-NEXT: %[[L0:.*]] = memref.load %{{.*}}[%[[I0]], %[[I1]]]
+ %v0 = memref.load %0[%x40, %x41] : memref<16x32xf32>
- // CHECK-NEXT: store %[[L0]], %{{.*}}[%[[I0]], %[[I1]]]
- store %v0, %1[%x40, %x41] : memref<16x32xf32>
+ // CHECK-NEXT: memref.store %[[L0]], %{{.*}}[%[[I0]], %[[I1]]]
+ memref.store %v0, %1[%x40, %x41] : memref<16x32xf32>
}
}
}
@@ -138,8 +138,8 @@ func @compose_affine_maps_2d_tile() {
// CHECK-LABEL: func @compose_affine_maps_dependent_loads() {
func @compose_affine_maps_dependent_loads() {
- %0 = alloc() : memref<16x32xf32>
- %1 = alloc() : memref<16x32xf32>
+ %0 = memref.alloc() : memref<16x32xf32>
+ %1 = memref.alloc() : memref<16x32xf32>
affine.for %i0 = 0 to 3 {
affine.for %i1 = 0 to 3 {
@@ -157,15 +157,15 @@ func @compose_affine_maps_dependent_loads() {
// CHECK: %[[I0:.*]] = affine.apply #[[$MAP9]](%{{.*}})
// CHECK: %[[I1:.*]] = affine.apply #[[$MAP4b]](%{{.*}})
// CHECK: %[[I2:.*]] = affine.apply #[[$MAP10]](%{{.*}})
- // CHECK-NEXT: %[[V0:.*]] = load %{{.*}}[%[[I0]], %[[I1]]]
- %v0 = load %0[%x00, %x01] : memref<16x32xf32>
+ // CHECK-NEXT: %[[V0:.*]] = memref.load %{{.*}}[%[[I0]], %[[I1]]]
+ %v0 = memref.load %0[%x00, %x01] : memref<16x32xf32>
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I0]], %[[I2]]]
- store %v0, %0[%x00, %x02] : memref<16x32xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I0]], %[[I2]]]
+ memref.store %v0, %0[%x00, %x02] : memref<16x32xf32>
// Swizzle %i0, %i1
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I1]], %[[I0]]]
- store %v0, %0[%x01, %x00] : memref<16x32xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I1]], %[[I0]]]
+ memref.store %v0, %0[%x01, %x00] : memref<16x32xf32>
// Swizzle %x00, %x01 and %c3, %c7
%x10 = affine.apply affine_map<(d0, d1)[s0, s1] -> (d0 * s1)>
@@ -175,8 +175,8 @@ func @compose_affine_maps_dependent_loads() {
// CHECK-NEXT: %[[I2A:.*]] = affine.apply #[[$MAP12]](%{{.*}})
// CHECK-NEXT: %[[I2B:.*]] = affine.apply #[[$MAP11]](%{{.*}})
- // CHECK-NEXT: store %[[V0]], %{{.*}}[%[[I2A]], %[[I2B]]]
- store %v0, %0[%x10, %x11] : memref<16x32xf32>
+ // CHECK-NEXT: memref.store %[[V0]], %{{.*}}[%[[I2A]], %[[I2B]]]
+ memref.store %v0, %0[%x10, %x11] : memref<16x32xf32>
}
}
}
@@ -198,8 +198,8 @@ func @compose_affine_maps_diamond_dependency(%arg0: f32, %arg1: memref<4x4xf32>)
%d1 = affine.apply affine_map<(d0, d1) -> (d1 floordiv 3)> (%b, %c)
// CHECK: %[[I0:.*]] = affine.apply #[[$MAP13A]](%{{.*}})
// CHECK: %[[I1:.*]] = affine.apply #[[$MAP13B]](%{{.*}})
- // CHECK-NEXT: store %arg0, %arg1[%[[I0]], %[[I1]]]
- store %arg0, %arg1[%d0, %d1] : memref<4x4xf32>
+ // CHECK-NEXT: memref.store %arg0, %arg1[%[[I0]], %[[I1]]]
+ memref.store %arg0, %arg1[%d0, %d1] : memref<4x4xf32>
}
return
@@ -223,16 +223,16 @@ func @compose_affine_maps_multiple_symbols(%arg0: index, %arg1: index) -> index
// CHECK-LABEL: func @arg_used_as_dim_and_symbol
func @arg_used_as_dim_and_symbol(%arg0: memref<100x100xf32>, %arg1: index, %arg2: f32) {
%c9 = constant 9 : index
- %1 = alloc() : memref<100x100xf32, 1>
- %2 = alloc() : memref<1xi32>
+ %1 = memref.alloc() : memref<100x100xf32, 1>
+ %2 = memref.alloc() : memref<1xi32>
affine.for %i0 = 0 to 100 {
affine.for %i1 = 0 to 100 {
%3 = affine.apply affine_map<(d0, d1)[s0, s1] -> (d1 + s0 + s1)>
(%i0, %i1)[%arg1, %c9]
%4 = affine.apply affine_map<(d0, d1, d3) -> (d3 - (d0 + d1))>
(%arg1, %c9, %3)
- // CHECK: store %arg2, %{{.*}}[%{{.*}}, %{{.*}}]
- store %arg2, %1[%4, %arg1] : memref<100x100xf32, 1>
+ // CHECK: memref.store %arg2, %{{.*}}[%{{.*}}, %{{.*}}]
+ memref.store %arg2, %1[%4, %arg1] : memref<100x100xf32, 1>
}
}
return
@@ -244,17 +244,17 @@ func @arg_used_as_dim_and_symbol(%arg0: memref<100x100xf32>, %arg1: index, %arg2
func @trivial_maps() {
// CHECK-NOT: affine.apply
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cst = constant 0.000000e+00 : f32
affine.for %i1 = 0 to 10 {
%1 = affine.apply affine_map<()[s0] -> (s0)>()[%c0]
- store %cst, %0[%1] : memref<10xf32>
- %2 = load %0[%c0] : memref<10xf32>
+ memref.store %cst, %0[%1] : memref<10xf32>
+ %2 = memref.load %0[%c0] : memref<10xf32>
%3 = affine.apply affine_map<()[] -> (0)>()[]
- store %cst, %0[%3] : memref<10xf32>
- store %2, %0[%c0] : memref<10xf32>
+ memref.store %cst, %0[%3] : memref<10xf32>
+ memref.store %2, %0[%c0] : memref<10xf32>
}
return
}
@@ -422,7 +422,7 @@ func @symbolic_semi_affine(%M: index, %N: index, %A: memref<?xf32>) {
%1 = affine.apply affine_map<()[s0] -> (s0 + 1)> ()[%M]
%2 = affine.apply affine_map<(d0)[s0] -> (d0 floordiv s0)> (%i0)[%1]
// CHECK-DAG: {{.*}} = affine.apply #[[$symbolic_semi_affine]](%{{.*}})[%{{.*}}]
- store %f1, %A[%2] : memref<?xf32>
+ memref.store %f1, %A[%2] : memref<?xf32>
}
return
}
@@ -667,7 +667,7 @@ func @affine_parallel_const_bounds() {
%cst = constant 1.0 : f32
%c0 = constant 0 : index
%c4 = constant 4 : index
- %0 = alloc() : memref<4xf32>
+ %0 = memref.alloc() : memref<4xf32>
// CHECK: affine.parallel (%{{.*}}) = (0) to (4)
affine.parallel (%i) = (%c0) to (%c0 + %c4) {
%1 = affine.apply #map3(%i)
@@ -686,11 +686,11 @@ func @compose_affine_maps_div_symbol(%A : memref<i64>, %i0 : index, %i1 : index)
%1 = affine.apply affine_map<()[s0] -> (3 * s0)> ()[%i0]
%2 = affine.apply affine_map<(d0)[s0, s1] -> (d0 mod s1 + s0 * s1 + s0 * 4)> (%i1)[%0, %1]
%3 = index_cast %2: index to i64
- store %3, %A[]: memref<i64>
+ memref.store %3, %A[]: memref<i64>
affine.for %i2 = 0 to 3 {
%4 = affine.apply affine_map<(d0)[s0, s1] -> (d0 ceildiv s1 + s0 + s0 * 3)> (%i2)[%0, %1]
%5 = index_cast %4: index to i64
- store %5, %A[]: memref<i64>
+ memref.store %5, %A[]: memref<i64>
}
return
}
diff --git a/mlir/test/Dialect/Affine/dma-generate.mlir b/mlir/test/Dialect/Affine/dma-generate.mlir
index 19885b317a24..be4b1dc68c71 100644
--- a/mlir/test/Dialect/Affine/dma-generate.mlir
+++ b/mlir/test/Dialect/Affine/dma-generate.mlir
@@ -14,21 +14,21 @@
// CHECK-LABEL: func @loop_nest_1d() {
func @loop_nest_1d() {
- %A = alloc() : memref<256 x f32>
- %B = alloc() : memref<512 x f32>
- %F = alloc() : memref<256 x f32, 2>
+ %A = memref.alloc() : memref<256 x f32>
+ %B = memref.alloc() : memref<512 x f32>
+ %F = memref.alloc() : memref<256 x f32, 2>
// First DMA buffer.
- // CHECK: alloc() : memref<256xf32>
- // CHECK: alloc() : memref<256xf32, 2>
+ // CHECK: memref.alloc() : memref<256xf32>
+ // CHECK: memref.alloc() : memref<256xf32, 2>
// Tag for first DMA.
- // CHECK: alloc() : memref<1xi32>
+ // CHECK: memref.alloc() : memref<1xi32>
// First DMA transfer.
// CHECK: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32>, memref<256xf32, 2>, memref<1xi32>
// CHECK: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
// Second DMA buffer.
- // CHECK: alloc() : memref<256xf32, 2>
+ // CHECK: memref.alloc() : memref<256xf32, 2>
// Tag for second DMA.
- // CHECK: alloc() : memref<1xi32>
+ // CHECK: memref.alloc() : memref<1xi32>
// Second DMA transfer.
// CHECK: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<512xf32>, memref<256xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
@@ -58,13 +58,13 @@ func @loop_nest_1d() {
// CHECK-LABEL: func @loop_nest_high_d
// CHECK: %{{.*}} = constant 16384 : index
-// CHECK-DAG: [[BUFB:%[0-9]+]] = alloc() : memref<512x32xf32, 2>
-// CHECK-DAG: [[BUFA:%[0-9]+]] = alloc() : memref<512x32xf32, 2>
-// CHECK-DAG: [[BUFC:%[0-9]+]] = alloc() : memref<512x32xf32, 2>
-// CHECK-DAG: [[TAGB:%[0-9]+]] = alloc() : memref<1xi32>
-// CHECK-DAG: [[TAGA:%[0-9]+]] = alloc() : memref<1xi32>
-// CHECK-DAG: [[TAGC:%[0-9]+]] = alloc() : memref<1xi32>
-// CHECK-DAG: [[TAGC_W:%[0-9]+]] = alloc() : memref<1xi32>
+// CHECK-DAG: [[BUFB:%[0-9]+]] = memref.alloc() : memref<512x32xf32, 2>
+// CHECK-DAG: [[BUFA:%[0-9]+]] = memref.alloc() : memref<512x32xf32, 2>
+// CHECK-DAG: [[BUFC:%[0-9]+]] = memref.alloc() : memref<512x32xf32, 2>
+// CHECK-DAG: [[TAGB:%[0-9]+]] = memref.alloc() : memref<1xi32>
+// CHECK-DAG: [[TAGA:%[0-9]+]] = memref.alloc() : memref<1xi32>
+// CHECK-DAG: [[TAGC:%[0-9]+]] = memref.alloc() : memref<1xi32>
+// CHECK-DAG: [[TAGC_W:%[0-9]+]] = memref.alloc() : memref<1xi32>
// INCOMING DMA for B
// CHECK-DAG: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], [[BUFB]][%{{.*}}, %{{.*}}], [[TAGB]][%{{.*}}], %{{.*}} : memref<512x32xf32>, memref<512x32xf32, 2>, memref<1xi32>
// CHECK-DAG: affine.dma_wait [[TAGB]][%{{.*}}], %{{.*}} : memref<1xi32>
@@ -145,10 +145,10 @@ func @loop_nest_high_d(%A: memref<512 x 32 x f32>,
// region within a 256 x 8 memref.
//
// CHECK-LABEL: func @loop_nest_modulo() {
-// CHECK: alloc() : memref<256x8xf32>
+// CHECK: memref.alloc() : memref<256x8xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 32 step 4 {
-// CHECK: alloc() : memref<1x2xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<1x2xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// Composition of the affine map for '%{{.*}}' causes '%{{.*}}' to be added as a symbol.
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}, 0], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256x8xf32>, memref<1x2xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
@@ -161,7 +161,7 @@ func @loop_nest_high_d(%A: memref<512 x 32 x f32>,
// CHECK-NEXT: }
// CHECK-NEXT: return
func @loop_nest_modulo() {
- %A = alloc() : memref<256 x 8 x f32>
+ %A = memref.alloc() : memref<256 x 8 x f32>
affine.for %i = 0 to 32 step 4 {
// DMAs will be performed at this level (%j is the first unit stride loop)
affine.for %j = 0 to 8 {
@@ -179,11 +179,11 @@ func @loop_nest_modulo() {
// dependent on outer loop IVs.
// CHECK-LABEL: func @loop_nest_tiled() -> memref<256x1024xf32> {
func @loop_nest_tiled() -> memref<256x1024xf32> {
- %0 = alloc() : memref<256x1024xf32>
+ %0 = memref.alloc() : memref<256x1024xf32>
affine.for %i0 = 0 to 256 step 32 {
affine.for %i1 = 0 to 1024 step 32 {
-// CHECK: alloc() : memref<32x32xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<32x32xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// Strided DMA here: 32 x 32 tile in a 256 x 1024 memref.
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}, %{{.*}}, %{{.*}} : memref<256x1024xf32>, memref<32x32xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait
@@ -206,8 +206,8 @@ func @loop_nest_tiled() -> memref<256x1024xf32> {
func @dma_constant_dim_access(%A : memref<100x100xf32>) {
%one = constant 1 : index
%N = constant 100 : index
- // CHECK: alloc() : memref<1x100xf32, 2>
- // CHECK-NEXT: alloc() : memref<1xi32>
+ // CHECK: memref.alloc() : memref<1x100xf32, 2>
+ // CHECK-NEXT: memref.alloc() : memref<1xi32>
// No strided DMA needed here.
// CHECK: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<100x100xf32>, memref<1x100xf32, 2>,
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
@@ -232,8 +232,8 @@ func @dma_with_symbolic_accesses(%A : memref<100x100xf32>, %M : index) {
}
}
return
-// CHECK: alloc() : memref<100x100xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<100x100xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[0, symbol(%{{.*}}) + 9], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}}
// CHECK-NEXT: affine.for %[[IV0:.*]] = 0 to 100 {
@@ -251,8 +251,8 @@ func @dma_with_symbolic_loop_bounds(%A : memref<100x100xf32>, %M : index, %N: in
%K = constant 9 : index
// The buffer size can't be bound by a constant smaller than the original
// memref size; so the DMA buffer is the entire 100x100.
-// CHECK: alloc() : memref<100x100xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<100x100xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<100x100xf32>, memref<100x100xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
affine.for %i = 0 to 100 {
@@ -269,8 +269,8 @@ func @dma_with_symbolic_loop_bounds(%A : memref<100x100xf32>, %M : index, %N: in
// CHECK-LABEL: func @dma_unknown_size
func @dma_unknown_size(%arg0: memref<?x?xf32>) {
%c0 = constant 0 : index
- %M = dim %arg0, %c0 : memref<? x ? x f32>
- %N = dim %arg0, %c0 : memref<? x ? x f32>
+ %M = memref.dim %arg0, %c0 : memref<? x ? x f32>
+ %N = memref.dim %arg0, %c0 : memref<? x ? x f32>
affine.for %i = 0 to %M {
affine.for %j = 0 to %N {
// If this loop nest isn't tiled, the access requires a non-constant DMA
@@ -315,7 +315,7 @@ func @dma_memref_3d(%arg0: memref<1024x1024x1024xf32>) {
// CHECK-LABEL: func @multi_load_store_union() {
func @multi_load_store_union() {
- %A = alloc() : memref<512 x 512 x f32>
+ %A = memref.alloc() : memref<512 x 512 x f32>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 256 {
%idx = affine.apply affine_map<(d0) -> (d0 + 64)>(%i)
@@ -335,12 +335,12 @@ func @multi_load_store_union() {
}
return
}
-// CHECK: alloc() : memref<512x512xf32>
-// CHECK-NEXT: alloc() : memref<382x446xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<512x512xf32>
+// CHECK-NEXT: memref.alloc() : memref<382x446xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}, %{{.*}}, %{{.*}} : memref<512x512xf32>, memref<382x446xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 256 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 256 {
// CHECK: affine.load %{{.*}}[%{{.*}}, %{{.*}} + 126] : memref<382x446xf32, 2>
@@ -363,7 +363,7 @@ func @multi_load_store_union() {
func @dma_loop_straightline_interspersed() {
%c0 = constant 0 : index
%c255 = constant 255 : index
- %A = alloc() : memref<256 x f32>
+ %A = memref.alloc() : memref<256 x f32>
%v = affine.load %A[%c0] : memref<256 x f32>
affine.for %i = 1 to 255 {
affine.load %A[%i] : memref<256 x f32>
@@ -374,16 +374,16 @@ func @dma_loop_straightline_interspersed() {
}
// There are three regions here - the 'load' preceding the loop, the loop
// itself, and the operations appearing after the scf.
-// CHECK: alloc() : memref<256xf32>
-// CHECK-NEXT: alloc() : memref<1xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<256xf32>
+// CHECK-NEXT: memref.alloc() : memref<1xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32>, memref<1xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
// CHECK-NEXT: affine.load %{{.*}}[0] : memref<1xf32, 2>
// CHECK-NEXT: dealloc %{{.*}} : memref<1xi32>
// CHECK-NEXT: dealloc %{{.*}} : memref<1xf32, 2>
-// CHECK-NEXT: alloc() : memref<254xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK-NEXT: memref.alloc() : memref<254xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32>, memref<254xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 1 to 255 {
@@ -391,11 +391,11 @@ func @dma_loop_straightline_interspersed() {
// CHECK-NEXT: }
// CHECK-NEXT: dealloc %{{.*}} : memref<1xi32>
// CHECK-NEXT: dealloc %{{.*}} : memref<254xf32, 2>
-// CHECK-NEXT: alloc() : memref<256xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK-NEXT: memref.alloc() : memref<256xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32>, memref<256xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.load %{{.*}}[255] : memref<256xf32, 2>
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[0] : memref<256xf32, 2>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32, 2>, memref<256xf32>, memref<1xi32>
@@ -410,7 +410,7 @@ func @dma_loop_straightline_interspersed() {
// CHECK-LABEL: func @dma_mixed_loop_blocks() {
func @dma_mixed_loop_blocks() {
%c0 = constant 0 : index
- %A = alloc() : memref<256 x 256 x vector<8 x f32>>
+ %A = memref.alloc() : memref<256 x 256 x vector<8 x f32>>
affine.for %i = 0 to 256 {
%v = affine.load %A[%c0, %c0] : memref<256 x 256 x vector<8 x f32>>
"foo"(%v) : (vector<8 x f32>) -> ()
@@ -421,9 +421,9 @@ func @dma_mixed_loop_blocks() {
}
return
}
-// CHECK-DAG: [[MEM:%[0-9]+]] = alloc() : memref<256x256xvector<8xf32>>
-// CHECK-DAG: [[BUF:%[0-9]+]] = alloc() : memref<256x256xvector<8xf32>, 2>
-// CHECK-DAG: [[TAG:%[0-9]+]] = alloc() : memref<1xi32>
+// CHECK-DAG: [[MEM:%[0-9]+]] = memref.alloc() : memref<256x256xvector<8xf32>>
+// CHECK-DAG: [[BUF:%[0-9]+]] = memref.alloc() : memref<256x256xvector<8xf32>, 2>
+// CHECK-DAG: [[TAG:%[0-9]+]] = memref.alloc() : memref<1xi32>
// CHECK: affine.dma_start [[MEM]][%{{.*}}, %{{.*}}], [[BUF]][%{{.*}}, %{{.*}}], [[TAG]][%{{.*}}], %{{.*}} : memref<256x256xvector<8xf32>>, memref<256x256xvector<8xf32>, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait [[TAG]][%{{.*}}], %{{.*}} : memref<1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 256 {
@@ -443,8 +443,8 @@ func @relative_loop_bounds(%arg0: memref<1027xf32>) {
}
return
}
-// CHECK: [[BUF:%[0-9]+]] = alloc() : memref<1027xf32, 2>
-// CHECK-NEXT: [[MEM:%[0-9]+]] = alloc() : memref<1xi32>
+// CHECK: [[BUF:%[0-9]+]] = memref.alloc() : memref<1027xf32, 2>
+// CHECK-NEXT: [[MEM:%[0-9]+]] = memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 1024 {
// CHECK-NEXT: affine.for %[[I2:.*]] = {{#map[0-9]+}}(%{{.*}}) to {{#map[0-9]+}}(%{{.*}}) {
// CHECK: affine.store %{{.*}}, [[BUF]][%[[I2]]] : memref<1027xf32, 2>
@@ -456,7 +456,7 @@ func @relative_loop_bounds(%arg0: memref<1027xf32>) {
// -----
func @test_read_write_region_union() {
- %0 = alloc() : memref<256xf32>
+ %0 = memref.alloc() : memref<256xf32>
affine.for %i0 = 0 to 10 {
// memref dims: [0, 256)
// read region: [100, 110)
@@ -470,12 +470,12 @@ func @test_read_write_region_union() {
return
}
-// CHECK: alloc() : memref<256xf32>
-// CHECK-NEXT: alloc() : memref<85xf32, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<256xf32>
+// CHECK-NEXT: memref.alloc() : memref<85xf32, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<256xf32>, memref<85xf32, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 10 {
// CHECK: affine.load %{{.*}}[%{{.*}} + 75] : memref<85xf32, 2>
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<85xf32, 2>
@@ -493,8 +493,8 @@ func @test_read_write_region_union() {
// CHECK-LABEL: func @test_analysis_util
func @test_analysis_util(%arg0: memref<4x4x16x1xf32>, %arg1: memref<144x9xf32>, %arg2: memref<2xf32>) -> (memref<144x9xf32>, memref<2xf32>) {
%c0 = constant 0 : index
- %0 = alloc() : memref<64x1xf32>
- %1 = alloc() : memref<144x4xf32>
+ %0 = memref.alloc() : memref<64x1xf32>
+ %1 = memref.alloc() : memref<144x4xf32>
%2 = constant 0.0 : f32
affine.for %i8 = 0 to 9 step 3 {
affine.for %i9 = #map_lb(%i8) to #map_ub(%i8) {
@@ -510,7 +510,7 @@ func @test_analysis_util(%arg0: memref<4x4x16x1xf32>, %arg1: memref<144x9xf32>,
return %arg1, %arg2 : memref<144x9xf32>, memref<2xf32>
}
// CHECK: affine.for %{{.*}} = 0 to 9 step 3 {
-// CHECK: [[BUF:%[0-9]+]] = alloc() : memref<2xf32, 2>
+// CHECK: [[BUF:%[0-9]+]] = memref.alloc() : memref<2xf32, 2>
// CHECK: affine.dma_start %{{.*}}[%{{.*}} floordiv 8], [[BUF]]
// CHECK: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
// CHECK: affine.for %{{.*}} =
@@ -539,8 +539,8 @@ func @test_memref_bounds(%arg0: memref<4x4x16x1xvector<8x128xf32>>, %arg1: memre
return %arg1, %arg2 : memref<144x9xvector<8x128xf32>>, memref<2xvector<8x128xf32>>
}
-// CHECK: alloc() : memref<4x4x16x1xvector<8x128xf32>, 2>
-// CHECK-NEXT: alloc() : memref<1xi32>
+// CHECK: memref.alloc() : memref<4x4x16x1xvector<8x128xf32>, 2>
+// CHECK-NEXT: memref.alloc() : memref<1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}], %{{.*}}[%{{.*}}], %{{.*}} : memref<4x4x16x1xvector<8x128xf32>>, memref<4x4x16x1xvector<8x128xf32>, 2>, memref<1xi32>
// CHECK-NEXT: affine.dma_wait %{{.*}}[%{{.*}}], %{{.*}} : memref<1xi32>
@@ -553,7 +553,7 @@ func @test_memref_bounds(%arg0: memref<4x4x16x1xvector<8x128xf32>>, %arg1: memre
func @load_store_same_memref(%arg0: memref<256x1024xf32>) {
// FAST-MEM-16KB: affine.for %{{.*}} = 0 to 256 step 4
affine.for %i0 = 0 to 256 step 4 {
- // FAST-MEM-16KB: [[BUF:%[0-9]+]] = alloc() : memref<4x1024xf32, 2>
+ // FAST-MEM-16KB: [[BUF:%[0-9]+]] = memref.alloc() : memref<4x1024xf32, 2>
// FAST-MEM-16KB: affine.dma_start %{{.*}}
// FAST-MEM-16KB-NEXT: affine.dma_wait
// FAST-MEM-16KB: affine.for %{{.*}}
diff --git a/mlir/test/Dialect/Affine/dma.mlir b/mlir/test/Dialect/Affine/dma.mlir
index 6feb267596a1..3ac1d03e0f82 100644
--- a/mlir/test/Dialect/Affine/dma.mlir
+++ b/mlir/test/Dialect/Affine/dma.mlir
@@ -4,9 +4,9 @@
// Test with loop IVs.
func @test0(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
- %1 = alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100x100xf32>
+ %1 = memref.alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -25,9 +25,9 @@ func @test0(%arg0 : index, %arg1 : index) {
// Test with loop IVs and optional stride arguments.
func @test1(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
- %1 = alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100x100xf32>
+ %1 = memref.alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c0 = constant 0 : index
%c64 = constant 64 : index
%c128 = constant 128 : index
@@ -48,9 +48,9 @@ func @test1(%arg0 : index, %arg1 : index) {
// Test with loop IVs and symbols (without symbol keyword).
func @test2(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
- %1 = alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100x100xf32>
+ %1 = memref.alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -70,9 +70,9 @@ func @test2(%arg0 : index, %arg1 : index) {
// Test with loop IVs and symbols (with symbol keyword).
func @test3(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
- %1 = alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100x100xf32>
+ %1 = memref.alloc() : memref<100x100xf32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -93,9 +93,9 @@ func @test3(%arg0 : index, %arg1 : index) {
// Test with loop IVs, symbols and constants in nested affine expressions.
func @test4(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
- %1 = alloc() : memref<100x100xf32, 2>
- %2 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<100x100xf32>
+ %1 = memref.alloc() : memref<100x100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32>
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
diff --git a/mlir/test/Dialect/Affine/invalid.mlir b/mlir/test/Dialect/Affine/invalid.mlir
index 56562481aad8..0e464ac4976c 100644
--- a/mlir/test/Dialect/Affine/invalid.mlir
+++ b/mlir/test/Dialect/Affine/invalid.mlir
@@ -121,9 +121,9 @@ func @affine_if_invalid_sym() {
func @affine_if_invalid_dimop_dim(%arg0: index, %arg1: index, %arg2: index, %arg3: index) {
affine.for %n0 = 0 to 7 {
- %0 = alloc(%arg0, %arg1, %arg2, %arg3) : memref<?x?x?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1, %arg2, %arg3) : memref<?x?x?x?xf32>
%c0 = constant 0 : index
- %dim = dim %0, %c0 : memref<?x?x?x?xf32>
+ %dim = memref.dim %0, %c0 : memref<?x?x?x?xf32>
// expected-error at +1 {{operand cannot be used as a symbol}}
affine.if #set0(%dim)[%n0] {}
@@ -253,7 +253,7 @@ func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
// expected-error at +1 {{reduction must be specified for each output}}
%1 = affine.parallel (%i, %j) = (0, 0) to (100, 100) step (10, 10) -> (f32) {
%2 = affine.load %0[%i, %j] : memref<100x100xf32>
@@ -265,7 +265,7 @@ func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
// expected-error at +1 {{invalid reduction value: "bad"}}
%1 = affine.parallel (%i, %j) = (0, 0) to (100, 100) step (10, 10) reduce ("bad") -> (f32) {
%2 = affine.load %0[%i, %j] : memref<100x100xf32>
@@ -277,7 +277,7 @@ func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<100x100xi32>
+ %0 = memref.alloc() : memref<100x100xi32>
%1 = affine.parallel (%i, %j) = (0, 0) to (100, 100) step (10, 10) reduce ("minf") -> (f32) {
%2 = affine.load %0[%i, %j] : memref<100x100xi32>
// expected-error at +1 {{types mismatch between yield op and its parent}}
@@ -289,7 +289,7 @@ func @affine_parallel(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @vector_load_invalid_vector_type() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
affine.for %i0 = 0 to 16 step 8 {
// expected-error at +1 {{requires memref and vector types of the same elemental type}}
%1 = affine.vector_load %0[%i0] : memref<100xf32>, vector<8xf64>
@@ -300,7 +300,7 @@ func @vector_load_invalid_vector_type() {
// -----
func @vector_store_invalid_vector_type() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%1 = constant dense<7.0> : vector<8xf64>
affine.for %i0 = 0 to 16 step 8 {
// expected-error at +1 {{requires memref and vector types of the same elemental type}}
@@ -312,7 +312,7 @@ func @vector_store_invalid_vector_type() {
// -----
func @vector_load_vector_memref() {
- %0 = alloc() : memref<100xvector<8xf32>>
+ %0 = memref.alloc() : memref<100xvector<8xf32>>
affine.for %i0 = 0 to 4 {
// expected-error at +1 {{requires memref and vector types of the same elemental type}}
%1 = affine.vector_load %0[%i0] : memref<100xvector<8xf32>>, vector<8xf32>
@@ -323,7 +323,7 @@ func @vector_load_vector_memref() {
// -----
func @vector_store_vector_memref() {
- %0 = alloc() : memref<100xvector<8xf32>>
+ %0 = memref.alloc() : memref<100xvector<8xf32>>
%1 = constant dense<7.0> : vector<8xf32>
affine.for %i0 = 0 to 4 {
// expected-error at +1 {{requires memref and vector types of the same elemental type}}
diff --git a/mlir/test/Dialect/Affine/load-store-invalid.mlir b/mlir/test/Dialect/Affine/load-store-invalid.mlir
index 5bc31a2184f3..4c2948afc2cf 100644
--- a/mlir/test/Dialect/Affine/load-store-invalid.mlir
+++ b/mlir/test/Dialect/Affine/load-store-invalid.mlir
@@ -63,7 +63,7 @@ func @store_too_few_subscripts_map(%arg0: memref<?x?xf32>, %arg1: index, %val: f
// -----
func @load_non_affine_index(%arg0 : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
%1 = muli %i0, %arg0 : index
// expected-error at +1 {{op index must be a dimension or symbol identifier}}
@@ -75,7 +75,7 @@ func @load_non_affine_index(%arg0 : index) {
// -----
func @store_non_affine_index(%arg0 : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%1 = constant 11.0 : f32
affine.for %i0 = 0 to 10 {
%2 = muli %i0, %arg0 : index
@@ -88,7 +88,7 @@ func @store_non_affine_index(%arg0 : index) {
// -----
func @invalid_prefetch_rw(%i : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
// expected-error at +1 {{rw specifier has to be 'read' or 'write'}}
affine.prefetch %0[%i], rw, locality<0>, data : memref<10xf32>
return
@@ -97,7 +97,7 @@ func @invalid_prefetch_rw(%i : index) {
// -----
func @invalid_prefetch_cache_type(%i : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
// expected-error at +1 {{cache type has to be 'data' or 'instr'}}
affine.prefetch %0[%i], read, locality<0>, false : memref<10xf32>
return
@@ -106,9 +106,9 @@ func @invalid_prefetch_cache_type(%i : index) {
// -----
func @dma_start_non_affine_src_index(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
- %1 = alloc() : memref<100xf32, 2>
- %2 = alloc() : memref<1xi32, 4>
+ %0 = memref.alloc() : memref<100xf32>
+ %1 = memref.alloc() : memref<100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32, 4>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -123,9 +123,9 @@ func @dma_start_non_affine_src_index(%arg0 : index) {
// -----
func @dma_start_non_affine_dst_index(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
- %1 = alloc() : memref<100xf32, 2>
- %2 = alloc() : memref<1xi32, 4>
+ %0 = memref.alloc() : memref<100xf32>
+ %1 = memref.alloc() : memref<100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32, 4>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -140,9 +140,9 @@ func @dma_start_non_affine_dst_index(%arg0 : index) {
// -----
func @dma_start_non_affine_tag_index(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
- %1 = alloc() : memref<100xf32, 2>
- %2 = alloc() : memref<1xi32, 4>
+ %0 = memref.alloc() : memref<100xf32>
+ %1 = memref.alloc() : memref<100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32, 4>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
@@ -157,9 +157,9 @@ func @dma_start_non_affine_tag_index(%arg0 : index) {
// -----
func @dma_wait_non_affine_tag_index(%arg0 : index) {
- %0 = alloc() : memref<100xf32>
- %1 = alloc() : memref<100xf32, 2>
- %2 = alloc() : memref<1xi32, 4>
+ %0 = memref.alloc() : memref<100xf32>
+ %1 = memref.alloc() : memref<100xf32, 2>
+ %2 = memref.alloc() : memref<1xi32, 4>
%c0 = constant 0 : index
%c64 = constant 64 : index
affine.for %i0 = 0 to 10 {
diff --git a/mlir/test/Dialect/Affine/load-store.mlir b/mlir/test/Dialect/Affine/load-store.mlir
index 84613f436a1e..fa667f5f36a1 100644
--- a/mlir/test/Dialect/Affine/load-store.mlir
+++ b/mlir/test/Dialect/Affine/load-store.mlir
@@ -4,7 +4,7 @@
// Test with just loop IVs.
func @test0(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[%i0, %i1] : memref<100x100xf32>
@@ -18,7 +18,7 @@ func @test0(%arg0 : index, %arg1 : index) {
// Test with loop IVs and constants.
func @test1(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[%i0 + 3, %i1 + 7] : memref<100x100xf32>
@@ -35,7 +35,7 @@ func @test1(%arg0 : index, %arg1 : index) {
// Test with loop IVs and function args without 'symbol' keyword (should
// be parsed as dim identifiers).
func @test2(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[%i0 + %arg0, %i1 + %arg1] : memref<100x100xf32>
@@ -52,7 +52,7 @@ func @test2(%arg0 : index, %arg1 : index) {
// Test with loop IVs and function args with 'symbol' keyword (should
// be parsed as symbol identifiers).
func @test3(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[%i0 + symbol(%arg0), %i1 + symbol(%arg1)]
@@ -70,7 +70,7 @@ func @test3(%arg0 : index, %arg1 : index) {
// Test with loop IVs, symbols and constants in nested affine expressions.
func @test4(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[(%i0 + symbol(%arg0)) floordiv 3 + 11,
@@ -88,7 +88,7 @@ func @test4(%arg0 : index, %arg1 : index) {
// Test with swizzled loop IVs.
func @test5(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<10x10x10xf32>
+ %0 = memref.alloc() : memref<10x10x10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
affine.for %i2 = 0 to 10 {
@@ -108,7 +108,7 @@ func @test5(%arg0 : index, %arg1 : index) {
// Dim identifiers are assigned in parse order:
// d0 = %i2, d1 = %arg0, d2 = %i0, d3 = %i1, d4 = %arg1
func @test6(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<10x10x10xf32>
+ %0 = memref.alloc() : memref<10x10x10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
affine.for %i2 = 0 to 10 {
@@ -131,7 +131,7 @@ func @test6(%arg0 : index, %arg1 : index) {
// d0 = %i2, d1 = %i0, d2 = %i1
// s0 = %arg0, s1 = %arg1
func @test6(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<10x10x10xf32>
+ %0 = memref.alloc() : memref<10x10x10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
affine.for %i2 = 0 to 10 {
@@ -157,7 +157,7 @@ func @test6(%arg0 : index, %arg1 : index) {
// Test with operands without special SSA name.
func @test7() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
%1 = affine.apply affine_map<(d1) -> (d1 + 1)>(%i0)
%2 = affine.load %0[%1] : memref<10xf32>
@@ -183,7 +183,7 @@ func @zero_dim(%arg0 : memref<i32>, %arg1 : memref<i32>) {
// Test with loop IVs and constants.
func @test_prefetch(%arg0 : index, %arg1 : index) {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%1 = affine.load %0[%i0 + 3, %i1 + 7] : memref<100x100xf32>
@@ -200,12 +200,12 @@ func @test_prefetch(%arg0 : index, %arg1 : index) {
// Test with just loop IVs.
func @vector_load_vector_store_iv() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 16 {
affine.for %i1 = 0 to 16 step 8 {
%1 = affine.vector_load %0[%i0, %i1] : memref<100x100xf32>, vector<8xf32>
affine.vector_store %1, %0[%i0, %i1] : memref<100x100xf32>, vector<8xf32>
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK-NEXT: affine.for %[[i0:.*]] = 0
// CHECK-NEXT: affine.for %[[i1:.*]] = 0
// CHECK-NEXT: %[[val:.*]] = affine.vector_load %[[buf]][%[[i0]], %[[i1]]] : memref<100x100xf32>, vector<8xf32>
@@ -219,12 +219,12 @@ func @vector_load_vector_store_iv() {
// Test with loop IVs and constants.
func @vector_load_vector_store_iv_constant() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 16 step 4 {
%1 = affine.vector_load %0[%i0 + 3, %i1 + 7] : memref<100x100xf32>, vector<4xf32>
affine.vector_store %1, %0[%i0 + 3, %i1 + 7] : memref<100x100xf32>, vector<4xf32>
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK-NEXT: affine.for %[[i0:.*]] = 0
// CHECK-NEXT: affine.for %[[i1:.*]] = 0
// CHECK-NEXT: %[[val:.*]] = affine.vector_load %{{.*}}[%{{.*}} + 3, %{{.*}} + 7] : memref<100x100xf32>, vector<4xf32>
@@ -237,12 +237,12 @@ func @vector_load_vector_store_iv_constant() {
// -----
func @vector_load_vector_store_2d() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
affine.for %i0 = 0 to 16 step 2{
affine.for %i1 = 0 to 16 step 8 {
%1 = affine.vector_load %0[%i0, %i1] : memref<100x100xf32>, vector<2x8xf32>
affine.vector_store %1, %0[%i0, %i1] : memref<100x100xf32>, vector<2x8xf32>
-// CHECK: %[[buf:.*]] = alloc
+// CHECK: %[[buf:.*]] = memref.alloc
// CHECK-NEXT: affine.for %[[i0:.*]] = 0
// CHECK-NEXT: affine.for %[[i1:.*]] = 0
// CHECK-NEXT: %[[val:.*]] = affine.vector_load %[[buf]][%[[i0]], %[[i1]]] : memref<100x100xf32>, vector<2x8xf32>
diff --git a/mlir/test/Dialect/Affine/loop-tiling-parametric.mlir b/mlir/test/Dialect/Affine/loop-tiling-parametric.mlir
index 938d632f1541..0618a0c8cfec 100644
--- a/mlir/test/Dialect/Affine/loop-tiling-parametric.mlir
+++ b/mlir/test/Dialect/Affine/loop-tiling-parametric.mlir
@@ -73,7 +73,7 @@ func @loop_tiling_non_unit_step(%t0: index, %t1: index, %t2: index){
#ub = affine_map<()[s0, s1] -> (s0, 4096 floordiv s1)>
func @tile_loop_with_div_in_upper_bound(%t5 : index, %A : memref<? x i32>, %L : index, %U : index) {
%c0 = constant 0 : index
- %M = dim %A, %c0 : memref<? x i32>
+ %M = memref.dim %A, %c0 : memref<? x i32>
affine.for %i = 0 to min #ub()[%M, %U] {
addi %i, %i : index
}
@@ -93,7 +93,7 @@ func @tile_loop_with_div_in_upper_bound(%t5 : index, %A : memref<? x i32>, %L :
#ub = affine_map<()[s0, s1] -> (s0, 4096 floordiv s1)>
func @tile_loop_with_div_in_upper_bound_non_unit_step(%t5 : index, %A : memref<? x i32>, %L : index, %U : index) {
%c0 = constant 0 : index
- %M = dim %A, %c0 : memref<? x i32>
+ %M = memref.dim %A, %c0 : memref<? x i32>
affine.for %i = 0 to min #ub()[%M, %U] step 4 {
addi %i, %i : index
}
@@ -191,7 +191,7 @@ func @simple_matmul(%t6 : index, %t7 : index, %t8 : index, %arg0: memref<256x256
func @tile_with_symbolic_loop_upper_bounds(%t9 : index, %t10: index, %arg0: memref<?x?xf32>, %arg1: memref<?x?xf32>, %arg2: memref<?x?xf32>) {
%cst = constant 0.000000e+00 : f32
%c0 = constant 0 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32>
affine.for %i0 = 0 to %0 {
affine.for %i1 = 0 to %0 {
affine.store %cst, %arg2[%i0, %i1] : memref<?x?xf32>
@@ -217,7 +217,7 @@ func @tile_with_symbolic_loop_upper_bounds(%t9 : index, %t10: index, %arg0: memr
// CHECK: func @tile_with_loop_upper_bounds_in_two_symbols([[ARG0:%arg[0-9]+]]: index{{.*}}){{.*}}
func @tile_with_loop_upper_bounds_in_two_symbols(%t11 : index, %arg0: memref<?xf32>, %limit: index) {
%c0 = constant 0 : index
- %dim0 = dim %arg0, %c0 : memref<?xf32>
+ %dim0 = memref.dim %arg0, %c0 : memref<?xf32>
affine.for %i0 = 0 to affine_map<()[s0, s1] -> (s0 + s1)> ()[%dim0, %limit] {
%v0 = affine.load %arg0[%i0] : memref<?xf32>
}
diff --git a/mlir/test/Dialect/Affine/loop-tiling-validity.mlir b/mlir/test/Dialect/Affine/loop-tiling-validity.mlir
index 6468d7ed3dc0..786313bacdc1 100644
--- a/mlir/test/Dialect/Affine/loop-tiling-validity.mlir
+++ b/mlir/test/Dialect/Affine/loop-tiling-validity.mlir
@@ -9,7 +9,7 @@
// CHECK-LABEL: func @legal_loop()
func @legal_loop() {
- %0 = alloc() : memref<64xf32>
+ %0 = memref.alloc() : memref<64xf32>
affine.for %i = 0 to 64 {
%1 = affine.load %0[%i] : memref<64xf32>
@@ -32,7 +32,7 @@ func @legal_loop() {
// CHECK-LABEL: func @illegal_loop_with_diag_dependence
func @illegal_loop_with_diag_dependence() {
- %A = alloc() : memref<64x64xf32>
+ %A = memref.alloc() : memref<64x64xf32>
affine.for %i = 0 to 64 {
// expected-remark at above {{tiled code is illegal due to dependences}}
diff --git a/mlir/test/Dialect/Affine/loop-tiling.mlir b/mlir/test/Dialect/Affine/loop-tiling.mlir
index 0b6542f08e53..2fcbb2622ade 100644
--- a/mlir/test/Dialect/Affine/loop-tiling.mlir
+++ b/mlir/test/Dialect/Affine/loop-tiling.mlir
@@ -67,7 +67,7 @@ func @loop_tiling() {
// CHECK-LABEL: func @loop_max_min_bound(%{{.*}}: memref<?xi32>, %{{.*}}: index, %{{.*}}: index) {
func @loop_max_min_bound(%A : memref<? x i32>, %L : index, %U : index) {
%c0 = constant 0 : index
- %M = dim %A, %c0 : memref<? x i32>
+ %M = memref.dim %A, %c0 : memref<? x i32>
affine.for %i = max #lb()[%L] to min #ub()[%M, %U] {
addi %i, %i : index
}
@@ -113,7 +113,7 @@ func @simple_matmul(%arg0: memref<256x256xvector<64xf32>>, %arg1: memref<256x256
func @tile_with_symbolic_loop_upper_bounds(%arg0: memref<?x?xf32>, %arg1: memref<?x?xf32>, %arg2: memref<?x?xf32>) {
%cst = constant 0.000000e+00 : f32
%c0 = constant 0 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32>
affine.for %i0 = 0 to %0 {
affine.for %i1 = 0 to %0 {
affine.store %cst, %arg2[%i0, %i1] : memref<?x?xf32>
@@ -130,7 +130,7 @@ func @tile_with_symbolic_loop_upper_bounds(%arg0: memref<?x?xf32>, %arg1: memref
return
}
-// CHECK: dim %{{.*}}, %c0 : memref<?x?xf32>
+// CHECK: memref.dim %{{.*}}, %c0 : memref<?x?xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to %{{.*}} step 32 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to %{{.*}} step 32 {
// CHECK-NEXT: affine.for %{{.*}} = #map0(%{{.*}}) to min [[$UBMAP]](%{{.*}})[%{{.*}}] {
@@ -158,14 +158,14 @@ func @tile_with_symbolic_loop_upper_bounds(%arg0: memref<?x?xf32>, %arg1: memref
func @tile_with_loop_upper_bounds_in_two_symbols(%arg0: memref<?xf32>, %limit: index) {
%c0 = constant 0 : index
- %dim0 = dim %arg0, %c0 : memref<?xf32>
+ %dim0 = memref.dim %arg0, %c0 : memref<?xf32>
affine.for %i0 = 0 to affine_map<()[s0, s1] -> (s0 + s1)> ()[%dim0, %limit] {
%v0 = affine.load %arg0[%i0] : memref<?xf32>
}
return
}
-// CHECK: dim %{{.*}}, %c0 : memref<?xf32>
+// CHECK: memref.dim %{{.*}}, %c0 : memref<?xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to [[MAP1]]()[%{{.*}}, %{{.*}}] step 32 {
// CHECK-NEXT: affine.for %{{.*}} = [[MAP0]](%{{.*}}) to min [[$UBMAP]](%{{.*}})[%{{.*}}, %{{.*}}] {
// CHECK-NEXT: affine.load
diff --git a/mlir/test/Dialect/Affine/memref-stride-calculation.mlir b/mlir/test/Dialect/Affine/memref-stride-calculation.mlir
index a52de3d858d8..d410e9d825bc 100644
--- a/mlir/test/Dialect/Affine/memref-stride-calculation.mlir
+++ b/mlir/test/Dialect/Affine/memref-stride-calculation.mlir
@@ -2,79 +2,79 @@
func @f(%0: index) {
// CHECK-LABEL: Testing: f
- %1 = alloc() : memref<3x4x5xf32>
+ %1 = memref.alloc() : memref<3x4x5xf32>
// CHECK: MemRefType offset: 0 strides: 20, 5, 1
- %2 = alloc(%0) : memref<3x4x?xf32>
+ %2 = memref.alloc(%0) : memref<3x4x?xf32>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %3 = alloc(%0) : memref<3x?x5xf32>
+ %3 = memref.alloc(%0) : memref<3x?x5xf32>
// CHECK: MemRefType offset: 0 strides: ?, 5, 1
- %4 = alloc(%0) : memref<?x4x5xf32>
+ %4 = memref.alloc(%0) : memref<?x4x5xf32>
// CHECK: MemRefType offset: 0 strides: 20, 5, 1
- %5 = alloc(%0, %0) : memref<?x4x?xf32>
+ %5 = memref.alloc(%0, %0) : memref<?x4x?xf32>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %6 = alloc(%0, %0, %0) : memref<?x?x?xf32>
+ %6 = memref.alloc(%0, %0, %0) : memref<?x?x?xf32>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %11 = alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %11 = memref.alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: 20, 5, 1
- %b11 = alloc() : memref<3x4x5xf32, offset: 0, strides: [20, 5, 1]>
+ %b11 = memref.alloc() : memref<3x4x5xf32, offset: 0, strides: [20, 5, 1]>
// CHECK: MemRefType offset: 0 strides: 20, 5, 1
- %12 = alloc(%0) : memref<3x4x?xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %12 = memref.alloc(%0) : memref<3x4x?xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %13 = alloc(%0) : memref<3x?x5xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %13 = memref.alloc(%0) : memref<3x?x5xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: ?, 5, 1
- %14 = alloc(%0) : memref<?x4x5xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %14 = memref.alloc(%0) : memref<?x4x5xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: 20, 5, 1
- %15 = alloc(%0, %0) : memref<?x4x?xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %15 = memref.alloc(%0, %0) : memref<?x4x?xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %16 = alloc(%0, %0, %0) : memref<?x?x?xf32, affine_map<(i, j, k)->(i, j, k)>>
+ %16 = memref.alloc(%0, %0, %0) : memref<?x?x?xf32, affine_map<(i, j, k)->(i, j, k)>>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %21 = alloc()[%0] : memref<3x4x5xf32, affine_map<(i, j, k)[M]->(32 * i + 16 * j + M * k + 1)>>
+ %21 = memref.alloc()[%0] : memref<3x4x5xf32, affine_map<(i, j, k)[M]->(32 * i + 16 * j + M * k + 1)>>
// CHECK: MemRefType offset: 1 strides: 32, 16, ?
- %22 = alloc()[%0] : memref<3x4x5xf32, affine_map<(i, j, k)[M]->(32 * i + M * j + 16 * k + 3)>>
+ %22 = memref.alloc()[%0] : memref<3x4x5xf32, affine_map<(i, j, k)[M]->(32 * i + M * j + 16 * k + 3)>>
// CHECK: MemRefType offset: 3 strides: 32, ?, 16
- %b22 = alloc(%0)[%0, %0] : memref<3x4x?xf32, offset: 0, strides: [?, ?, 1]>
+ %b22 = memref.alloc(%0)[%0, %0] : memref<3x4x?xf32, offset: 0, strides: [?, ?, 1]>
// CHECK: MemRefType offset: 0 strides: ?, ?, 1
- %23 = alloc(%0)[%0] : memref<3x?x5xf32, affine_map<(i, j, k)[M]->(M * i + 32 * j + 16 * k + 7)>>
+ %23 = memref.alloc(%0)[%0] : memref<3x?x5xf32, affine_map<(i, j, k)[M]->(M * i + 32 * j + 16 * k + 7)>>
// CHECK: MemRefType offset: 7 strides: ?, 32, 16
- %b23 = alloc(%0)[%0] : memref<3x?x5xf32, offset: 0, strides: [?, 5, 1]>
+ %b23 = memref.alloc(%0)[%0] : memref<3x?x5xf32, offset: 0, strides: [?, 5, 1]>
// CHECK: MemRefType offset: 0 strides: ?, 5, 1
- %24 = alloc(%0)[%0] : memref<3x?x5xf32, affine_map<(i, j, k)[M]->(M * i + 32 * j + 16 * k + M)>>
+ %24 = memref.alloc(%0)[%0] : memref<3x?x5xf32, affine_map<(i, j, k)[M]->(M * i + 32 * j + 16 * k + M)>>
// CHECK: MemRefType offset: ? strides: ?, 32, 16
- %b24 = alloc(%0)[%0, %0] : memref<3x?x5xf32, offset: ?, strides: [?, 32, 16]>
+ %b24 = memref.alloc(%0)[%0, %0] : memref<3x?x5xf32, offset: ?, strides: [?, 32, 16]>
// CHECK: MemRefType offset: ? strides: ?, 32, 16
- %25 = alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, affine_map<(i, j, k)[M, N]->(M * i + N * j + k + 1)>>
+ %25 = memref.alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, affine_map<(i, j, k)[M, N]->(M * i + N * j + k + 1)>>
// CHECK: MemRefType offset: 1 strides: ?, ?, 1
- %b25 = alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, offset: 1, strides: [?, ?, 1]>
+ %b25 = memref.alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, offset: 1, strides: [?, ?, 1]>
// CHECK: MemRefType offset: 1 strides: ?, ?, 1
- %26 = alloc(%0)[] : memref<?xf32, affine_map<(i)[M]->(i)>>
+ %26 = memref.alloc(%0)[] : memref<?xf32, affine_map<(i)[M]->(i)>>
// CHECK: MemRefType offset: 0 strides: 1
- %27 = alloc()[%0] : memref<5xf32, affine_map<(i)[M]->(M)>>
+ %27 = memref.alloc()[%0] : memref<5xf32, affine_map<(i)[M]->(M)>>
// CHECK: MemRefType memref<5xf32, affine_map<(d0)[s0] -> (s0)>> cannot be converted to strided form
- %28 = alloc()[%0] : memref<5xf32, affine_map<(i)[M]->(123)>>
+ %28 = memref.alloc()[%0] : memref<5xf32, affine_map<(i)[M]->(123)>>
// CHECK: MemRefType memref<5xf32, affine_map<(d0)[s0] -> (123)>> cannot be converted to strided form
- %29 = alloc()[%0] : memref<f32, affine_map<()[M]->(M)>>
+ %29 = memref.alloc()[%0] : memref<f32, affine_map<()[M]->(M)>>
// CHECK: MemRefType offset: ? strides:
- %30 = alloc()[%0] : memref<f32, affine_map<()[M]->(123)>>
+ %30 = memref.alloc()[%0] : memref<f32, affine_map<()[M]->(123)>>
// CHECK: MemRefType offset: 123 strides:
- %100 = alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, affine_map<(i, j, k)[M, N]->(i + j, j, k)>, affine_map<(i, j, k)[M, N]->(M * i + N * j + k + 1)>>
+ %100 = memref.alloc(%0, %0)[%0, %0] : memref<?x?x16xf32, affine_map<(i, j, k)[M, N]->(i + j, j, k)>, affine_map<(i, j, k)[M, N]->(M * i + N * j + k + 1)>>
// CHECK: MemRefType memref<?x?x16xf32, affine_map<(d0, d1, d2)[s0, s1] -> (d0 + d1, d1, d2)>, affine_map<(d0, d1, d2)[s0, s1] -> (d0 * s0 + d1 * s1 + d2 + 1)>> cannot be converted to strided form
- %101 = alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i floordiv 4 + j + k)>>
+ %101 = memref.alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i floordiv 4 + j + k)>>
// CHECK: MemRefType memref<3x4x5xf32, affine_map<(d0, d1, d2) -> (d0 floordiv 4 + d1 + d2)>> cannot be converted to strided form
- %102 = alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i ceildiv 4 + j + k)>>
+ %102 = memref.alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i ceildiv 4 + j + k)>>
// CHECK: MemRefType memref<3x4x5xf32, affine_map<(d0, d1, d2) -> (d0 ceildiv 4 + d1 + d2)>> cannot be converted to strided form
- %103 = alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i mod 4 + j + k)>>
+ %103 = memref.alloc() : memref<3x4x5xf32, affine_map<(i, j, k)->(i mod 4 + j + k)>>
// CHECK: MemRefType memref<3x4x5xf32, affine_map<(d0, d1, d2) -> (d0 mod 4 + d1 + d2)>> cannot be converted to strided form
- %200 = alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * i + N * i + N * j + K * k - (M + N - 20)* i)>>
+ %200 = memref.alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * i + N * i + N * j + K * k - (M + N - 20)* i)>>
// CHECK: MemRefType offset: 0 strides: 20, ?, ?
- %201 = alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * i + N * i + N * K * j + K * K * k - (M + N - 20) * (i + 1))>>
+ %201 = memref.alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * i + N * i + N * K * j + K * K * k - (M + N - 20) * (i + 1))>>
// CHECK: MemRefType offset: ? strides: 20, ?, ?
- %202 = alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * (i + 1) + j + k - M)>>
+ %202 = memref.alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M * (i + 1) + j + k - M)>>
// CHECK: MemRefType offset: 0 strides: ?, 1, 1
- %203 = alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M + M * (i + N * (j + K * k)))>>
+ %203 = memref.alloc()[%0, %0, %0] : memref<3x4x5xf32, affine_map<(i, j, k)[M, N, K]->(M + M * (i + N * (j + K * k)))>>
// CHECK: MemRefType offset: ? strides: ?, ?, ?
return
diff --git a/mlir/test/Dialect/Affine/ops.mlir b/mlir/test/Dialect/Affine/ops.mlir
index ad2fb5fd6a59..ae21d531e9aa 100644
--- a/mlir/test/Dialect/Affine/ops.mlir
+++ b/mlir/test/Dialect/Affine/ops.mlir
@@ -97,14 +97,14 @@ func @affine_max(%arg0 : index, %arg1 : index, %arg2 : index) {
func @valid_symbols(%arg0: index, %arg1: index, %arg2: index) {
%c1 = constant 1 : index
%c0 = constant 0 : index
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
affine.for %arg3 = 0 to %arg2 step 768 {
- %13 = dim %0, %c1 : memref<?x?xf32>
+ %13 = memref.dim %0, %c1 : memref<?x?xf32>
affine.for %arg4 = 0 to %13 step 264 {
- %18 = dim %0, %c0 : memref<?x?xf32>
- %20 = std.subview %0[%c0, %c0][%18,%arg4][%c1,%c1] : memref<?x?xf32>
+ %18 = memref.dim %0, %c0 : memref<?x?xf32>
+ %20 = memref.subview %0[%c0, %c0][%18,%arg4][%c1,%c1] : memref<?x?xf32>
to memref<?x?xf32, offset : ?, strides : [?, ?]>
- %24 = dim %20, %c0 : memref<?x?xf32, offset : ?, strides : [?, ?]>
+ %24 = memref.dim %20, %c0 : memref<?x?xf32, offset : ?, strides : [?, ?]>
affine.for %arg5 = 0 to %24 step 768 {
"foo"() : () -> ()
}
diff --git a/mlir/test/Dialect/Affine/parallelize.mlir b/mlir/test/Dialect/Affine/parallelize.mlir
index 5ca2f4652827..ca72e67f9108 100644
--- a/mlir/test/Dialect/Affine/parallelize.mlir
+++ b/mlir/test/Dialect/Affine/parallelize.mlir
@@ -4,8 +4,8 @@
// CHECK-LABEL: func @reduce_window_max() {
func @reduce_window_max() {
%cst = constant 0.000000e+00 : f32
- %0 = alloc() : memref<1x8x8x64xf32>
- %1 = alloc() : memref<1x18x18x64xf32>
+ %0 = memref.alloc() : memref<1x8x8x64xf32>
+ %1 = memref.alloc() : memref<1x18x18x64xf32>
affine.for %arg0 = 0 to 1 {
affine.for %arg1 = 0 to 8 {
affine.for %arg2 = 0 to 8 {
@@ -40,8 +40,8 @@ func @reduce_window_max() {
}
// CHECK: %[[cst:.*]] = constant 0.000000e+00 : f32
-// CHECK: %[[v0:.*]] = alloc() : memref<1x8x8x64xf32>
-// CHECK: %[[v1:.*]] = alloc() : memref<1x18x18x64xf32>
+// CHECK: %[[v0:.*]] = memref.alloc() : memref<1x8x8x64xf32>
+// CHECK: %[[v1:.*]] = memref.alloc() : memref<1x18x18x64xf32>
// CHECK: affine.parallel (%[[arg0:.*]]) = (0) to (1) {
// CHECK: affine.parallel (%[[arg1:.*]]) = (0) to (8) {
// CHECK: affine.parallel (%[[arg2:.*]]) = (0) to (8) {
@@ -75,9 +75,9 @@ func @reduce_window_max() {
// CHECK: }
func @loop_nest_3d_outer_two_parallel(%N : index) {
- %0 = alloc() : memref<1024 x 1024 x vector<64xf32>>
- %1 = alloc() : memref<1024 x 1024 x vector<64xf32>>
- %2 = alloc() : memref<1024 x 1024 x vector<64xf32>>
+ %0 = memref.alloc() : memref<1024 x 1024 x vector<64xf32>>
+ %1 = memref.alloc() : memref<1024 x 1024 x vector<64xf32>>
+ %2 = memref.alloc() : memref<1024 x 1024 x vector<64xf32>>
affine.for %i = 0 to %N {
affine.for %j = 0 to %N {
%7 = affine.load %2[%i, %j] : memref<1024x1024xvector<64xf32>>
@@ -108,10 +108,10 @@ func @unknown_op_conservative() {
// CHECK-LABEL: non_affine_load
func @non_affine_load() {
- %0 = alloc() : memref<100 x f32>
+ %0 = memref.alloc() : memref<100 x f32>
affine.for %i = 0 to 100 {
// CHECK: affine.for %{{.*}} = 0 to 100 {
- load %0[%i] : memref<100 x f32>
+ memref.load %0[%i] : memref<100 x f32>
}
return
}
diff --git a/mlir/test/Dialect/Affine/slicing-utils.mlir b/mlir/test/Dialect/Affine/slicing-utils.mlir
index e11a66b0d0eb..9ce5cb1fdd72 100644
--- a/mlir/test/Dialect/Affine/slicing-utils.mlir
+++ b/mlir/test/Dialect/Affine/slicing-utils.mlir
@@ -17,7 +17,7 @@
// FWDBWD-LABEL: slicing_test
func @slicing_test() {
// Fake 0 to align on 1 and match ASCII art.
- %0 = alloc() : memref<1xi32>
+ %0 = memref.alloc() : memref<1xi32>
// FWD: matched: %[[v1:.*]] {{.*}} forward static slice:
// FWD-NEXT: %[[v5:.*]] {{.*}} -> i5
diff --git a/mlir/test/Dialect/Affine/unroll.mlir b/mlir/test/Dialect/Affine/unroll.mlir
index 67fa20e4716a..f78c6c325ea2 100644
--- a/mlir/test/Dialect/Affine/unroll.mlir
+++ b/mlir/test/Dialect/Affine/unroll.mlir
@@ -265,9 +265,9 @@ func @loop_nest_outer_unroll() {
// count threshold set to 2.
// SHORT-LABEL: func @loop_nest_seq_long() -> i32 {
func @loop_nest_seq_long() -> i32 {
- %A = alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %B = alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- %C = alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %A = memref.alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %B = memref.alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %C = memref.alloc() : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
%zero = constant 0 : i32
%one = constant 1 : i32
@@ -279,9 +279,9 @@ func @loop_nest_seq_long() -> i32 {
affine.for %n0 = 0 to 512 {
// CHECK: affine.for %arg1 = 0 to 8
affine.for %n1 = 0 to 8 {
- store %one, %A[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- store %two, %B[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
- store %zero, %C[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ memref.store %one, %A[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ memref.store %two, %B[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ memref.store %zero, %C[%n0, %n1] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
}
}
@@ -292,27 +292,27 @@ func @loop_nest_seq_long() -> i32 {
// CHECK-NOT: affine.for
// CHECK: %{{[0-9]+}} = affine.apply
%b2 = "affine.apply" (%y, %arg2) {map = affine_map<(d0, d1) -> (16*d0 + d1)>} : (index, index) -> index
- %z = load %B[%x, %b2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %z = memref.load %B[%x, %b2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
"op1"(%z) : (i32) -> ()
}
affine.for %j1 = 0 to 8 {
affine.for %j2 = 0 to 8 {
%a2 = "affine.apply" (%y, %j2) {map = affine_map<(d0, d1) -> (16*d0 + d1)>} : (index, index) -> index
- %v203 = load %A[%j1, %a2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %v203 = memref.load %A[%j1, %a2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
"op2"(%v203) : (i32) -> ()
}
affine.for %k2 = 0 to 8 {
%s0 = "op3"() : () -> i32
%c2 = "affine.apply" (%x, %k2) {map = affine_map<(d0, d1) -> (16*d0 + d1)>} : (index, index) -> index
- %s1 = load %C[%j1, %c2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %s1 = memref.load %C[%j1, %c2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
%s2 = "addi32"(%s0, %s1) : (i32, i32) -> i32
- store %s2, %C[%j1, %c2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ memref.store %s2, %C[%j1, %c2] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
}
}
"op4"() : () -> ()
}
}
- %ret = load %C[%zero_idx, %zero_idx] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
+ %ret = memref.load %C[%zero_idx, %zero_idx] : memref<512 x 512 x i32, affine_map<(d0, d1) -> (d0, d1)>, 2>
return %ret : i32
}
diff --git a/mlir/test/Dialect/Async/async-parallel-for.mlir b/mlir/test/Dialect/Async/async-parallel-for.mlir
index ad21650ab0ee..d498fe80906b 100644
--- a/mlir/test/Dialect/Async/async-parallel-for.mlir
+++ b/mlir/test/Dialect/Async/async-parallel-for.mlir
@@ -6,14 +6,14 @@ func @loop_1d(%arg0: index, %arg1: index, %arg2: index, %arg3: memref<?xf32>) {
// CHECK: scf.for
// CHECK: %[[TOKEN:.*]] = async.execute {
// CHECK: scf.for
- // CHECK: store
+ // CHECK: memref.store
// CHECK: async.yield
// CHECK: }
// CHECK: async.add_to_group %[[TOKEN]], %[[GROUP]]
// CHECK: async.await_all %[[GROUP]]
scf.parallel (%i) = (%arg0) to (%arg1) step (%arg2) {
%one = constant 1.0 : f32
- store %one, %arg3[%i] : memref<?xf32>
+ memref.store %one, %arg3[%i] : memref<?xf32>
}
return
@@ -29,7 +29,7 @@ func @loop_2d(%arg0: index, %arg1: index, %arg2: index, // lb, ub, step
// CHECK: %[[TOKEN:.*]] = async.execute {
// CHECK: scf.for
// CHECK: scf.for
- // CHECK: store
+ // CHECK: memref.store
// CHECK: async.yield
// CHECK: }
// CHECK: async.add_to_group %[[TOKEN]], %[[GROUP]]
@@ -37,7 +37,7 @@ func @loop_2d(%arg0: index, %arg1: index, %arg2: index, // lb, ub, step
scf.parallel (%i0, %i1) = (%arg0, %arg3) to (%arg1, %arg4)
step (%arg2, %arg5) {
%one = constant 1.0 : f32
- store %one, %arg6[%i0, %i1] : memref<?x?xf32>
+ memref.store %one, %arg6[%i0, %i1] : memref<?x?xf32>
}
return
diff --git a/mlir/test/Dialect/Async/async-to-async-runtime.mlir b/mlir/test/Dialect/Async/async-to-async-runtime.mlir
index 7d4d05538629..a57c283b3de2 100644
--- a/mlir/test/Dialect/Async/async-to-async-runtime.mlir
+++ b/mlir/test/Dialect/Async/async-to-async-runtime.mlir
@@ -4,7 +4,7 @@
func @execute_no_async_args(%arg0: f32, %arg1: memref<1xf32>) {
%token = async.execute {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
async.await %token : !async.token
@@ -28,7 +28,7 @@ func @execute_no_async_args(%arg0: f32, %arg1: memref<1xf32>) {
// Resume coroutine after suspension.
// CHECK: ^[[RESUME]]:
-// CHECK: store
+// CHECK: memref.store
// CHECK: async.runtime.set_available %[[TOKEN]]
// Delete coroutine.
@@ -50,12 +50,12 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
%token1 = async.execute {
%c1 = constant 1: index
- store %arg0, %arg2[%c0] : memref<1xf32>
+ memref.store %arg0, %arg2[%c0] : memref<1xf32>
async.yield
}
async.await %token1 : !async.token
- store %arg1, %arg2[%c0] : memref<1xf32>
+ memref.store %arg1, %arg2[%c0] : memref<1xf32>
async.yield
}
// CHECK: async.runtime.await %[[TOKEN]]
@@ -77,7 +77,7 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
// CHECK-SAME: ^[[SUSPEND:.*]], ^[[RESUME:.*]], ^[[CLEANUP:.*]]
// CHECK: ^[[RESUME]]:
-// CHECK: store
+// CHECK: memref.store
// CHECK: async.runtime.set_available %[[TOKEN]]
// Function outlined from the outer async.execute operation.
@@ -103,7 +103,7 @@ func @nested_async_execute(%arg0: f32, %arg1: f32, %arg2: memref<1xf32>) {
// Set token available after second resumption.
// CHECK: ^[[RESUME_1]]:
-// CHECK: store
+// CHECK: memref.store
// CHECK: async.runtime.set_available %[[TOKEN]]
// CHECK: ^[[CLEANUP]]:
@@ -116,13 +116,13 @@ func @async_execute_token_dependency(%arg0: f32, %arg1: memref<1xf32>) {
// CHECK: %[[TOKEN:.*]] = call @async_execute_fn
%token = async.execute {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
// CHECK: call @async_execute_fn_0(%[[TOKEN]], %arg0, %arg1)
%token_0 = async.execute [%token] {
%c0 = constant 0 : index
- store %arg0, %arg1[%c0] : memref<1xf32>
+ memref.store %arg0, %arg1[%c0] : memref<1xf32>
async.yield
}
return
@@ -157,7 +157,7 @@ func @async_execute_token_dependency(%arg0: f32, %arg1: memref<1xf32>) {
// Emplace result token after second resumption.
// CHECK: ^[[RESUME_1]]:
-// CHECK: store
+// CHECK: memref.store
// CHECK: async.runtime.set_available %[[TOKEN]]
// CHECK: ^[[CLEANUP]]:
diff --git a/mlir/test/Dialect/GPU/all-reduce-max.mlir b/mlir/test/Dialect/GPU/all-reduce-max.mlir
index 04cb68aaf7b7..5995a9209a88 100644
--- a/mlir/test/Dialect/GPU/all-reduce-max.mlir
+++ b/mlir/test/Dialect/GPU/all-reduce-max.mlir
@@ -121,7 +121,7 @@ gpu.module @kernels {
// CHECK: cond_br [[VAL_84]], ^bb22, ^bb41
// CHECK: ^bb22:
// CHECK: [[VAL_85:%.*]] = index_cast [[VAL_27]] : i32 to index
- // CHECK: [[VAL_86:%.*]] = load [[VAL_1]]{{\[}}[[VAL_85]]] : memref<32xf32, 3>
+ // CHECK: [[VAL_86:%.*]] = memref.load [[VAL_1]]{{\[}}[[VAL_85]]] : memref<32xf32, 3>
// CHECK: [[VAL_87:%.*]] = cmpi slt, [[VAL_83]], [[VAL_5]] : i32
// CHECK: cond_br [[VAL_87]], ^bb23, ^bb39
// CHECK: ^bb23:
diff --git a/mlir/test/Dialect/GPU/all-reduce.mlir b/mlir/test/Dialect/GPU/all-reduce.mlir
index 4cc4869dfee5..22e91a6bb6c7 100644
--- a/mlir/test/Dialect/GPU/all-reduce.mlir
+++ b/mlir/test/Dialect/GPU/all-reduce.mlir
@@ -111,7 +111,7 @@ gpu.module @kernels {
// CHECK: cond_br [[VAL_74]], ^bb22, ^bb41
// CHECK: ^bb22:
// CHECK: [[VAL_75:%.*]] = index_cast [[VAL_27]] : i32 to index
- // CHECK: [[VAL_76:%.*]] = load [[VAL_1]]{{\[}}[[VAL_75]]] : memref<32xf32, 3>
+ // CHECK: [[VAL_76:%.*]] = memref.load [[VAL_1]]{{\[}}[[VAL_75]]] : memref<32xf32, 3>
// CHECK: [[VAL_77:%.*]] = cmpi slt, [[VAL_73]], [[VAL_5]] : i32
// CHECK: cond_br [[VAL_77]], ^bb23, ^bb39
// CHECK: ^bb23:
diff --git a/mlir/test/Dialect/GPU/multiple-all-reduce.mlir b/mlir/test/Dialect/GPU/multiple-all-reduce.mlir
index 497adb50c142..5f6619898407 100644
--- a/mlir/test/Dialect/GPU/multiple-all-reduce.mlir
+++ b/mlir/test/Dialect/GPU/multiple-all-reduce.mlir
@@ -1,19 +1,19 @@
// RUN: mlir-opt --gpu-kernel-outlining --convert-gpu-to-nvvm %s | FileCheck %s
func @main() {
- %data = alloc() : memref<2x6xf32>
- %sum = alloc() : memref<2xf32>
- %mul = alloc() : memref<2xf32>
+ %data = memref.alloc() : memref<2x6xf32>
+ %sum = memref.alloc() : memref<2xf32>
+ %mul = memref.alloc() : memref<2xf32>
%c1 = constant 1 : index
// ADD + MUL
gpu.launch blocks(%bx, %by, %bz) in (%grid_x = %c1, %grid_y = %c1, %grid_z = %c1)
threads(%tx, %ty, %tz) in (%block_x = %c1, %block_y = %c1, %block_z = %c1) {
- %val = load %data[%bx, %tx] : memref<2x6xf32>
+ %val = memref.load %data[%bx, %tx] : memref<2x6xf32>
%reduced0 = "gpu.all_reduce"(%val) ({}) { op = "add" } : (f32) -> (f32)
- store %reduced0, %sum[%bx] : memref<2xf32>
+ memref.store %reduced0, %sum[%bx] : memref<2xf32>
%reduced1 = "gpu.all_reduce"(%val) ({}) { op = "mul" } : (f32) -> (f32)
- store %reduced1, %mul[%bx] : memref<2xf32>
+ memref.store %reduced1, %mul[%bx] : memref<2xf32>
gpu.terminator
}
diff --git a/mlir/test/Dialect/GPU/ops.mlir b/mlir/test/Dialect/GPU/ops.mlir
index 5cea772ce599..7c1420115fc7 100644
--- a/mlir/test/Dialect/GPU/ops.mlir
+++ b/mlir/test/Dialect/GPU/ops.mlir
@@ -59,7 +59,7 @@ module attributes {gpu.container_module} {
"gpu.barrier"() : () -> ()
"some_op"(%bIdX, %tIdX) : (index, index) -> ()
- %42 = load %arg1[%bIdX] : memref<?xf32, 1>
+ %42 = memref.load %arg1[%bIdX] : memref<?xf32, 1>
gpu.return
}
diff --git a/mlir/test/Dialect/GPU/outlining.mlir b/mlir/test/Dialect/GPU/outlining.mlir
index cb98ebc1544b..ddfa6588b63d 100644
--- a/mlir/test/Dialect/GPU/outlining.mlir
+++ b/mlir/test/Dialect/GPU/outlining.mlir
@@ -29,7 +29,7 @@ func @launch() {
%block_z = %bDimZ) {
"use"(%0): (f32) -> ()
"some_op"(%bx, %block_x) : (index, index) -> ()
- %42 = load %1[%tx] : memref<?xf32, 1>
+ %42 = memref.load %1[%tx] : memref<?xf32, 1>
gpu.terminator
}
return
@@ -55,7 +55,7 @@ func @launch() {
// CHECK-NEXT: ^[[BLOCK]]:
// CHECK-NEXT: "use"(%[[KERNEL_ARG0]]) : (f32) -> ()
// CHECK-NEXT: "some_op"(%[[BID]], %[[BDIM]]) : (index, index) -> ()
-// CHECK-NEXT: = load %[[KERNEL_ARG1]][%[[TID]]] : memref<?xf32, 1>
+// CHECK-NEXT: = memref.load %[[KERNEL_ARG1]][%[[TID]]] : memref<?xf32, 1>
// -----
@@ -118,7 +118,7 @@ func @extra_constants(%arg0: memref<?xf32>) {
%cst = constant 8 : index
%cst2 = constant 2 : index
%c0 = constant 0 : index
- %cst3 = dim %arg0, %c0 : memref<?xf32>
+ %cst3 = memref.dim %arg0, %c0 : memref<?xf32>
// CHECK: gpu.launch_func @extra_constants_kernel::@extra_constants_kernel blocks in (%[[CST]], %[[CST]], %[[CST]]) threads in (%[[CST]], %[[CST]], %[[CST]]) args(%[[ARG0]] : memref<?xf32>)
gpu.launch blocks(%bx, %by, %bz) in (%grid_x = %cst, %grid_y = %cst,
%grid_z = %cst)
@@ -134,7 +134,7 @@ func @extra_constants(%arg0: memref<?xf32>) {
// CHECK-SAME: %[[KARG0:.*]]: memref<?xf32>
// CHECK: constant 2
// CHECK: constant 0
-// CHECK: dim %[[KARG0]]
+// CHECK: memref.dim %[[KARG0]]
// -----
@@ -145,8 +145,8 @@ func @extra_constants_noarg(%arg0: memref<?xf32>, %arg1: memref<?xf32>) {
%cst = constant 8 : index
%cst2 = constant 2 : index
%c0 = constant 0 : index
- // CHECK: dim %[[ARG1]]
- %cst3 = dim %arg1, %c0 : memref<?xf32>
+ // CHECK: memref.dim %[[ARG1]]
+ %cst3 = memref.dim %arg1, %c0 : memref<?xf32>
// CHECK: gpu.launch_func @extra_constants_noarg_kernel::@extra_constants_noarg_kernel blocks in (%[[CST]], %[[CST]], %[[CST]]) threads in (%[[CST]], %[[CST]], %[[CST]]) args(%[[ARG0]] : memref<?xf32>, {{.*}} : index)
gpu.launch blocks(%bx, %by, %bz) in (%grid_x = %cst, %grid_y = %cst,
%grid_z = %cst)
@@ -192,10 +192,10 @@ func @multiple_uses(%arg0 : memref<?xf32>) {
func @multiple_uses2(%arg0 : memref<*xf32>) {
%c1 = constant 1 : index
%c2 = constant 2 : index
- %d = dim %arg0, %c2 : memref<*xf32>
+ %d = memref.dim %arg0, %c2 : memref<*xf32>
// CHECK: gpu.func {{.*}} {
// CHECK: %[[C2:.*]] = constant 2 : index
- // CHECK: %[[D:.*]] = dim %[[ARG:.*]], %[[C2]]
+ // CHECK: %[[D:.*]] = memref.dim %[[ARG:.*]], %[[C2]]
// CHECK: "use1"(%[[D]])
// CHECK: "use2"(%[[C2]], %[[C2]])
// CHECK: "use3"(%[[ARG]])
diff --git a/mlir/test/Dialect/GPU/promotion.mlir b/mlir/test/Dialect/GPU/promotion.mlir
index 9df7bf19dff6..a9a08eda55d6 100644
--- a/mlir/test/Dialect/GPU/promotion.mlir
+++ b/mlir/test/Dialect/GPU/promotion.mlir
@@ -26,7 +26,7 @@ gpu.module @foo {
// CHECK: scf.for %[[i2:.*]] =
// Verify that the copy is emitted and uses only the last two loops.
- // CHECK: %[[v:.*]] = load %[[arg]][%[[i1]], %[[i2]]]
+ // CHECK: %[[v:.*]] = memref.load %[[arg]][%[[i1]], %[[i2]]]
// CHECK: store %[[v]], %[[promoted]][%[[i1]], %[[i2]]]
// Verify that the use has been rewritten.
@@ -42,7 +42,7 @@ gpu.module @foo {
// CHECK: scf.for %[[i2:.*]] =
// Verify that the copy is emitted and uses only the last two loops.
- // CHECK: %[[v:.*]] = load %[[promoted]][%[[i1]], %[[i2]]]
+ // CHECK: %[[v:.*]] = memref.load %[[promoted]][%[[i1]], %[[i2]]]
// CHECK: store %[[v]], %[[arg]][%[[i1]], %[[i2]]]
gpu.return
}
@@ -80,7 +80,7 @@ gpu.module @foo {
// CHECK: scf.for %[[i4:.*]] =
// Verify that the copy is emitted.
- // CHECK: %[[v:.*]] = load %[[arg]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
+ // CHECK: %[[v:.*]] = memref.load %[[arg]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
// CHECK: store %[[v]], %[[promoted]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
// Verify that the use has been rewritten.
@@ -95,7 +95,7 @@ gpu.module @foo {
// CHECK: scf.for %[[i4:.*]] =
// Verify that the copy is emitted.
- // CHECK: %[[v:.*]] = load %[[promoted]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
+ // CHECK: %[[v:.*]] = memref.load %[[promoted]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
// CHECK: store %[[v]], %[[arg]][%[[i0]], %[[i1]], %[[i2]], %[[i3]], %[[i4]]]
gpu.return
}
diff --git a/mlir/test/Dialect/Linalg/affine.mlir b/mlir/test/Dialect/Linalg/affine.mlir
index 0df7db06e4c4..4de404db7afd 100644
--- a/mlir/test/Dialect/Linalg/affine.mlir
+++ b/mlir/test/Dialect/Linalg/affine.mlir
@@ -12,9 +12,9 @@
func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %A = view %arg0[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
- %B = view %arg0[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
- %C = view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
+ %A = memref.view %arg0[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
+ %B = memref.view %arg0[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
+ %C = memref.view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
linalg.matmul ins(%A, %B: memref<?x?xf32>, memref<?x?xf32>)
outs(%C: memref<?x?xf32>)
return
@@ -24,9 +24,9 @@ func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECK-SAME: [[M:arg[0-9]+]]: index
// CHECK-SAME: [[N:arg[0-9]+]]: index
// CHECK-SAME: [[K:arg[0-9]+]]: index
-// CHECK: %[[A:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[B:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[C:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[A:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[B:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[C:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
// CHECK: affine.for %{{.*}} = 0 to %{{.*}} {
@@ -44,11 +44,11 @@ func @conv_view3(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1:
// CHECK-LABEL: func @conv_view3(
// CHECK: %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>) {
-// CHECK: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECK: %[[Q:.*]] = dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECK: %[[K:.*]] = dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECK: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECK: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECK: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECK: %[[Q:.*]] = memref.dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECK: %[[K:.*]] = memref.dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECK: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECK: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
// CHECK: affine.for %{{.*}} = 0 to %[[B]] {
// CHECK: affine.for %{{.*}} = 0 to %[[X0]] {
// CHECK: affine.for %{{.*}} = 0 to %[[K]] {
@@ -71,13 +71,13 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32>,
// CHECK-LABEL: func @conv_padding
// CHECK: %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>) {
// CHECK: %[[ZERO:.*]] = constant 0.000000e+00 : f32
-// CHECK: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?x?xf32>
-// CHECK: %[[Z1:.*]] = dim %arg0, %c1 : memref<?x?x?x?xf32>
-// CHECK: %[[Q:.*]] = dim %arg0, %c2 : memref<?x?x?x?xf32>
-// CHECK: %[[K:.*]] = dim %arg0, %c3 : memref<?x?x?x?xf32>
-// CHECK: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?x?xf32>
-// CHECK: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?x?xf32>
-// CHECK: %[[X1:.*]] = dim %arg2, %c2 : memref<?x?x?x?xf32>
+// CHECK: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?x?xf32>
+// CHECK: %[[Z1:.*]] = memref.dim %arg0, %c1 : memref<?x?x?x?xf32>
+// CHECK: %[[Q:.*]] = memref.dim %arg0, %c2 : memref<?x?x?x?xf32>
+// CHECK: %[[K:.*]] = memref.dim %arg0, %c3 : memref<?x?x?x?xf32>
+// CHECK: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?x?xf32>
+// CHECK: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?x?xf32>
+// CHECK: %[[X1:.*]] = memref.dim %arg2, %c2 : memref<?x?x?x?xf32>
// CHECK: affine.for %{{.*}} = 0 to %[[B]] {
// CHECK: affine.for %{{.*}} = 0 to %[[X0]] {
// CHECK: affine.for %{{.*}} = 0 to %[[X1]] {
@@ -90,8 +90,8 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32>,
// CHECK: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[SUM0]])
// CHECK: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[SUM1]])
// Padded conv involves an affine.max in the memory access and this is not
-// allowed by affine.load. Use std.load in such cases.
-// CHECK: %{{.*}} = load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
+// allowed by affine.load. Use memref.load in such cases.
+// CHECK: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
// CHECK: %{{.*}} = select %{{.*}}, %{{.*}}, %{{.*}} : f32
// CHECK: %{{.*}} = affine.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECK: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
@@ -111,10 +111,10 @@ func @named_batch_matmul(%A: memref<?x?x?xf32>, %B: memref<?x?x?xf32>, %C: memre
// CHECK-SAME: %[[mA:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECK-SAME: %[[mB:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECK-SAME: %[[mC:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
-// CHECK: %[[B:.*]] = dim %[[mA]], %c0 : memref<?x?x?xf32>
-// CHECK: %[[M:.*]] = dim %[[mA]], %c1 : memref<?x?x?xf32>
-// CHECK: %[[K:.*]] = dim %[[mA]], %c2 : memref<?x?x?xf32>
-// CHECK: %[[N:.*]] = dim %[[mB]], %c2 : memref<?x?x?xf32>
+// CHECK: %[[B:.*]] = memref.dim %[[mA]], %c0 : memref<?x?x?xf32>
+// CHECK: %[[M:.*]] = memref.dim %[[mA]], %c1 : memref<?x?x?xf32>
+// CHECK: %[[K:.*]] = memref.dim %[[mA]], %c2 : memref<?x?x?xf32>
+// CHECK: %[[N:.*]] = memref.dim %[[mB]], %c2 : memref<?x?x?xf32>
// CHECK: affine.for %[[b:.*]] = 0 to %[[B]] {
// CHECK: affine.for %[[m:.*]] = 0 to %[[M]] {
// CHECK: affine.for %[[n:.*]] = 0 to %[[N]] {
diff --git a/mlir/test/Dialect/Linalg/bufferize.mlir b/mlir/test/Dialect/Linalg/bufferize.mlir
index 1b5b047b9bdd..e0027b765d25 100644
--- a/mlir/test/Dialect/Linalg/bufferize.mlir
+++ b/mlir/test/Dialect/Linalg/bufferize.mlir
@@ -3,7 +3,7 @@
#map0 = affine_map<(d0) -> (d0)>
// In-depth checking of a basic case, this is testing
-// - tensor_to_memref / tensor_load materializations are properly inserted
+// - memref.buffer_cast / memref.tensor_load materializations are properly inserted
// - payload is correctly carried over
// - affine maps are correctly carried over
// Later tests will not check all these details.
@@ -11,8 +11,8 @@
// CHECK: #map = affine_map<(d0) -> (d0)>
// CHECK-LABEL: func @basic(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<4xf32>) -> tensor<4xf32> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<4xf32>
-// CHECK: %[[RESULT_MEMREF:.*]] = alloc() : memref<4xf32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<4xf32>
+// CHECK: %[[RESULT_MEMREF:.*]] = memref.alloc() : memref<4xf32>
// CHECK: linalg.generic {indexing_maps = [#map, #map], iterator_types = ["parallel"]}
// CHECK-SAME: ins(%[[MEMREF]] : memref<4xf32>)
// CHECK-SAME: outs(%[[RESULT_MEMREF]] : memref<4xf32>) {
@@ -20,7 +20,7 @@
// CHECK: %[[DIM1:.*]] = math.exp %[[RESULT1]] : f32
// CHECK: linalg.yield %[[DIM1]] : f32
// CHECK: }
-// CHECK: %[[RESULT:.*]] = tensor_load %[[RESULT_MEMREF]] : memref<4xf32>
+// CHECK: %[[RESULT:.*]] = memref.tensor_load %[[RESULT_MEMREF]] : memref<4xf32>
// CHECK: return %[[RESULT]] : tensor<4xf32>
func @basic(%arg0: tensor<4xf32>) -> tensor<4xf32> {
%0 = linalg.generic {
@@ -45,8 +45,8 @@ func @basic(%arg0: tensor<4xf32>) -> tensor<4xf32> {
// CHECK: #map = affine_map<(d0) -> (d0)>
// CHECK-LABEL: func @init_tensor(
// CHECK-SAME: %[[IN:.*]]: tensor<?xf32>, %[[SIZE:.*]]: index)
-// CHECK: %[[OUT_BUF:.*]] = alloc(%[[SIZE]]) : memref<?xf32>
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[IN]] : memref<?xf32>
+// CHECK: %[[OUT_BUF:.*]] = memref.alloc(%[[SIZE]]) : memref<?xf32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[IN]] : memref<?xf32>
// CHECK: linalg.generic
// CHECK-SAME: ins(%[[MEMREF]] : memref<?xf32>)
// CHECK-SAME: outs(%[[OUT_BUF]] : memref<?xf32>) {
@@ -70,8 +70,8 @@ func @init_tensor(%in : tensor<?xf32>, %size: index) -> tensor<?xf32> {
#map0 = affine_map<(d0) -> (d0)>
// CHECK-LABEL: func @multiple_results
-// CHECK: %[[RESULT0:.*]] = alloc() : memref<4xf32>
-// CHECK: %[[RESULT1:.*]] = alloc() : memref<4xf32>
+// CHECK: %[[RESULT0:.*]] = memref.alloc() : memref<4xf32>
+// CHECK: %[[RESULT1:.*]] = memref.alloc() : memref<4xf32>
// CHECK: linalg.generic
// CHECK-SAME: ins(%{{.*}} : memref<4xf32>)
// CHECK-SAME: outs(%[[RESULT0]], %[[RESULT1]] : memref<4xf32>, memref<4xf32>)
@@ -94,8 +94,8 @@ func @multiple_results(%arg0: tensor<4xf32>) -> (tensor<4xf32>, tensor<4xf32>) {
#map0 = affine_map<(d0) -> (d0)>
// CHECK-LABEL: func @multiple_results_indexed
-// CHECK: %[[RESULT0:.*]] = alloc() : memref<4xi32>
-// CHECK: %[[RESULT1:.*]] = alloc() : memref<4xi32>
+// CHECK: %[[RESULT0:.*]] = memref.alloc() : memref<4xi32>
+// CHECK: %[[RESULT1:.*]] = memref.alloc() : memref<4xi32>
// CHECK: linalg.indexed_generic
// CHECK-SAME: ins(%{{.*}} : memref<4xi32>)
// CHECK-SAME: outs(%[[RESULT0]], %[[RESULT1]] : memref<4xi32>, memref<4xi32>)
@@ -126,11 +126,11 @@ func @multiple_results_indexed(%arg0: tensor<4xi32>)
// CHECK-SAME: %[[ARG:.*]]: tensor<?x?xf32>
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[MEMREF_ARG:.*]] = tensor_to_memref %[[ARG]] : memref<?x?xf32>
-// CHECK: %[[DIM0:.*]] = dim %[[ARG]], %[[C0]] : tensor<?x?xf32>
-// CHECK: %[[DIM1:.*]] = dim %[[ARG]], %[[C1]] : tensor<?x?xf32>
-// CHECK: %[[RESULT0:.*]] = alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
-// CHECK: %[[RESULT1:.*]] = alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
+// CHECK: %[[MEMREF_ARG:.*]] = memref.buffer_cast %[[ARG]] : memref<?x?xf32>
+// CHECK: %[[DIM0:.*]] = memref.dim %[[ARG]], %[[C0]] : tensor<?x?xf32>
+// CHECK: %[[DIM1:.*]] = memref.dim %[[ARG]], %[[C1]] : tensor<?x?xf32>
+// CHECK: %[[RESULT0:.*]] = memref.alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
+// CHECK: %[[RESULT1:.*]] = memref.alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
// CHECK: linalg.generic
// CHECK-SAME: ins(%[[MEMREF_ARG]] : memref<?x?xf32>)
// CHECK-SAME: outs(%[[RESULT0]], %[[RESULT1]] : memref<?x?xf32>, memref<?x?xf32>)
@@ -165,9 +165,9 @@ func @dynamic_results(%arg0: tensor<?x?xf32>)
// CHECK-LABEL: func @generic_with_init_tensor(
// CHECK-SAME: %[[ARG0_TENSOR:.*]]: tensor<2x3x4xvector<3x4xi4>>,
// CHECK-SAME: %[[ARG1_TENSOR:.*]]: tensor<3x2xf32>) -> tensor<3x2xf32> {
-// CHECK: %[[ARG0_MEMREF:.*]] = tensor_to_memref %[[ARG0_TENSOR]] : memref<2x3x4xvector<3x4xi4>>
-// CHECK: %[[ARG1_MEMREF:.*]] = tensor_to_memref %[[ARG1_TENSOR]] : memref<3x2xf32>
-// CHECK: %[[INIT_BUFFER:.*]] = alloc() : memref<3x2xf32>
+// CHECK: %[[ARG0_MEMREF:.*]] = memref.buffer_cast %[[ARG0_TENSOR]] : memref<2x3x4xvector<3x4xi4>>
+// CHECK: %[[ARG1_MEMREF:.*]] = memref.buffer_cast %[[ARG1_TENSOR]] : memref<3x2xf32>
+// CHECK: %[[INIT_BUFFER:.*]] = memref.alloc() : memref<3x2xf32>
// CHECK: linalg.copy(%[[ARG1_MEMREF]], %[[INIT_BUFFER]]) : memref<3x2xf32>, memref<3x2xf32>
// CHECK: linalg.generic
// CHECK-SAME: ins(%[[ARG0_MEMREF]] : memref<2x3x4xvector<3x4xi4>>)
@@ -198,20 +198,20 @@ func @bufferize_subtensor(%t : tensor<?x?xf32>) -> (tensor<2x3xf32>, tensor<2x?x
// CHECK: %[[IDX:.*]] = call @make_index() : () -> index
%i0 = call @make_index() : () -> index
- // CHECK: %[[M0:.*]] = tensor_to_memref %[[T]] : memref<?x?xf32>
- // CHECK-NEXT: %[[A0:.*]] = alloc() : memref<2x3xf32>
- // CHECK-NEXT: %[[SM0:.*]] = subview %[[M0]][0, 0] [2, 3] [1, 1]
+ // CHECK: %[[M0:.*]] = memref.buffer_cast %[[T]] : memref<?x?xf32>
+ // CHECK-NEXT: %[[A0:.*]] = memref.alloc() : memref<2x3xf32>
+ // CHECK-NEXT: %[[SM0:.*]] = memref.subview %[[M0]][0, 0] [2, 3] [1, 1]
// CHECK-SAME: memref<?x?xf32> to memref<2x3xf32, #[[$MAP0]]>
// CHECK-NEXT: linalg.copy(%[[SM0]], %[[A0]]) : memref<2x3xf32, #[[$MAP0]]>, memref<2x3xf32>
- // CHECK-NEXT: %[[RT0:.*]] = tensor_load %[[A0]] : memref<2x3xf32>
+ // CHECK-NEXT: %[[RT0:.*]] = memref.tensor_load %[[A0]] : memref<2x3xf32>
%st0 = subtensor %t[0, 0][2, 3][1, 1] : tensor<?x?xf32> to tensor<2x3xf32>
- // CHECK: %[[M1:.*]] = tensor_to_memref %[[T]] : memref<?x?xf32>
- // CHECK-NEXT: %[[A1:.*]] = alloc(%[[IDX]]) : memref<2x?xf32>
- // CHECK-NEXT: %[[SM1:.*]] = subview %[[M1]][0, %[[IDX]]] [2, %[[IDX]]] [1, 2]
+ // CHECK: %[[M1:.*]] = memref.buffer_cast %[[T]] : memref<?x?xf32>
+ // CHECK-NEXT: %[[A1:.*]] = memref.alloc(%[[IDX]]) : memref<2x?xf32>
+ // CHECK-NEXT: %[[SM1:.*]] = memref.subview %[[M1]][0, %[[IDX]]] [2, %[[IDX]]] [1, 2]
// CHECK-SAME: memref<?x?xf32> to memref<2x?xf32, #[[$MAP1]]>
// CHECK-NEXT: linalg.copy(%[[SM1]], %[[A1]]) : memref<2x?xf32, #[[$MAP1]]>, memref<2x?xf32>
- // CHECK-NEXT: %[[RT1:.*]] = tensor_load %[[A1]] : memref<2x?xf32>
+ // CHECK-NEXT: %[[RT1:.*]] = memref.tensor_load %[[A1]] : memref<2x?xf32>
%st1 = subtensor %t[0, %i0][2, %i0][1, 2] : tensor<?x?xf32> to tensor<2x?xf32>
// CHECK-NEXT: return %[[RT0]], %[[RT1]]
@@ -239,26 +239,26 @@ func @bufferize_subtensor_insert(%t : tensor<?x?xf32>, %st0 : tensor<2x3xf32>, %
// CHECK: %[[IDX:.*]] = call @make_index() : () -> index
- // CHECK-DAG: %[[M0:.*]] = tensor_to_memref %[[T]] : memref<?x?xf32>
- // CHECK-DAG: %[[SM0:.*]] = tensor_to_memref %[[ST0]] : memref<2x3xf32>
- // CHECK-NEXT: %[[DIM0:.*]] = dim %[[T]], %[[C0]] : tensor<?x?xf32>
- // CHECK-NEXT: %[[DIM1:.*]] = dim %[[T]], %[[C1]] : tensor<?x?xf32>
- // CHECK-NEXT: %[[M0_COPY:.*]] = alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
+ // CHECK-DAG: %[[M0:.*]] = memref.buffer_cast %[[T]] : memref<?x?xf32>
+ // CHECK-DAG: %[[SM0:.*]] = memref.buffer_cast %[[ST0]] : memref<2x3xf32>
+ // CHECK-NEXT: %[[DIM0:.*]] = memref.dim %[[T]], %[[C0]] : tensor<?x?xf32>
+ // CHECK-NEXT: %[[DIM1:.*]] = memref.dim %[[T]], %[[C1]] : tensor<?x?xf32>
+ // CHECK-NEXT: %[[M0_COPY:.*]] = memref.alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
// CHECK-NEXT: linalg.copy(%[[M0]], %[[M0_COPY]]) : memref<?x?xf32>, memref<?x?xf32>
- // CHECK-NEXT: %[[SUBVIEW0:.*]] = subview %[[M0_COPY]][0, 0] [2, 3] [1, 1]
+ // CHECK-NEXT: %[[SUBVIEW0:.*]] = memref.subview %[[M0_COPY]][0, 0] [2, 3] [1, 1]
// CHECK-SAME: memref<?x?xf32> to memref<2x3xf32, #[[$MAP0]]>
// CHECK-NEXT: linalg.copy(%[[SM0]], %[[SUBVIEW0]]) : memref<2x3xf32>, memref<2x3xf32, #[[$MAP0]]>
- // CHECK-NEXT: %[[RT0:.*]] = tensor_load %[[M0_COPY]] : memref<?x?xf32>
+ // CHECK-NEXT: %[[RT0:.*]] = memref.tensor_load %[[M0_COPY]] : memref<?x?xf32>
%t0 = subtensor_insert %st0 into %t[0, 0][2, 3][1, 1] : tensor<2x3xf32> into tensor<?x?xf32>
- // CHECK-DAG: %[[M1:.*]] = tensor_to_memref %[[T]] : memref<?x?xf32>
- // CHECK-DAG: %[[SM1:.*]] = tensor_to_memref %[[ST1]] : memref<2x?xf32>
- // CHECK-NEXT: %[[M1_COPY:.*]] = alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
+ // CHECK-DAG: %[[M1:.*]] = memref.buffer_cast %[[T]] : memref<?x?xf32>
+ // CHECK-DAG: %[[SM1:.*]] = memref.buffer_cast %[[ST1]] : memref<2x?xf32>
+ // CHECK-NEXT: %[[M1_COPY:.*]] = memref.alloc(%[[DIM0]], %[[DIM1]]) : memref<?x?xf32>
// CHECK-NEXT: linalg.copy(%[[M1]], %[[M1_COPY]]) : memref<?x?xf32>, memref<?x?xf32>
- // CHECK-NEXT: %[[SUBVIEW1:.*]] = subview %[[M1_COPY]][0, %[[IDX]]] [2, %[[IDX]]] [1, 2]
+ // CHECK-NEXT: %[[SUBVIEW1:.*]] = memref.subview %[[M1_COPY]][0, %[[IDX]]] [2, %[[IDX]]] [1, 2]
// CHECK-SAME: memref<?x?xf32> to memref<2x?xf32, #[[$MAP1]]>
// CHECK-NEXT: linalg.copy(%[[SM1]], %[[SUBVIEW1]]) : memref<2x?xf32>, memref<2x?xf32, #[[$MAP1]]>
- // CHECK-NEXT: %[[RT1:.*]] = tensor_load %[[M1_COPY]] : memref<?x?xf32>
+ // CHECK-NEXT: %[[RT1:.*]] = memref.tensor_load %[[M1_COPY]] : memref<?x?xf32>
%t1 = subtensor_insert %st1 into %t[0, %i0][2, %i0][1, 2] : tensor<2x?xf32> into tensor<?x?xf32>
// CHECK: return %[[RT0]], %[[RT1]]
diff --git a/mlir/test/Dialect/Linalg/canonicalize.mlir b/mlir/test/Dialect/Linalg/canonicalize.mlir
index 1ab790ad84de..693e94f63698 100644
--- a/mlir/test/Dialect/Linalg/canonicalize.mlir
+++ b/mlir/test/Dialect/Linalg/canonicalize.mlir
@@ -6,9 +6,9 @@ func @memref_cast(%a: index, %b: index) -> memref<?x?xf32> {
%c1 = constant 1 : index
%c8 = constant 8 : index
%c16 = constant 16 : index
- %1 = alloc (%b) : memref<?xi8>
- %2 = view %1[%c0][] : memref<?xi8> to memref<16x16xf32>
- %3 = memref_cast %2 : memref<16x16xf32> to memref<?x?xf32>
+ %1 = memref.alloc (%b) : memref<?xi8>
+ %2 = memref.view %1[%c0][] : memref<?xi8> to memref<16x16xf32>
+ %3 = memref.cast %2 : memref<16x16xf32> to memref<?x?xf32>
// CHECK: linalg.matmul ins({{.*}}memref<16x16xf32>, memref<16x16xf32>) outs({{.*}}memref<16x16xf32>)
linalg.matmul ins(%3, %3: memref<?x?xf32>, memref<?x?xf32>)
@@ -339,11 +339,12 @@ func @linalg_effects(%a : tensor<?x?xf32>, %b : memref<?x?xf32>, %c : tensor<?x?
%t = linalg.matmul ins(%a, %b : tensor<?x?xf32>, memref<?x?xf32>)
outs(%c : tensor<?x?xf32>) -> tensor<?x?xf32>
- // CHECK-NOT: %{{.*}} = linalg.matmul
+ // CHECK: linalg.matmul
linalg.matmul ins(%a, %c : tensor<?x?xf32>, tensor<?x?xf32>)
outs(%b : memref<?x?xf32>)
return
}
+
// -----
func @init_tensor_canonicalize() -> (tensor<4x5x?xf32>) {
@@ -363,8 +364,8 @@ func @init_tensor_static_dim() -> (index, index) {
%c2 = constant 2 : index
%c6 = constant 6 : index
%0 = linalg.init_tensor [4, 5, %c6] : tensor<4x5x?xf32>
- %1 = dim %0, %c2 : tensor<4x5x?xf32>
- %2 = dim %0, %c0 : tensor<4x5x?xf32>
+ %1 = memref.dim %0, %c2 : tensor<4x5x?xf32>
+ %2 = memref.dim %0, %c0 : tensor<4x5x?xf32>
return %1, %2 : index, index
}
// CHECK: func @init_tensor_static_dim
@@ -377,7 +378,7 @@ func @init_tensor_static_dim() -> (index, index) {
func @init_tensor_dynamic_dim(%arg0 : index) -> (index) {
%c2 = constant 2 : index
%0 = linalg.init_tensor [4, 5, %arg0] : tensor<4x5x?xf32>
- %1 = dim %0, %c2 : tensor<4x5x?xf32>
+ %1 = memref.dim %0, %c2 : tensor<4x5x?xf32>
return %1 : index
}
// CHECK: func @init_tensor_dynamic_dim
@@ -390,8 +391,8 @@ func @init_tensor_dynamic_dim2(%arg0 : index, %arg1 : index) -> (index, index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%0 = linalg.init_tensor [%arg0, %arg1] : tensor<?x?xf32>
- %1 = dim %0, %c0 : tensor<?x?xf32>
- %2 = dim %0, %c1 : tensor<?x?xf32>
+ %1 = memref.dim %0, %c0 : tensor<?x?xf32>
+ %2 = memref.dim %0, %c1 : tensor<?x?xf32>
return %1, %2 : index, index
}
// CHECK: func @init_tensor_dynamic_dim2
@@ -417,7 +418,7 @@ func @remove_dim_result_uses
%2 = addf %1, %arg5 : f32
linalg.yield %2 : f32
} -> tensor<?x?xf32>
- %3 = dim %0, %c0 : tensor<?x?xf32>
+ %3 = memref.dim %0, %c0 : tensor<?x?xf32>
return %3 : index
}
// CHECK: #[[MAP:.+]] = affine_map<()[s0, s1] -> (s0 + s1)>
@@ -427,8 +428,8 @@ func @remove_dim_result_uses
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: tensor<?x?xf32>
// CHECK-DAG: %[[C0:.+]] = constant 0 : index
// CHECK-DAG: %[[C1:.+]] = constant 1 : index
-// CHECK-DAG: %[[T0:.+]] = dim %[[ARG0]], %[[C0]]
-// CHECK-DAG: %[[T1:.+]] = dim %[[ARG1]], %[[C1]]
+// CHECK-DAG: %[[T0:.+]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[T1:.+]] = memref.dim %[[ARG1]], %[[C1]]
// CHECK: %[[T2:.+]] = affine.apply #[[MAP]]()[%[[T0]], %[[T1]]]
// CHECK: return %[[T2]]
@@ -438,7 +439,7 @@ func @remove_dim_result_uses_outs
(%arg0 : tensor<?xf32>, %arg1 : index) -> (index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %d0 = dim %arg0, %c0 : tensor<?xf32>
+ %d0 = memref.dim %arg0, %c0 : tensor<?xf32>
%0 = linalg.init_tensor [%d0, %arg1] : tensor<?x?xf32>
%1 = linalg.generic
{indexing_maps = [affine_map<(d0, d1) -> (d0)>,
@@ -448,7 +449,7 @@ func @remove_dim_result_uses_outs
^bb0(%arg2: f32, %arg3: f32) :
linalg.yield %arg2 : f32
} -> tensor<?x?xf32>
- %2 = dim %1, %c1 : tensor<?x?xf32>
+ %2 = memref.dim %1, %c1 : tensor<?x?xf32>
return %2 : index
}
// CHECK: func @remove_dim_result_uses_outs
@@ -464,8 +465,8 @@ func @remove_dim_result_uses_sequence
%c1 = constant 1 : index
%0 = linalg.matmul ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
outs(%arg2 : tensor<?x?xf32>) -> tensor<?x?xf32>
- %1 = dim %0, %c0 : tensor<?x?xf32>
- %2 = dim %0, %c1 : tensor<?x?xf32>
+ %1 = memref.dim %0, %c0 : tensor<?x?xf32>
+ %2 = memref.dim %0, %c1 : tensor<?x?xf32>
%3 = linalg.generic
{indexing_maps = [affine_map<(d0, d1, d2) -> (d1, d0)>,
affine_map<(d0, d1, d2) -> (d0, d2)>,
@@ -478,8 +479,8 @@ func @remove_dim_result_uses_sequence
%5 = addf %4, %arg5 : f32
linalg.yield %5 : f32
} -> tensor<?x?xf32>
- %6 = dim %3, %c0 : tensor<?x?xf32>
- %7 = dim %3, %c1 : tensor<?x?xf32>
+ %6 = memref.dim %3, %c0 : tensor<?x?xf32>
+ %7 = memref.dim %3, %c1 : tensor<?x?xf32>
return %1, %2, %6, %7 : index, index, index, index
}
// CHECK-LABEL: func @remove_dim_result_uses_sequence
@@ -488,10 +489,10 @@ func @remove_dim_result_uses_sequence
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: tensor<?x?xf32>
// CHECK-DAG: %[[C0:.+]] = constant 0 : index
// CHECK-DAG: %[[C1:.+]] = constant 1 : index
-// CHECK-DAG: %[[T0:.+]] = dim %[[ARG0]], %[[C0]]
-// CHECK-DAG: %[[T1:.+]] = dim %[[ARG1]], %[[C1]]
-// CHECK-DAG: %[[T2:.+]] = dim %[[ARG0]], %[[C1]]
-// CHECK-DAG: %[[T3:.+]] = dim %[[ARG1]], %[[C1]]
+// CHECK-DAG: %[[T0:.+]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[T1:.+]] = memref.dim %[[ARG1]], %[[C1]]
+// CHECK-DAG: %[[T2:.+]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK-DAG: %[[T3:.+]] = memref.dim %[[ARG1]], %[[C1]]
// CHECK: return %[[T0]], %[[T1]], %[[T2]], %[[T3]]
// -----
@@ -500,7 +501,7 @@ func @keep_result_dim_uses_sequence2
(%arg0 : tensor<?xf32>, %arg1 : index) -> (index, index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %d0 = dim %arg0, %c0 : tensor<?xf32>
+ %d0 = memref.dim %arg0, %c0 : tensor<?xf32>
%0 = linalg.init_tensor [%d0, %arg1] : tensor<?x?xf32>
%1 = linalg.generic
{indexing_maps = [affine_map<(d0, d1) -> (d0)>,
@@ -510,15 +511,15 @@ func @keep_result_dim_uses_sequence2
^bb0(%arg2: f32, %arg3 : f32):
linalg.yield %arg2 : f32
} -> tensor<?x?xf32>
- %2 = dim %1, %c0 : tensor<?x?xf32>
- %3 = dim %1, %c1 : tensor<?x?xf32>
+ %2 = memref.dim %1, %c0 : tensor<?x?xf32>
+ %3 = memref.dim %1, %c1 : tensor<?x?xf32>
return %2, %3 : index, index
}
// CHECK: func @keep_result_dim_uses_sequence2
// CHECK-SAME: %[[ARG0:[a-zA-Z0-9_]+]]: tensor<?xf32>
// CHECK-SAME: %[[ARG1:[a-zA-Z0-9_]+]]: index
// CHECK-DAG: %[[C0:.+]] = constant 0 : index
-// CHECK-DAG: %[[T0:.+]] = dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[T0:.+]] = memref.dim %[[ARG0]], %[[C0]]
// CHECK: return %[[T0]], %[[ARG1]]
// -----
@@ -537,16 +538,16 @@ func @init_tensor_dim_of_linalg_result(%arg_0 : tensor<?xf32>,
} -> tensor<?xf32>, tensor<?xf32>
%c0 = constant 0 : index
- %num_elem_0 = dim %0, %c0 : tensor<?xf32>
+ %num_elem_0 = memref.dim %0, %c0 : tensor<?xf32>
- %num_elem_1 = dim %1, %c0 : tensor<?xf32>
+ %num_elem_1 = memref.dim %1, %c0 : tensor<?xf32>
return %num_elem_0, %num_elem_1 : index, index
}
// CHECK: func @init_tensor_dim_of_linalg_result(
// CHECK-SAME: %[[ARG_0:[a-zA-Z0-9_]+]]: tensor<?xf32>
// CHECK-SAME: %[[ARG_1:[a-zA-Z0-9_]+]]: tensor<?xf32>)
-// CHECK: %[[R0:.+]] = dim %[[ARG_0]]
-// CHECK: %[[R1:.+]] = dim %[[ARG_0]]
+// CHECK: %[[R0:.+]] = memref.dim %[[ARG_0]]
+// CHECK: %[[R1:.+]] = memref.dim %[[ARG_0]]
// CHECK: return %[[R0]], %[[R1]]
// -----
@@ -593,9 +594,9 @@ func @remove_no_op(%arg0 : tensor<?x?x?xf32>, %arg1 : tensor<?x?x?xf32>)
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %arg0, %c0 : tensor<?x?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?x?xf32>
- %2 = dim %arg0, %c2 : tensor<?x?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?x?xf32>
+ %2 = memref.dim %arg0, %c2 : tensor<?x?x?xf32>
%3 = linalg.init_tensor [%0, %1, %2] : tensor<?x?x?xf32>
%4, %5 = linalg.generic {
indexing_maps = [#map, #map, #map, #map],
@@ -619,8 +620,8 @@ func @keep_not_noop(%arg0 : tensor<?x?xf32>) -> tensor<?x?xf32> {
%c0 = constant 0 : index
%c1 = constant 1 : index
%cst = constant 1.000000e+00 : f32
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xf32>
br ^bb1(%cst : f32)
@@ -645,8 +646,8 @@ func @keep_not_noop(%arg0 : tensor<?x?xf32>, %arg1 : tensor<?x?xf32>)
%c0 = constant 0 : index
%c1 = constant 1 : index
%cst = constant 1.000000e+00 : f32
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xf32>
br ^bb1(%cst : f32)
@@ -729,9 +730,9 @@ func @dim_reshape_expansion(%arg0 : tensor<6x5x?xf32>) -> (index, index, index)
affine_map<(d0, d1, d2, d3, d4, d5) -> (d2)>,
affine_map<(d0, d1, d2, d3, d4, d5) -> (d3, d4, d5)>] :
tensor<6x5x?xf32> into tensor<2x3x5x4x?x7xf32>
- %1 = dim %0, %c1 : tensor<2x3x5x4x?x7xf32>
- %2 = dim %0, %c3 : tensor<2x3x5x4x?x7xf32>
- %3 = dim %0, %c4 : tensor<2x3x5x4x?x7xf32>
+ %1 = memref.dim %0, %c1 : tensor<2x3x5x4x?x7xf32>
+ %2 = memref.dim %0, %c3 : tensor<2x3x5x4x?x7xf32>
+ %3 = memref.dim %0, %c4 : tensor<2x3x5x4x?x7xf32>
return %1, %2, %3 : index, index, index
}
// CHECK: #[[MAP:.+]] = affine_map<()[s0] -> (s0 floordiv 28)>
@@ -740,7 +741,7 @@ func @dim_reshape_expansion(%arg0 : tensor<6x5x?xf32>) -> (index, index, index)
// CHECK-DAG: %[[C2:.+]] = constant 2 : index
// CHECK-DAG: %[[C3:.+]] = constant 3 : index
// CHECK-DAG: %[[C4:.+]] = constant 4 : index
-// CHECK: %[[D0:.+]] = dim %[[ARG0]], %[[C2]]
+// CHECK: %[[D0:.+]] = memref.dim %[[ARG0]], %[[C2]]
// CHECK: %[[D1:.+]] = affine.apply #[[MAP]]()[%[[D0]]]
// CHECK: return %[[C3]], %[[C4]], %[[D1]]
@@ -755,8 +756,8 @@ func @dim_reshape_collapse(%arg0 : tensor<2x3x5x4x?x7xf32>) -> (index, index)
affine_map<(d0, d1, d2, d3, d4, d5) -> (d2)>,
affine_map<(d0, d1, d2, d3, d4, d5) -> (d3, d4, d5)>] :
tensor<2x3x5x4x?x7xf32> into tensor<6x5x?xf32>
- %1 = dim %0, %c1 : tensor<6x5x?xf32>
- %2 = dim %0, %c2 : tensor<6x5x?xf32>
+ %1 = memref.dim %0, %c1 : tensor<6x5x?xf32>
+ %2 = memref.dim %0, %c2 : tensor<6x5x?xf32>
return %1, %2 : index, index
}
// CHECK: #[[MAP:.+]] = affine_map<()[s0] -> (s0 * 28)>
@@ -764,7 +765,7 @@ func @dim_reshape_collapse(%arg0 : tensor<2x3x5x4x?x7xf32>) -> (index, index)
// CHECK-SAME: %[[ARG0:[a-zA-Z0-9_]+]]: tensor<2x3x5x4x?x7xf32>
// CHECK-DAG: %[[C4:.+]] = constant 4 : index
// CHECK-DAG: %[[C5:.+]] = constant 5 : index
-// CHECK: %[[D0:.+]] = dim %[[ARG0]], %[[C4]]
+// CHECK: %[[D0:.+]] = memref.dim %[[ARG0]], %[[C4]]
// CHECK: %[[D1:.+]] = affine.apply #[[MAP]]()[%[[D0]]]
// CHECK: return %[[C5]], %[[D1]]
@@ -778,8 +779,8 @@ func @propogate_casts(%arg0 : tensor<?x?xf32>, %arg1 : f32, %arg2 : index,
%c42 = constant 42 : index
%0 = linalg.init_tensor [%c21, %c42] : tensor<?x?xf32>
%1 = linalg.fill(%0, %arg1) : tensor<?x?xf32>, f32 -> tensor<?x?xf32>
- %2 = dim %arg0, %c0 : tensor<?x?xf32>
- %3 = dim %arg0, %c1 : tensor<?x?xf32>
+ %2 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %3 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%4 = subtensor_insert %arg0 into %1[%arg2, %arg3] [%2, %3] [1, 1] : tensor<?x?xf32> into tensor<?x?xf32>
return %4 : tensor<?x?xf32>
}
diff --git a/mlir/test/Dialect/Linalg/convert-elementwise-to-linalg.mlir b/mlir/test/Dialect/Linalg/convert-elementwise-to-linalg.mlir
index 55faf663085f..047bd15edfb3 100644
--- a/mlir/test/Dialect/Linalg/convert-elementwise-to-linalg.mlir
+++ b/mlir/test/Dialect/Linalg/convert-elementwise-to-linalg.mlir
@@ -93,11 +93,11 @@ func @cmpf(%arg0: tensor<f32>, %arg1: tensor<f32>) -> tensor<i1> {
// CHECK-SAME: %[[ARG1:[0-9a-zA-Z]*]]: tensor<4x?x?x8x2x?xf32>
func @cmpf(%arg0: tensor<4x?x?x8x2x?xf32>, %arg1: tensor<4x?x?x8x2x?xf32>) -> tensor<4x?x?x8x2x?xi1> {
// CHECK: %[[C1:.*]] = constant 1 : index
- // CHECK: %[[D1:.*]] = dim %[[ARG0]], %[[C1]] : tensor<4x?x?x8x2x?xf32>
+ // CHECK: %[[D1:.*]] = memref.dim %[[ARG0]], %[[C1]] : tensor<4x?x?x8x2x?xf32>
// CHECK: %[[C2:.*]] = constant 2 : index
- // CHECK: %[[D2:.*]] = dim %[[ARG0]], %[[C2]] : tensor<4x?x?x8x2x?xf32>
+ // CHECK: %[[D2:.*]] = memref.dim %[[ARG0]], %[[C2]] : tensor<4x?x?x8x2x?xf32>
// CHECK: %[[C5:.*]] = constant 5 : index
- // CHECK: %[[D5:.*]] = dim %[[ARG0]], %[[C5]] : tensor<4x?x?x8x2x?xf32>
+ // CHECK: %[[D5:.*]] = memref.dim %[[ARG0]], %[[C5]] : tensor<4x?x?x8x2x?xf32>
// CHECK: %[[INIT:.*]] = linalg.init_tensor [4, %[[D1]], %[[D2]], 8, 2, %[[D5]]] : tensor<4x?x?x8x2x?xi1>
// CHECK: linalg.generic
// CHECK-SAME: ins(%[[ARG0]], %[[ARG1]]
diff --git a/mlir/test/Dialect/Linalg/fold-affine-min-scf.mlir b/mlir/test/Dialect/Linalg/fold-affine-min-scf.mlir
index 72d76b3d1869..200ed8475950 100644
--- a/mlir/test/Dialect/Linalg/fold-affine-min-scf.mlir
+++ b/mlir/test/Dialect/Linalg/fold-affine-min-scf.mlir
@@ -13,21 +13,21 @@ func @scf_for(%A : memref<i64>, %step : index) {
// CHECK: scf.for
// CHECK-NEXT: %[[C2:.*]] = constant 2 : index
// CHECK-NEXT: %[[C2I64:.*]] = index_cast %[[C2:.*]]
- // CHECK-NEXT: store %[[C2I64]], %{{.*}}[] : memref<i64>
+ // CHECK-NEXT: memref.store %[[C2I64]], %{{.*}}[] : memref<i64>
scf.for %i = %c0 to %c4 step %c2 {
%1 = affine.min affine_map<(d0, d1)[] -> (2, d1 - d0)> (%i, %c4)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// CHECK: scf.for
// CHECK-NEXT: %[[C2:.*]] = constant 2 : index
// CHECK-NEXT: %[[C2I64:.*]] = index_cast %[[C2:.*]]
- // CHECK-NEXT: store %[[C2I64]], %{{.*}}[] : memref<i64>
+ // CHECK-NEXT: memref.store %[[C2I64]], %{{.*}}[] : memref<i64>
scf.for %i = %c1 to %c7 step %c2 {
%1 = affine.min affine_map<(d0)[s0] -> (s0 - d0, 2)> (%i)[%c7]
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This should not canonicalize because: 4 - %i may take the value 1 < 2.
@@ -37,7 +37,7 @@ func @scf_for(%A : memref<i64>, %step : index) {
scf.for %i = %c1 to %c4 step %c2 {
%1 = affine.min affine_map<(d0)[s0] -> (2, s0 - d0)> (%i)[%c4]
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This should not canonicalize because: 16 - %i may take the value 15 < 1024.
@@ -47,7 +47,7 @@ func @scf_for(%A : memref<i64>, %step : index) {
scf.for %i = %c1 to %c16 step %c1024 {
%1 = affine.min affine_map<(d0) -> (1024, 16 - d0)> (%i)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This example should simplify but affine_map is currently missing
@@ -62,7 +62,7 @@ func @scf_for(%A : memref<i64>, %step : index) {
scf.for %i = %c0 to %ub step %step {
%1 = affine.min affine_map<(d0, d1, d2) -> (d0, d1 - d2)> (%step, %ub, %i)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This example should simplify but affine_map is currently missing
@@ -79,7 +79,7 @@ func @scf_for(%A : memref<i64>, %step : index) {
scf.for %i = %c0 to %ub2 step %step {
%1 = affine.min affine_map<(d0, d1, d2) -> (d0, d2 - d1)> (%step, %i, %ub2)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
return
@@ -96,21 +96,21 @@ func @scf_parallel(%A : memref<i64>, %step : index) {
// CHECK: scf.parallel
// CHECK-NEXT: %[[C2:.*]] = constant 2 : index
// CHECK-NEXT: %[[C2I64:.*]] = index_cast %[[C2:.*]]
- // CHECK-NEXT: store %[[C2I64]], %{{.*}}[] : memref<i64>
+ // CHECK-NEXT: memref.store %[[C2I64]], %{{.*}}[] : memref<i64>
scf.parallel (%i) = (%c0) to (%c4) step (%c2) {
%1 = affine.min affine_map<(d0, d1)[] -> (2, d1 - d0)> (%i, %c4)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// CHECK: scf.parallel
// CHECK-NEXT: %[[C2:.*]] = constant 2 : index
// CHECK-NEXT: %[[C2I64:.*]] = index_cast %[[C2:.*]]
- // CHECK-NEXT: store %[[C2I64]], %{{.*}}[] : memref<i64>
+ // CHECK-NEXT: memref.store %[[C2I64]], %{{.*}}[] : memref<i64>
scf.parallel (%i) = (%c1) to (%c7) step (%c2) {
%1 = affine.min affine_map<(d0)[s0] -> (2, s0 - d0)> (%i)[%c7]
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This example should simplify but affine_map is currently missing
@@ -125,7 +125,7 @@ func @scf_parallel(%A : memref<i64>, %step : index) {
scf.parallel (%i) = (%c0) to (%ub) step (%step) {
%1 = affine.min affine_map<(d0, d1, d2) -> (d0, d2 - d1)> (%step, %i, %ub)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
// This example should simplify but affine_map is currently missing
@@ -140,7 +140,7 @@ func @scf_parallel(%A : memref<i64>, %step : index) {
scf.parallel (%i) = (%c0) to (%ub2) step (%step) {
%1 = affine.min affine_map<(d0, d1, d2) -> (d0, d2 - d1)> (%step, %i, %ub2)
%2 = index_cast %1: index to i64
- store %2, %A[]: memref<i64>
+ memref.store %2, %A[]: memref<i64>
}
return
diff --git a/mlir/test/Dialect/Linalg/forward-vector-transfers.mlir b/mlir/test/Dialect/Linalg/forward-vector-transfers.mlir
index 35e9b7f13a7e..788b1566628f 100644
--- a/mlir/test/Dialect/Linalg/forward-vector-transfers.mlir
+++ b/mlir/test/Dialect/Linalg/forward-vector-transfers.mlir
@@ -4,17 +4,17 @@
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.fill
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_read %[[ARG0]]
// CHECK-NOT: masked
func @testAllocRead(%in: memref<? x f32>) -> vector<32 x f32> {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<32 x f32>
- %subview = subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<32 x f32>
+ %subview = memref.subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
linalg.copy(%in, %subview): memref<? x f32>, memref<16 x f32>
%0 = vector.transfer_read %alloc[%c0], %f0 {masked = [false]} : memref<32 x f32>, vector<32 x f32>
- dealloc %alloc : memref<32 x f32>
+ memref.dealloc %alloc : memref<32 x f32>
return %0: vector<32 x f32>
}
@@ -22,18 +22,18 @@ func @testAllocRead(%in: memref<? x f32>) -> vector<32 x f32> {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.fill
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_read %[[ARG0]]
// CHECK-NOT: masked
func @testAllocFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<32 x f32>
+ %alloc = memref.alloc() : memref<32 x f32>
linalg.fill(%alloc, %f0): memref<32 x f32>, f32
- %subview = subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %subview = memref.subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
linalg.copy(%in, %subview): memref<? x f32>, memref<16 x f32>
%0 = vector.transfer_read %alloc[%c0], %f0 {masked = [false]} : memref<32 x f32>, vector<32 x f32>
- dealloc %alloc : memref<32 x f32>
+ memref.dealloc %alloc : memref<32 x f32>
return %0: vector<32 x f32>
}
@@ -41,18 +41,18 @@ func @testAllocFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.fill
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_read %[[ARG0]]
// CHECK-NOT: masked
func @testViewRead(%in: memref<? x f32>) -> vector<32 x f32> {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<128 x i8>
- %view = view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
- %subview = subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<128 x i8>
+ %view = memref.view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
+ %subview = memref.subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
linalg.copy(%in, %subview): memref<? x f32>, memref<16 x f32>
%0 = vector.transfer_read %view[%c0], %f0 {masked = [false]} : memref<32 x f32>, vector<32 x f32>
- dealloc %alloc : memref<128 x i8>
+ memref.dealloc %alloc : memref<128 x i8>
return %0: vector<32 x f32>
}
@@ -60,19 +60,19 @@ func @testViewRead(%in: memref<? x f32>) -> vector<32 x f32> {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.fill
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_read %[[ARG0]]
// CHECK-NOT: masked
func @testViewFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<128 x i8>
- %view = view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
- %subview = subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<128 x i8>
+ %view = memref.view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
+ %subview = memref.subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
linalg.fill(%view, %f0): memref<32 x f32>, f32
linalg.copy(%in, %subview): memref<? x f32>, memref<16 x f32>
%0 = vector.transfer_read %view[%c0], %f0 {masked = [false]} : memref<32 x f32>, vector<32 x f32>
- dealloc %alloc : memref<128 x i8>
+ memref.dealloc %alloc : memref<128 x i8>
return %0: vector<32 x f32>
}
@@ -80,17 +80,17 @@ func @testViewFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: vector
// CHECK-SAME: %[[ARG1:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_write %[[ARG0]], %[[ARG1]]
// CHECK-NOT: masked
func @testAllocWrite(%vec: vector<32 x f32>, %out: memref<? x f32>) {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<32 x f32>
- %subview = subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<32 x f32>
+ %subview = memref.subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
vector.transfer_write %vec, %alloc[%c0] {masked = [false]} : vector<32 x f32>, memref<32 x f32>
linalg.copy(%subview, %out): memref<16 x f32>, memref<? x f32>
- dealloc %alloc : memref<32 x f32>
+ memref.dealloc %alloc : memref<32 x f32>
return
}
@@ -98,18 +98,18 @@ func @testAllocWrite(%vec: vector<32 x f32>, %out: memref<? x f32>) {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: vector
// CHECK-SAME: %[[ARG1:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: linalg.copy
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_write %[[ARG0]], %[[ARG1]]
// CHECK-NOT: masked
func @testViewWrite(%vec: vector<32 x f32>, %out: memref<? x f32>) {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<128 x i8>
- %view = view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
- %subview = subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<128 x i8>
+ %view = memref.view %alloc[%c0][] : memref<128 x i8> to memref<32 x f32>
+ %subview = memref.subview %view[0][16][1] : memref<32 x f32> to memref<16 x f32>
vector.transfer_write %vec, %view[%c0] {masked = [false]} : vector<32 x f32>, memref<32 x f32>
linalg.copy(%subview, %out): memref<16 x f32>, memref<? x f32>
- dealloc %alloc : memref<128 x i8>
+ memref.dealloc %alloc : memref<128 x i8>
return
}
@@ -121,20 +121,20 @@ func @testViewWrite(%vec: vector<32 x f32>, %out: memref<? x f32>) {
// CHECK-LABEL: failAllocFillRead
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: vector.transfer_read %[[ARG0]]
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: linalg.copy
// CHECK: vector.transfer_read %[[ALLOC]]
func @failAllocFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
%c0 = constant 0: index
%f0 = constant 0.0: f32
%f1 = constant 1.0: f32
- %alloc = alloc() : memref<32 x f32>
+ %alloc = memref.alloc() : memref<32 x f32>
linalg.fill(%alloc, %f0): memref<32 x f32>, f32
- %subview = subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %subview = memref.subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
linalg.copy(%in, %subview): memref<? x f32>, memref<16 x f32>
"some_interleaved_use"(%subview) : (memref<16 x f32>) -> ()
%0 = vector.transfer_read %alloc[%c0], %f1: memref<32 x f32>, vector<32 x f32>
- dealloc %alloc : memref<32 x f32>
+ memref.dealloc %alloc : memref<32 x f32>
return %0: vector<32 x f32>
}
@@ -143,17 +143,17 @@ func @failAllocFillRead(%in: memref<? x f32>) -> vector<32 x f32> {
// CHECK-SAME: %[[ARG0:[0-9a-zA-Z]*]]: vector
// CHECK-SAME: %[[ARG1:[0-9a-zA-Z]*]]: memref
// CHECK-NOT: vector.transfer_write %[[ARG0]], %[[ARG1]]
-// CHECK: %[[ALLOC:.*]] = alloc
+// CHECK: %[[ALLOC:.*]] = memref.alloc
// CHECK: vector.transfer_write %[[ARG0]], %[[ALLOC]]
// CHECK: linalg.copy
func @failAllocWrite(%vec: vector<32 x f32>, %out: memref<? x f32>) {
%c0 = constant 0: index
%f0 = constant 0.0: f32
- %alloc = alloc() : memref<32 x f32>
- %subview = subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
+ %alloc = memref.alloc() : memref<32 x f32>
+ %subview = memref.subview %alloc[0][16][1] : memref<32 x f32> to memref<16 x f32>
vector.transfer_write %vec, %alloc[%c0] : vector<32 x f32>, memref<32 x f32>
"some_interleaved_use"(%subview) : (memref<16 x f32>) -> ()
linalg.copy(%subview, %out): memref<16 x f32>, memref<? x f32>
- dealloc %alloc : memref<32 x f32>
+ memref.dealloc %alloc : memref<32 x f32>
return
}
diff --git a/mlir/test/Dialect/Linalg/fusion-2-level.mlir b/mlir/test/Dialect/Linalg/fusion-2-level.mlir
index 27154a5e277c..3d0a09a22646 100644
--- a/mlir/test/Dialect/Linalg/fusion-2-level.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-2-level.mlir
@@ -9,26 +9,26 @@ func @f1(%A: memref<?x?xf32, offset: ?, strides: [?, 1]>, %B: memref<?x?xf32, of
%c40 = constant 40 : index
%c30 = constant 30 : index
%c20 = constant 20 : index
- %0 = dim %C, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %1 = dim %C, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %2 = dim %D, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %0 = memref.dim %C, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %1 = memref.dim %C, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
linalg.matmul ins(%A, %B: memref<?x?xf32, offset: ?, strides: [?, 1]>, memref<?x?xf32, offset: ?, strides: [?, 1]>)
outs(%C: memref<?x?xf32, offset: ?, strides: [?, 1]>)
scf.for %arg5 = %c0 to %0 step %c20 {
scf.for %arg6 = %c0 to %2 step %c30 {
scf.for %arg7 = %c0 to %1 step %c40 {
- %5 = std.subview %C[%arg5, %arg7][%c20, %c40][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %D[%arg7, %arg6][%c40, %c30][%c1, %c1]: memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c20, %c40][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %9 = dim %5, %c0 : memref<?x?xf32, offset: ?, strides: [?, ?]>
- %10 = dim %5, %c1 : memref<?x?xf32, offset: ?, strides: [?, ?]>
- %11 = dim %7, %c1 : memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %5 = memref.subview %C[%arg5, %arg7][%c20, %c40][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %7 = memref.subview %D[%arg7, %arg6][%c40, %c30][%c1, %c1]: memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %8 = memref.subview %E[%arg5, %arg6][%c20, %c40][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %9 = memref.dim %5, %c0 : memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %10 = memref.dim %5, %c1 : memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %11 = memref.dim %7, %c1 : memref<?x?xf32, offset: ?, strides: [?, ?]>
scf.for %arg8 = %c0 to %9 step %c2 {
scf.for %arg9 = %c0 to %11 step %c3 {
scf.for %arg10 = %c0 to %10 step %c4 {
- %14 = std.subview %5[%arg8, %arg10][%c2, %c4][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %16 = std.subview %7[%arg10, %arg9][%c4, %c3][%c1, %c1]: memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %17 = std.subview %8[%arg8, %arg9][%c2, %c4][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %14 = memref.subview %5[%arg8, %arg10][%c2, %c4][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %16 = memref.subview %7[%arg10, %arg9][%c4, %c3][%c1, %c1]: memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
+ %17 = memref.subview %8[%arg8, %arg9][%c2, %c4][%c1, %c1] : memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%14, %16: memref<?x?xf32, offset: ?, strides: [?, ?]>, memref<?x?xf32, offset: ?, strides: [?, ?]>)
outs(%17: memref<?x?xf32, offset: ?, strides: [?, ?]>)
}
diff --git a/mlir/test/Dialect/Linalg/fusion-indexed-generic.mlir b/mlir/test/Dialect/Linalg/fusion-indexed-generic.mlir
index 3b4948a3b4f1..76e9148cdff0 100644
--- a/mlir/test/Dialect/Linalg/fusion-indexed-generic.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-indexed-generic.mlir
@@ -21,15 +21,15 @@ func @fuse_indexed_generic_consumer(%A: memref<?x?xf32>,
%c0 = constant 0 : index
%c25 = constant 25 : index
%c10 = constant 10 : index
- %0 = dim %C, %c0 : memref<?x?xf32>
- %1 = dim %C, %c1 : memref<?x?xf32>
- %2 = dim %D, %c0 : memref<?x?xf32>
- %3 = dim %D, %c1 : memref<?x?xf32>
+ %0 = memref.dim %C, %c0 : memref<?x?xf32>
+ %1 = memref.dim %C, %c1 : memref<?x?xf32>
+ %2 = memref.dim %D, %c0 : memref<?x?xf32>
+ %3 = memref.dim %D, %c1 : memref<?x?xf32>
scf.for %arg2 = %c0 to %0 step %c10 {
scf.for %arg3 = %c0 to %1 step %c25 {
- %4 = std.subview %C[%arg2, %arg3][%c10, %c25][%c1, %c1] :
+ %4 = memref.subview %C[%arg2, %arg3][%c10, %c25][%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
- %5 = std.subview %D[%arg2, %arg3][%c10, %c25][%c1, %c1] :
+ %5 = memref.subview %D[%arg2, %arg3][%c10, %c25][%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
linalg.indexed_generic {
indexing_maps = [#id_2d, #id_2d],
@@ -89,14 +89,14 @@ func @fuse_indexed_generic_producer(%A: memref<?x?xf32>,
%out = addf %tmp, %j_float : f32
linalg.yield %out : f32
}
- %C_X = dim %C, %c0 : memref<?x?xf32>
- %C_Y = dim %C, %c1 : memref<?x?xf32>
- %D_X = dim %D, %c0 : memref<?x?xf32>
- %D_Y = dim %D, %c1 : memref<?x?xf32>
+ %C_X = memref.dim %C, %c0 : memref<?x?xf32>
+ %C_Y = memref.dim %C, %c1 : memref<?x?xf32>
+ %D_X = memref.dim %D, %c0 : memref<?x?xf32>
+ %D_Y = memref.dim %D, %c1 : memref<?x?xf32>
scf.parallel (%arg2, %arg3) = (%c0, %c0) to (%C_X, %C_Y) step (%c10, %c25) {
- %C_view = std.subview %C[%arg2, %arg3][%c10, %c25][%c1, %c1] :
+ %C_view = memref.subview %C[%arg2, %arg3][%c10, %c25][%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
- %D_view = std.subview %D[%arg2, %arg3][%c10, %c25][%c1, %c1] :
+ %D_view = memref.subview %D[%arg2, %arg3][%c10, %c25][%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
linalg.generic {
indexing_maps = [#id_2d, #id_2d],
@@ -150,18 +150,18 @@ func @fuse_indexed_generic_producer_tile_second_dim_only(%A: memref<?x?xf32>,
%out = addf %tmp, %j_float : f32
linalg.yield %out : f32
}
- %C_X = dim %C, %c0 : memref<?x?xf32>
- %C_Y = dim %C, %c1 : memref<?x?xf32>
- %D_X = dim %D, %c0 : memref<?x?xf32>
- %D_Y = dim %D, %c1 : memref<?x?xf32>
+ %C_X = memref.dim %C, %c0 : memref<?x?xf32>
+ %C_Y = memref.dim %C, %c1 : memref<?x?xf32>
+ %D_X = memref.dim %D, %c0 : memref<?x?xf32>
+ %D_Y = memref.dim %D, %c1 : memref<?x?xf32>
%3 = linalg.range %c0 : %C_Y : %c3 : !linalg.range
scf.parallel (%j) = (%c0) to (%C_Y) step (%c3) {
%0 = affine.min affine_map<(d0, d1, d2) -> (d0, d1 - d2)>(%c3, %C_Y, %j)
- %C_view = subview %C[%c0, %j] [%C_X, %0] [%c1, %c1] :
+ %C_view = memref.subview %C[%c0, %j] [%C_X, %0] [%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
%1 = affine.min affine_map<(d0, d1, d2) -> (d0, d1 - d2)>(%c3, %D_Y, %j)
- %D_view = subview %D[%c0, %j] [%D_X, %1] [%c1, %c1] :
+ %D_view = memref.subview %D[%c0, %j] [%D_X, %1] [%c1, %c1] :
memref<?x?xf32> to memref<?x?xf32, #map>
linalg.generic {
diff --git a/mlir/test/Dialect/Linalg/fusion-pattern.mlir b/mlir/test/Dialect/Linalg/fusion-pattern.mlir
index ca30a32e75b1..aefeeb5e3ada 100644
--- a/mlir/test/Dialect/Linalg/fusion-pattern.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-pattern.mlir
@@ -28,35 +28,35 @@ module {
// CHECK-DAG: %[[CST:.+]] = constant 0.0{{.*}} : f32
// CHECK-DAG: linalg.fill(%[[ARG2]], %[[CST]])
// CHECK-SAME: __internal_linalg_transform__ = "after_basic_fusion_original"
-// CHECK-DAG: %[[M:.+]] = dim %[[ARG0]], %[[C0]]
-// CHECK-DAG: %[[N:.+]] = dim %[[ARG1]], %[[C1]]
+// CHECK-DAG: %[[M:.+]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[N:.+]] = memref.dim %[[ARG1]], %[[C1]]
// CHECK: scf.parallel (%[[IV0:.+]], %[[IV1:.+]]) =
// CHECK-SAME: to (%[[M]], %[[N]])
// CHECK-SAME: step (%[[C32]], %[[C64]]) {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M]]]
-// CHECK: %[[K:.+]] = dim %[[ARG0]], %[[C1]]
-// CHECK: %[[SV1:.+]] = subview %[[ARG0]][%[[IV0]], 0]
+// CHECK: %[[K:.+]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK: %[[SV1:.+]] = memref.subview %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K]]]
-// CHECK: %[[K_2:.+]] = dim %[[ARG1]], %[[C0]]
+// CHECK: %[[K_2:.+]] = memref.dim %[[ARG1]], %[[C0]]
// CHECK: %[[TILE_N:.+]] = affine.min #[[MAP2]](%[[IV1]])[%[[N]]]
-// CHECK: %[[SV2:.+]] = subview %[[ARG1]][0, %[[IV1]]]
+// CHECK: %[[SV2:.+]] = memref.subview %[[ARG1]][0, %[[IV1]]]
// CHECK-SAME: %[[K_2]], %[[TILE_N]]
-// CHECK: %[[M_2:.+]] = dim %[[ARG2]], %[[C0]]
+// CHECK: %[[M_2:.+]] = memref.dim %[[ARG2]], %[[C0]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M_2]]]
-// CHECK: %[[N_2:.+]] = dim %[[ARG2]], %[[C1]]
+// CHECK: %[[N_2:.+]] = memref.dim %[[ARG2]], %[[C1]]
// CHECK: %[[TILE_N_2:.+]] = affine.min #[[MAP2]](%[[IV1]])[%[[N_2]]]
-// CHECK: %[[SV3:.+]] = subview %[[ARG2]][%[[IV0]], %[[IV1]]]
+// CHECK: %[[SV3:.+]] = memref.subview %[[ARG2]][%[[IV0]], %[[IV1]]]
// CHECK-SAME: [%[[TILE_M_2]], %[[TILE_N_2]]]
-// CHECK: %[[SV3_2:.+]] = subview %[[ARG2]][%[[IV0]], %[[IV1]]]
+// CHECK: %[[SV3_2:.+]] = memref.subview %[[ARG2]][%[[IV0]], %[[IV1]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_N]]]
// CHECK: linalg.fill(%[[SV3_2]], %[[CST]])
// CHECK-SAME: __internal_linalg_transform__ = "after_basic_fusion_producer"
// CHECK: scf.for %[[IV2:.+]] = %[[C0]] to %[[K]] step %[[C16]] {
// CHECK: %[[TILE_K:.+]] = affine.min #[[MAP3]](%[[IV2]])[%[[K]]]
-// CHECK: %[[SV4:.+]] = subview %[[SV1]][0, %[[IV2]]]
+// CHECK: %[[SV4:.+]] = memref.subview %[[SV1]][0, %[[IV2]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_K]]]
// CHECK: %[[TILE_K_2:.+]] = affine.min #[[MAP3]](%[[IV2]])[%[[K_2]]]
-// CHECK: %[[SV5:.+]] = subview %[[SV2]][%[[IV2]], 0]
+// CHECK: %[[SV5:.+]] = memref.subview %[[SV2]][%[[IV2]], 0]
// CHECK-SAME: [%[[TILE_K_2]], %[[TILE_N]]]
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_basic_fusion"
@@ -99,40 +99,40 @@ module {
// CHECK-DAG: %[[CST:.+]] = constant 0.0{{.*}} : f32
// CHECK-DAG: linalg.copy(%[[ARG1]], %[[ARG2]])
// CHECK-SAME: __internal_linalg_transform__ = "after_rhs_fusion_original"
-// CHECK-DAG: %[[N:.+]] = dim %[[ARG2]], %[[C1]]
+// CHECK-DAG: %[[N:.+]] = memref.dim %[[ARG2]], %[[C1]]
// CHECK: scf.parallel (%[[IV0:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[N]]) step (%[[C64]]) {
-// CHECK: %[[K:.+]] = dim %[[ARG2]], %[[C0]]
+// CHECK: %[[K:.+]] = memref.dim %[[ARG2]], %[[C0]]
// CHECK: %[[TILE_N:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[N]]]
-// CHECK: %[[SV1:.+]] = subview %[[ARG2]][0, %[[IV0]]]
+// CHECK: %[[SV1:.+]] = memref.subview %[[ARG2]][0, %[[IV0]]]
// CHECK-SAME: [%[[K]], %[[TILE_N]]]
-// CHECK: %[[M:.+]] = dim %[[ARG3]], %[[C0]]
-// CHECK: %[[N_2:.+]] = dim %[[ARG3]], %[[C1]]
+// CHECK: %[[M:.+]] = memref.dim %[[ARG3]], %[[C0]]
+// CHECK: %[[N_2:.+]] = memref.dim %[[ARG3]], %[[C1]]
// CHECK: %[[TILE_N_2:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[N_2]]]
-// CHECK: %[[SV2:.+]] = subview %[[ARG3]][0, %[[IV0]]]
+// CHECK: %[[SV2:.+]] = memref.subview %[[ARG3]][0, %[[IV0]]]
// CHECK-SAME: [%[[M]], %[[TILE_N_2]]]
-// CHECK: %[[K_2:.+]] = dim %[[ARG1]], %[[C0]]
-// CHECK: %[[SV3:.+]] = subview %[[ARG1]][0, %[[IV0]]]
+// CHECK: %[[K_2:.+]] = memref.dim %[[ARG1]], %[[C0]]
+// CHECK: %[[SV3:.+]] = memref.subview %[[ARG1]][0, %[[IV0]]]
// CHECK-SAME: [%[[K_2]], %[[TILE_N]]]
-// CHECK: %[[SV3_2:.+]] = subview %[[ARG2]][0, %[[IV0]]]
+// CHECK: %[[SV3_2:.+]] = memref.subview %[[ARG2]][0, %[[IV0]]]
// CHECK-SAME: [%[[K_2]], %[[TILE_N]]]
// CHECK: linalg.copy(%[[SV3]], %[[SV3_2]])
// CHECK-SAME: __internal_linalg_transform__ = "after_rhs_fusion_producer"
// CHECK-NOT: linalg.fill
-// CHECK-DAG: %[[M_2:.+]] = dim %[[ARG0]], %[[C0]]
-// CHECK-DAG: %[[K_2:.+]] = dim %[[ARG0]], %[[C1]]
+// CHECK-DAG: %[[M_2:.+]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[K_2:.+]] = memref.dim %[[ARG0]], %[[C1]]
// CHECK: scf.parallel (%[[IV1:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[M_2]]) step (%[[C32]]) {
// CHECK-NEXT: scf.for %[[IV2:.+]] = %[[C0]] to %[[K_2]] step %[[C16]] {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP2]](%[[IV1]])[%[[M_2]]]
// CHECK: %[[TILE_K:.+]] = affine.min #[[MAP3]](%[[IV2]])[%[[K_2]]]
-// CHECK: %[[SV4:.+]] = subview %[[ARG0]][%[[IV1]], %[[IV2]]]
+// CHECK: %[[SV4:.+]] = memref.subview %[[ARG0]][%[[IV1]], %[[IV2]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_K]]]
// CHECK: %[[TILE_K_2:.+]] = affine.min #[[MAP3]](%[[IV2]])[%[[K]]]
-// CHECK: %[[SV5:.+]] = subview %[[SV1]][%[[IV2]], 0]
+// CHECK: %[[SV5:.+]] = memref.subview %[[SV1]][%[[IV2]], 0]
// CHECK-SAME: [%[[TILE_K_2]], %[[TILE_N]]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP2]](%[[IV1]])[%[[M]]]
-// CHECK: %[[SV6:.+]] = subview %[[SV2]][%[[IV1]], 0]
+// CHECK: %[[SV6:.+]] = memref.subview %[[SV2]][%[[IV1]], 0]
// CHECK-SAME: [%[[TILE_M_2]], %[[TILE_N_2]]]
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_rhs_fusion"
@@ -179,43 +179,43 @@ module {
// CHECK-SAME: __internal_linalg_transform__ = "after_two_operand_fusion_original"
// CHECK: linalg.fill(%[[ARG3]], %[[CST]])
// CHECK-SAME: __internal_linalg_transform__ = "after_two_operand_fusion_original"
-// CHECK-DAG: %[[M:.+]] = dim %[[ARG1]], %[[C0]]
+// CHECK-DAG: %[[M:.+]] = memref.dim %[[ARG1]], %[[C0]]
// CHECK: scf.parallel (%[[IV0:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[M]]) step (%[[C32]]) {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M]]]
-// CHECK: %[[K:.+]] = dim %[[ARG1]], %[[C1]]
-// CHECK: %[[SV1:.+]] = subview %[[ARG1]][%[[IV0]], 0]
+// CHECK: %[[K:.+]] = memref.dim %[[ARG1]], %[[C1]]
+// CHECK: %[[SV1:.+]] = memref.subview %[[ARG1]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K]]]
-// CHECK: %[[M_2:.+]] = dim %[[ARG3]], %[[C0]]
+// CHECK: %[[M_2:.+]] = memref.dim %[[ARG3]], %[[C0]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M_2]]]
-// CHECK: %[[N:.+]] = dim %[[ARG3]], %[[C1]]
-// CHECK: %[[SV2:.+]] = subview %[[ARG3]][%[[IV0]], 0]
+// CHECK: %[[N:.+]] = memref.dim %[[ARG3]], %[[C1]]
+// CHECK: %[[SV2:.+]] = memref.subview %[[ARG3]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M_2]], %[[N]]]
-// CHECK: %[[SV2_2:.+]] = subview %[[ARG3]][%[[IV0]], 0]
+// CHECK: %[[SV2_2:.+]] = memref.subview %[[ARG3]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[N]]]
-// CHECK: %[[K_2:.+]] = dim %[[ARG0]], %[[C1]]
-// CHECK: %[[SV3:.+]] = subview %[[ARG0]][%[[IV0]], 0]
+// CHECK: %[[K_2:.+]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK: %[[SV3:.+]] = memref.subview %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K_2]]]
-// CHECK: %[[SV3_2:.+]] = subview %[[ARG1]][%[[IV0]], 0]
+// CHECK: %[[SV3_2:.+]] = memref.subview %[[ARG1]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K_2]]]
// CHECK: linalg.copy(%[[SV3]], %[[SV3_2]])
// CHECK-SAME: __internal_linalg_transform__ = "after_two_operand_fusion_producer"
// CHECK: linalg.fill(%[[SV2_2]], %[[CST]])
// CHECK-SAME: __internal_linalg_transform__ = "after_two_operand_fusion_producer"
-// CHECK-DAG: %[[N_2:.+]] = dim %[[ARG2]], %[[C1]]
+// CHECK-DAG: %[[N_2:.+]] = memref.dim %[[ARG2]], %[[C1]]
// CHECK: scf.parallel (%[[IV1:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[N_2]]) step (%[[C64]]) {
// CHECK-NEXT: scf.for %[[IV2:.+]] = %[[C0]] to %[[K]] step %[[C16]] {
// CHECK: %[[TILE_K:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[K]]]
-// CHECK: %[[SV4:.+]] = subview %[[SV1]][0, %[[IV2]]]
+// CHECK: %[[SV4:.+]] = memref.subview %[[SV1]][0, %[[IV2]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_K]]]
-// CHECK: %[[K_2:.+]] = dim %[[ARG2]], %[[C0]]
+// CHECK: %[[K_2:.+]] = memref.dim %[[ARG2]], %[[C0]]
// CHECK: %[[TILE_K_2:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[K_2]]]
// CHECK: %[[TILE_N:.+]] = affine.min #[[MAP3]](%[[IV1]])[%[[N_2]]]
-// CHECK: %[[SV5:.+]] = subview %[[ARG2]][%[[IV2]], %[[IV1]]]
+// CHECK: %[[SV5:.+]] = memref.subview %[[ARG2]][%[[IV2]], %[[IV1]]]
// CHECK-SAME: [%[[TILE_K_2]], %[[TILE_N]]]
// CHECK: %[[TILE_N_2:.+]] = affine.min #[[MAP3]](%[[IV1]])[%[[N]]]
-// CHECK: %[[SV6:.+]] = subview %[[SV2]][0, %[[IV1]]]
+// CHECK: %[[SV6:.+]] = memref.subview %[[SV2]][0, %[[IV1]]]
// CHECK-SAME: [%[[TILE_M_2]], %[[TILE_N_2]]]
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_two_operand_fusion"
@@ -259,44 +259,44 @@ module {
// CHECK-DAG: %[[C16:.+]] = constant 16 : index
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_lhs_fusion_original"
-// CHECK-DAG: %[[M:.+]] = dim %[[ARG2]], %[[C0]]
+// CHECK-DAG: %[[M:.+]] = memref.dim %[[ARG2]], %[[C0]]
// CHECK: scf.parallel (%[[IV0:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[M]]) step (%[[C32]]) {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M]]]
-// CHECK: %[[K2:.+]] = dim %[[ARG2]], %[[C1]]
-// CHECK: %[[SV1:.+]] = subview %[[ARG2]][%[[IV0]], 0]
+// CHECK: %[[K2:.+]] = memref.dim %[[ARG2]], %[[C1]]
+// CHECK: %[[SV1:.+]] = memref.subview %[[ARG2]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K2]]]
-// CHECK: %[[M_2:.+]] = dim %[[ARG4]], %[[C0]]
+// CHECK: %[[M_2:.+]] = memref.dim %[[ARG4]], %[[C0]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M_2]]]
-// CHECK: %[[N:.+]] = dim %[[ARG4]], %[[C1]]
-// CHECK: %[[SV2:.+]] = subview %[[ARG4]][%[[IV0]], 0]
+// CHECK: %[[N:.+]] = memref.dim %[[ARG4]], %[[C1]]
+// CHECK: %[[SV2:.+]] = memref.subview %[[ARG4]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M_2]], %[[N]]]
-// CHECK: %[[K2_2:.+]] = dim %[[ARG1]], %[[C1]]
-// CHECK: %[[K1:.+]] = dim %[[ARG0]], %[[C1]]
-// CHECK: %[[SV3:.+]] = subview %[[ARG0]][%[[IV0]], 0]
+// CHECK: %[[K2_2:.+]] = memref.dim %[[ARG1]], %[[C1]]
+// CHECK: %[[K1:.+]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK: %[[SV3:.+]] = memref.subview %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K1]]]
-// CHECK: %[[SV4:.+]] = subview %[[ARG1]][0, 0] [%[[K1]], %[[K2_2]]]
-// CHECK: %[[SV1_2:.+]] = subview %[[ARG2]][%[[IV0]], 0]
+// CHECK: %[[SV4:.+]] = memref.subview %[[ARG1]][0, 0] [%[[K1]], %[[K2_2]]]
+// CHECK: %[[SV1_2:.+]] = memref.subview %[[ARG2]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[K2_2]]]
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_lhs_fusion_producer"
// CHECK-SAME: ins(%[[SV3]], %[[SV4]]
// CHECK-SAME: : memref<?x?xf32, #[[MAP1]]>, memref<?x?xf32, #[[MAP1]]>)
// CHECK-SAME: outs(%[[SV1_2]] : memref<?x?xf32, #[[MAP1]]>)
-// CHECK-DAG: %[[N_2:.+]] = dim %[[ARG3]], %[[C1]]
+// CHECK-DAG: %[[N_2:.+]] = memref.dim %[[ARG3]], %[[C1]]
// CHECK: scf.parallel (%[[IV1:.+]]) =
// CHECK-SAME: (%[[C0]]) to (%[[N_2]]) step (%[[C64]]) {
// CHECK-NEXT: scf.for %[[IV2:.+]] = %[[C0]] to %[[K]] step %[[C16]] {
// CHECK: %[[TILE_K:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[K]]]
-// CHECK: %[[SV6:.+]] = subview %[[SV1]][0, %[[IV2]]]
+// CHECK: %[[SV6:.+]] = memref.subview %[[SV1]][0, %[[IV2]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_K]]]
-// CHECK: %[[K_2:.+]] = dim %[[ARG3]], %[[C0]]
+// CHECK: %[[K_2:.+]] = memref.dim %[[ARG3]], %[[C0]]
// CHECK: %[[TILE_K_2:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[K_2]]]
// CHECK: %[[TILE_N:.+]] = affine.min #[[MAP3]](%[[IV1]])[%[[N_2]]]
-// CHECK: %[[SV7:.+]] = subview %[[ARG3]][%[[IV2]], %[[IV1]]]
+// CHECK: %[[SV7:.+]] = memref.subview %[[ARG3]][%[[IV2]], %[[IV1]]]
// CHECK-SAME: [%[[TILE_K_2]], %[[TILE_N]]]
// CHECK: %[[TILE_N_2:.+]] = affine.min #[[MAP3]](%[[IV1]])[%[[N]]]
-// CHECK: %[[SV8:.+]] = subview %[[SV2]][0, %[[IV1]]]
+// CHECK: %[[SV8:.+]] = memref.subview %[[SV2]][0, %[[IV1]]]
// CHECK-SAME: [%[[TILE_M_2]], %[[TILE_N_2]]]
// CHECK: linalg.matmul
// CHECK-SAME: __internal_linalg_transform__ = "after_lhs_fusion"
@@ -316,9 +316,9 @@ module {
%arg2: memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg2, %c0 : memref<?x?xf32>
- %1 = dim %arg2, %c1 : memref<?x?xf32>
- %2 = alloc(%0, %1) : memref<?x?xf32>
+ %0 = memref.dim %arg2, %c0 : memref<?x?xf32>
+ %1 = memref.dim %arg2, %c1 : memref<?x?xf32>
+ %2 = memref.alloc(%0, %1) : memref<?x?xf32>
linalg.matmul ins(%arg0, %arg1 : memref<?x?xf32>, memref<?x?xf32>)
outs(%2 : memref<?x?xf32>)
linalg.generic
@@ -340,14 +340,14 @@ module {
// CHECK-SAME: %[[ARG0:[a-zA-Z0-9_]+]]: memref<?x?xf32>
// CHECK-SAME: %[[ARG1:[a-zA-Z0-9_]+]]: memref<?x?xf32>
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: memref<?x?xf32>
-// CHECK: %[[T2:.+]] = alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32>
+// CHECK: %[[T2:.+]] = memref.alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32>
// CHECK: linalg.matmul
// CHECK-SAME: after_transpose_fusion_original
// CHECK: scf.parallel (%[[ARG3:[a-zA-Z0-9_]+]], %[[ARG4:.[a-zA-Z0-9_]+]])
-// CHECK: %[[T5:.+]] = subview %[[T2]][%[[ARG3]], %[[ARG4]]]
-// CHECK: %[[T6:.+]] = subview %[[ARG2]][%[[ARG3]], %[[ARG4]]]
-// CHECK: %[[T8:.+]] = subview %[[ARG0]][%[[ARG3]], 0]
-// CHECK: %[[T9:.+]] = subview %[[ARG1]][0, %[[ARG4]]]
+// CHECK: %[[T5:.+]] = memref.subview %[[T2]][%[[ARG3]], %[[ARG4]]]
+// CHECK: %[[T6:.+]] = memref.subview %[[ARG2]][%[[ARG3]], %[[ARG4]]]
+// CHECK: %[[T8:.+]] = memref.subview %[[ARG0]][%[[ARG3]], 0]
+// CHECK: %[[T9:.+]] = memref.subview %[[ARG1]][0, %[[ARG4]]]
// CHECK: linalg.matmul
// CHECK-SAME: after_transpose_fusion_producer
// CHECK-SAME: ins(%[[T8]], %[[T9]]
@@ -366,9 +366,9 @@ module {
%arg2: memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg2, %c0 : memref<?x?xf32>
- %1 = dim %arg2, %c1 : memref<?x?xf32>
- %2 = alloc(%0, %1) : memref<?x?xf32>
+ %0 = memref.dim %arg2, %c0 : memref<?x?xf32>
+ %1 = memref.dim %arg2, %c1 : memref<?x?xf32>
+ %2 = memref.alloc(%0, %1) : memref<?x?xf32>
linalg.matmul ins(%arg0, %arg1 : memref<?x?xf32>, memref<?x?xf32>)
outs(%2 : memref<?x?xf32>)
linalg.generic
@@ -413,17 +413,17 @@ module {
%c16 = constant 16 : index
%cst = constant 0.000000e+00 : f32
linalg.fill(%arg2, %cst) : memref<?x?xf32>, f32
- %0 = dim %arg0, %c0 : memref<?x?xf32>
- %1 = dim %arg1, %c1 : memref<?x?xf32>
- %2 = dim %arg0, %c1 : memref<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32>
+ %1 = memref.dim %arg1, %c1 : memref<?x?xf32>
+ %2 = memref.dim %arg0, %c1 : memref<?x?xf32>
scf.parallel (%arg3, %arg4) = (%c0, %c0) to (%0, %1) step (%c32, %c64) {
scf.for %arg5 = %c0 to %2 step %c16 {
%3 = affine.min #map0(%arg3)[%0]
%4 = affine.min #map1(%arg4)[%1]
%5 = affine.min #map2(%arg5)[%2]
- %6 = subview %arg0[%arg3, %arg5] [%3, %5] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
- %7 = subview %arg1[%arg5, %arg4] [%5, %4] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
- %8 = subview %arg2[%arg3, %arg4] [%3, %4] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
+ %6 = memref.subview %arg0[%arg3, %arg5] [%3, %5] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
+ %7 = memref.subview %arg1[%arg5, %arg4] [%5, %4] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
+ %8 = memref.subview %arg2[%arg3, %arg4] [%3, %4] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #map3>
linalg.matmul {__internal_linalg_transform__ = "basic_fusion"}
ins(%6, %7 : memref<?x?xf32, #map3>, memref<?x?xf32, #map3>)
outs(%8 : memref<?x?xf32, #map3>)
diff --git a/mlir/test/Dialect/Linalg/fusion-sequence.mlir b/mlir/test/Dialect/Linalg/fusion-sequence.mlir
index 97d81378701e..bec19b325a7b 100644
--- a/mlir/test/Dialect/Linalg/fusion-sequence.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-sequence.mlir
@@ -6,9 +6,9 @@ module {
%cst = constant 0.000000e+00 : f32
%c0 = constant 0 : index
%c1 = constant 1 : index
- %d0 = dim %arg0, %c0 : memref<?x?xf32>
- %d1 = dim %arg1, %c1 : memref<?x?xf32>
- %0 = alloc(%d0, %d1) : memref<?x?xf32>
+ %d0 = memref.dim %arg0, %c0 : memref<?x?xf32>
+ %d1 = memref.dim %arg1, %c1 : memref<?x?xf32>
+ %0 = memref.alloc(%d0, %d1) : memref<?x?xf32>
linalg.fill(%0, %cst) : memref<?x?xf32>, f32
linalg.matmul ins(%arg0, %arg1 : memref<?x?xf32>, memref<?x?xf32>)
outs(%0 : memref<?x?xf32>)
@@ -34,13 +34,13 @@ module {
// CHECK-SAME: %[[ARG1:[a-zA-Z0-9_]+]]: memref<?x?xf32>
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: memref<?xf32>
// CHECK-SAME: %[[ARG3:[a-zA-Z0-9_]+]]: memref<?x?xf32>
-// CHECK: %[[TEMP:.+]] = alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32>
+// CHECK: %[[TEMP:.+]] = memref.alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32>
// CHECK: scf.parallel (%[[IV0:.+]], %[[IV1:.+]]) = {{.*}} {
-// CHECK-DAG: %[[SV_TEMP:.+]] = subview %[[TEMP]][%[[IV0]], %[[IV1]]]
-// CHECK-DAG: %[[SV_ARG2:.+]] = subview %[[ARG2]][%[[IV1]]]
-// CHECK-DAG: %[[SV_ARG3:.+]] = subview %[[ARG3]][%[[IV0]], %[[IV1]]]
-// CHECK-DAG: %[[SV_ARG0:.+]] = subview %[[ARG0]][%[[IV0]], 0]
-// CHECK-DAG: %[[SV_ARG1:.+]] = subview %[[ARG1]][0, %[[IV1]]]
+// CHECK-DAG: %[[SV_TEMP:.+]] = memref.subview %[[TEMP]][%[[IV0]], %[[IV1]]]
+// CHECK-DAG: %[[SV_ARG2:.+]] = memref.subview %[[ARG2]][%[[IV1]]]
+// CHECK-DAG: %[[SV_ARG3:.+]] = memref.subview %[[ARG3]][%[[IV0]], %[[IV1]]]
+// CHECK-DAG: %[[SV_ARG0:.+]] = memref.subview %[[ARG0]][%[[IV0]], 0]
+// CHECK-DAG: %[[SV_ARG1:.+]] = memref.subview %[[ARG1]][0, %[[IV1]]]
// CHECK: linalg.fill(%[[SV_TEMP]], %{{.+}})
// CHECK: linalg.matmul
// CHECK-SAME: ins(%[[SV_ARG0]], %[[SV_ARG1]]
@@ -62,12 +62,12 @@ module {
%cst = constant 0.000000e+00 : f32
%c0 = constant 0 : index
%c1 = constant 1 : index
- %m = dim %arg0, %c0 : memref<?x?xf32>
- %n1 = dim %arg1, %c1 : memref<?x?xf32>
- %n2 = dim %arg2, %c1 : memref<?x?xf32>
- %n3 = dim %arg3, %c1 : memref<?x?xf32>
- %0 = alloc(%m, %n1) : memref<?x?xf32>
- %1 = alloc(%m, %n2) : memref<?x?xf32>
+ %m = memref.dim %arg0, %c0 : memref<?x?xf32>
+ %n1 = memref.dim %arg1, %c1 : memref<?x?xf32>
+ %n2 = memref.dim %arg2, %c1 : memref<?x?xf32>
+ %n3 = memref.dim %arg3, %c1 : memref<?x?xf32>
+ %0 = memref.alloc(%m, %n1) : memref<?x?xf32>
+ %1 = memref.alloc(%m, %n2) : memref<?x?xf32>
linalg.fill(%0, %cst) : memref<?x?xf32>, f32
linalg.matmul ins(%arg0, %arg1 : memref<?x?xf32>, memref<?x?xf32>)
outs(%0 : memref<?x?xf32>)
@@ -92,30 +92,30 @@ module {
// CHECK-DAG: %[[C0:.+]] = constant 0 : index
// CHECK-DAG: %[[C1:.+]] = constant 1 : index
// CHECK-DAG: %[[C16:.+]] = constant 16 : index
-// CHECK-DAG: %[[M:.+]] = dim %[[ARG0]], %[[C0]]
-// CHECK-DAG: %[[N1:.+]] = dim %[[ARG1]], %[[C1]]
-// CHECK-DAG: %[[N2:.+]] = dim %[[ARG2]], %[[C1]]
-// CHECK: %[[ALLOC1:.+]] = alloc(%[[M]], %[[N1]])
-// CHECK: %[[ALLOC2:.+]] = alloc(%[[M]], %[[N2]])
+// CHECK-DAG: %[[M:.+]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[N1:.+]] = memref.dim %[[ARG1]], %[[C1]]
+// CHECK-DAG: %[[N2:.+]] = memref.dim %[[ARG2]], %[[C1]]
+// CHECK: %[[ALLOC1:.+]] = memref.alloc(%[[M]], %[[N1]])
+// CHECK: %[[ALLOC2:.+]] = memref.alloc(%[[M]], %[[N2]])
// CHECK: scf.parallel (%[[IV0:.+]]) = (%[[C0]]) to (%[[M]])
// CHECK-SAME: step (%[[C16]]) {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M]]]
-// CHECK: %[[SV_ALLOC2:.+]] = subview %[[ALLOC2]][%[[IV0]], 0]
+// CHECK: %[[SV_ALLOC2:.+]] = memref.subview %[[ALLOC2]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[N2]]]
-// CHECK: %[[M_2:.+]] = dim %[[ARG4]], %[[C0]]
+// CHECK: %[[M_2:.+]] = memref.dim %[[ARG4]], %[[C0]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M_2]]]
-// CHECK: %[[N3:.+]] = dim %[[ARG4]], %[[C1]]
-// CHECK: %[[SV_ARG4:.+]] = subview %[[ARG4]][%[[IV0]], 0]
+// CHECK: %[[N3:.+]] = memref.dim %[[ARG4]], %[[C1]]
+// CHECK: %[[SV_ARG4:.+]] = memref.subview %[[ARG4]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M_2]], %[[N3]]]
-// CHECK: %[[SV_ARG4_2:.+]] = subview %[[ARG4]][%[[IV0]], 0]
+// CHECK: %[[SV_ARG4_2:.+]] = memref.subview %[[ARG4]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[N3]]]
-// CHECK: %[[SV_ALLOC1:.+]] = subview %[[ALLOC1]][%[[IV0]], 0]
+// CHECK: %[[SV_ALLOC1:.+]] = memref.subview %[[ALLOC1]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[N1]]]
-// CHECK: %[[SV_ARG2:.+]] = subview %[[ARG2]][0, 0] [%[[N1]], %[[N2]]]
-// CHECK: %[[N0:.+]] = dim %[[ARG0]], %[[C1]]
-// CHECK: %[[SV_ARG0:.+]] = subview %[[ARG0]][%[[IV0]], 0]
+// CHECK: %[[SV_ARG2:.+]] = memref.subview %[[ARG2]][0, 0] [%[[N1]], %[[N2]]]
+// CHECK: %[[N0:.+]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK: %[[SV_ARG0:.+]] = memref.subview %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M:.+]], %[[N0]]]
-// CHECK: %[[SV_ARG1:.+]] = subview %[[ARG1]][0, 0] [%[[N0]], %[[N1]]]
+// CHECK: %[[SV_ARG1:.+]] = memref.subview %[[ARG1]][0, 0] [%[[N0]], %[[N1]]]
// CHECK: linalg.fill(%[[SV_ALLOC1]], %{{.+}})
// CHECK: linalg.matmul ins(%[[SV_ARG0]], %[[SV_ARG1]]
// CHECK-SAME: : memref<?x?xf32, #[[MAP1]]>, memref<?x?xf32, #[[MAP1]]>)
@@ -141,8 +141,8 @@ module {
%c1 = constant 1 : index
%0 = linalg.matmul ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
outs(%arg2 : tensor<?x?xf32>) -> tensor<?x?xf32>
- %1 = dim %0, %c0 : tensor<?x?xf32>
- %2 = dim %0, %c1 : tensor<?x?xf32>
+ %1 = memref.dim %0, %c0 : tensor<?x?xf32>
+ %2 = memref.dim %0, %c1 : tensor<?x?xf32>
%3 = linalg.init_tensor [%1, %2] : tensor<?x?xf32>
%4 = linalg.generic
{indexing_maps = [affine_map<(d0, d1) -> (d0, d1)>,
@@ -212,16 +212,16 @@ module {
// CHECK-DAG: %[[C1:.+]] = constant 1 : index
// CHECK: %[[R0:.+]] = scf.for %[[IV0:[a-zA-Z0-9_]+]] =
// CHECK-SAME: iter_args(%[[ARG8:.+]] = %[[ARG6]]) -> (tensor<?x?xf32>) {
-// CHECK: %[[N3:.+]] = dim %[[ARG8]], %[[C1]]
+// CHECK: %[[N3:.+]] = memref.dim %[[ARG8]], %[[C1]]
// CHECK: %[[STARG6:.+]] = subtensor %[[ARG8]][%[[IV0]], 0]
// CHECK-SAME: [%{{[a-zA-Z0-9_]+}}, %[[N3]]]
-// CHECK: %[[N2:.+]] = dim %[[ARG3]], %[[C1]]
-// CHECK: %[[N1:.+]] = dim %[[ARG1]], %[[C1]]
+// CHECK: %[[N2:.+]] = memref.dim %[[ARG3]], %[[C1]]
+// CHECK: %[[N1:.+]] = memref.dim %[[ARG1]], %[[C1]]
// CHECK: %[[STARG3:.+]] = subtensor %[[ARG3]][0, 0]
// CHECK-SAME: [%[[N1]], %[[N2]]]
// CHECK: %[[STARG4:.+]] = subtensor %[[ARG4]][%[[IV0]], 0]
// CHECK-SAME: [%{{[a-zA-Z0-9_]+}}, %[[N2]]]
-// CHECK: %[[N0:.+]] = dim %[[ARG0]], %[[C1]]
+// CHECK: %[[N0:.+]] = memref.dim %[[ARG0]], %[[C1]]
// CHECK: %[[STARG0:.+]] = subtensor %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%{{[a-zA-Z0-9_]+}}, %[[N0]]]
// CHECK: %[[STARG1:.+]] = subtensor %[[ARG1]][0, 0]
diff --git a/mlir/test/Dialect/Linalg/fusion-tensor-pattern.mlir b/mlir/test/Dialect/Linalg/fusion-tensor-pattern.mlir
index 10cd5b454a4a..bd0d61c8580e 100644
--- a/mlir/test/Dialect/Linalg/fusion-tensor-pattern.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-tensor-pattern.mlir
@@ -28,18 +28,18 @@ module {
// CHECK-DAG: %[[C32:.+]] = constant 32 : index
// CHECK-DAG: %[[C64:.+]] = constant 64 : index
// CHECK-DAG: %[[C16:.+]] = constant 16 : index
-// CHECK-DAG: %[[M:.+]] = dim %[[ARG0]], %[[C0]]
+// CHECK-DAG: %[[M:.+]] = memref.dim %[[ARG0]], %[[C0]]
// CHECK: %[[RESULT:.+]] = scf.for %[[IV0:[a-zA-Z0-9]+]] =
// CHECK-SAME: %[[C0]] to %[[M]] step %[[C32]]
// CHECK-SAME: iter_args(%[[ARG6:.+]] = %[[ARG4]]) -> (tensor<?x?xf32>) {
// CHECK: %[[TILE_M:.+]] = affine.min #[[MAP0]](%[[IV0]])[%[[M]]]
-// CHECK: %[[M_2:.+]] = dim %[[ARG6]], %[[C0]]
+// CHECK: %[[M_2:.+]] = memref.dim %[[ARG6]], %[[C0]]
// CHECK: %[[TILE_M_2:.+]] = affine.min #[[MAP1]](%[[M_2]], %[[IV0]])
-// CHECK: %[[N3:.+]] = dim %[[ARG6]], %[[C1]]
+// CHECK: %[[N3:.+]] = memref.dim %[[ARG6]], %[[C1]]
// CHECK: %[[ST_ARG6:.+]] = subtensor %[[ARG6]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M_2]], %[[N3]]]
-// CHECK: %[[N2:.+]] = dim %[[ARG1]], %[[C1]]
-// CHECK: %[[N1:.+]] = dim %[[ARG0]], %[[C1]]
+// CHECK: %[[N2:.+]] = memref.dim %[[ARG1]], %[[C1]]
+// CHECK: %[[N1:.+]] = memref.dim %[[ARG0]], %[[C1]]
// CHECK: %[[ST_ARG0:.+]] = subtensor %[[ARG0]][%[[IV0]], 0]
// CHECK-SAME: [%[[TILE_M]], %[[N1]]]
// CHECK: %[[ST_ARG1:.+]] = subtensor %[[ARG1]][0, 0]
@@ -50,7 +50,7 @@ module {
// CHECK-SAME: __internal_linalg_transform__ = "after_lhs_fusion_producer"
// CHECK-SAME: ins(%[[ST_ARG0]], %[[ST_ARG1]] : tensor<?x?xf32>, tensor<?x?xf32>)
// CHECK-SAME: outs(%[[ST_ARG2]] : tensor<?x?xf32>)
-// CHECK: %[[N3_2:.+]] = dim %[[ARG3]], %[[C1]]
+// CHECK: %[[N3_2:.+]] = memref.dim %[[ARG3]], %[[C1]]
// CHECK: %[[YIELD0:.+]] = scf.for %[[IV1:[a-zA-Z0-9]+]] =
// CHECK-SAME: %[[C0]] to %[[N3_2]] step %[[C64]]
// CHECK-SAME: iter_args(%[[ARG8:.+]] = %[[ST_ARG6]]) -> (tensor<?x?xf32>) {
@@ -60,13 +60,13 @@ module {
// CHECK: %[[TILE_N2:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[N2]]]
// CHECK: %[[ST_LHS:.+]] = subtensor %[[LHS]][0, %[[IV2]]]
// CHECK-SAME: [%[[TILE_M]], %[[TILE_N2]]]
-// CHECK: %[[N2_3:.+]] = dim %[[ARG3]], %[[C0]]
+// CHECK: %[[N2_3:.+]] = memref.dim %[[ARG3]], %[[C0]]
// CHECK: %[[TILE_N2_2:.+]] = affine.min #[[MAP2]](%[[IV2]])[%[[N2_3]]]
// CHECK: %[[TILE_N3:.+]] = affine.min #[[MAP3]](%[[IV1]])[%[[N3_2]]]
// CHECK: %[[ST_ARG3:.+]] = subtensor %[[ARG3]][%[[IV2]], %[[IV1]]]
// CHECK-SAME: [%[[TILE_N2_2]], %[[TILE_N3]]]
-// CHECK: %[[M_4:.+]] = dim %[[ARG10]], %[[C0]]
-// CHECK: %[[N3_3:.+]] = dim %[[ARG10]], %[[C1]]
+// CHECK: %[[M_4:.+]] = memref.dim %[[ARG10]], %[[C0]]
+// CHECK: %[[N3_3:.+]] = memref.dim %[[ARG10]], %[[C1]]
// CHECK: %[[TILE_N3_2:.+]] = affine.min #[[MAP4]](%[[N3_3]], %[[IV1]])
// CHECK: %[[ST_ARG4:.+]] = subtensor %[[ARG10]][0, %[[IV1]]]
// CHECK-SAME: [%[[M_4]], %[[TILE_N3_2]]]
@@ -94,12 +94,12 @@ module {
%arg2: tensor<?x?xf32>) -> tensor<?x?xf32>{
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg2, %c0 : tensor<?x?xf32>
- %1 = dim %arg2, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg2, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg2, %c1 : tensor<?x?xf32>
%2 = linalg.matmul ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
outs(%arg2 : tensor<?x?xf32>) -> tensor<?x?xf32>
- %3 = dim %2, %c0 : tensor<?x?xf32>
- %4 = dim %2, %c1 : tensor<?x?xf32>
+ %3 = memref.dim %2, %c0 : tensor<?x?xf32>
+ %4 = memref.dim %2, %c1 : tensor<?x?xf32>
%5 = linalg.init_tensor [%3, %4] : tensor<?x?xf32>
%6 = linalg.generic
{indexing_maps = [affine_map<(d0, d1) -> (d0, d1)>,
diff --git a/mlir/test/Dialect/Linalg/fusion-tensor.mlir b/mlir/test/Dialect/Linalg/fusion-tensor.mlir
index 6a67b5dff10e..a4071897b4d8 100644
--- a/mlir/test/Dialect/Linalg/fusion-tensor.mlir
+++ b/mlir/test/Dialect/Linalg/fusion-tensor.mlir
@@ -8,8 +8,8 @@ func @add_mul_fusion(%arg0: tensor<?x?xf32>, %arg1 : tensor<?x?xf32>, %arg2 : te
{
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xf32>
%3 = linalg.generic {indexing_maps = [#map0, #map0, #map0], iterator_types = ["parallel", "parallel"]}
ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
@@ -50,8 +50,8 @@ func @transpose_add_mul_fusion(%arg0: tensor<?x?xf32>, %arg1 : tensor<?x?xf32>,
{
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xf32>
%3 = linalg.generic {indexing_maps = [#map0, #map1, #map0], iterator_types = ["parallel", "parallel"]}
ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
@@ -84,8 +84,8 @@ func @add_transpose_mul_fusion(%arg0: tensor<?x?xf32>, %arg1 : tensor<?x?xf32>,
{
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xf32>
%3 = linalg.generic {indexing_maps = [#map0, #map1, #map0], iterator_types = ["parallel", "parallel"]}
ins(%arg0, %arg1 : tensor<?x?xf32>, tensor<?x?xf32>)
@@ -119,7 +119,7 @@ func @add_broadcast_mul_fusion(%arg0: tensor<?xf32>, %arg1 : tensor<?xf32>, %arg
{
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?xf32>
%1 = linalg.init_tensor [%0] : tensor<?xf32>
%2 = linalg.generic {indexing_maps = [#map2, #map2, #map2], iterator_types = ["parallel"]}
ins(%arg0, %arg1 : tensor<?xf32>, tensor<?xf32>)
@@ -130,7 +130,7 @@ func @add_broadcast_mul_fusion(%arg0: tensor<?xf32>, %arg1 : tensor<?xf32>, %arg
} -> tensor<?xf32>
// CHECK: linalg.generic {
// CHECK-SAME: indexing_maps = {{\[}}[[$MAP1]], [[$MAP1]], [[$MAP0]], [[$MAP0]]
- %3 = dim %arg2, %c1 : tensor<?x?xf32>
+ %3 = memref.dim %arg2, %c1 : tensor<?x?xf32>
%4 = linalg.init_tensor [%0, %3] : tensor<?x?xf32>
%5 = linalg.generic {indexing_maps = [#map1, #map0, #map0], iterator_types = ["parallel", "parallel"]}
ins(%2, %arg2 : tensor<?xf32>, tensor<?x?xf32>)
@@ -182,8 +182,8 @@ func @generic_op_constant_fusion(%arg0 : tensor<5x?x?xf32>) -> tensor<5x?x?xf32>
%c1 = constant 1 : index
%c2 = constant 2 : index
%cst = constant dense<42.0> : tensor<5xf32>
- %0 = dim %arg0, %c1 : tensor<5x?x?xf32>
- %1 = dim %arg0, %c2 : tensor<5x?x?xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<5x?x?xf32>
+ %1 = memref.dim %arg0, %c2 : tensor<5x?x?xf32>
%2 = linalg.init_tensor [5, %0, %1] : tensor<5x?x?xf32>
%3 = linalg.generic {
indexing_maps = [#map0, #map1, #map1],
@@ -214,8 +214,8 @@ func @indexed_generic_op_constant_fusion(%arg0 : tensor<5x?x?xf32>)
%c1 = constant 1 : index
%c2 = constant 2 : index
%cst = constant dense<42.0> : tensor<5xf32>
- %0 = dim %arg0, %c1 : tensor<5x?x?xf32>
- %1 = dim %arg0, %c2 : tensor<5x?x?xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<5x?x?xf32>
+ %1 = memref.dim %arg0, %c2 : tensor<5x?x?xf32>
%2 = linalg.init_tensor [5, %0, %1] : tensor<5x?x?xf32>
%3 = linalg.indexed_generic {
indexing_maps = [#map0, #map1, #map1],
@@ -250,8 +250,8 @@ func @generic_op_zero_dim_constant_fusion(%arg0 : tensor<5x?x?xf32>)
%c1 = constant 1 : index
%c2 = constant 2 : index
%cst = constant dense<42.0> : tensor<f32>
- %0 = dim %arg0, %c1 : tensor<5x?x?xf32>
- %1 = dim %arg0, %c2 : tensor<5x?x?xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<5x?x?xf32>
+ %1 = memref.dim %arg0, %c2 : tensor<5x?x?xf32>
%2 = linalg.init_tensor [5, %0, %1] : tensor<5x?x?xf32>
%3 = linalg.generic {
indexing_maps = [#map0, #map1, #map1],
@@ -282,8 +282,8 @@ func @indexed_generic_op_zero_dim_constant_fusion
%c1 = constant 1 : index
%c2 = constant 2 : index
%cst = constant dense<42.0> : tensor<f32>
- %0 = dim %arg0, %c1 : tensor<5x?x?xf32>
- %1 = dim %arg0, %c2 : tensor<5x?x?xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<5x?x?xf32>
+ %1 = memref.dim %arg0, %c2 : tensor<5x?x?xf32>
%2 = linalg.init_tensor [5, %0, %1] : tensor<5x?x?xf32>
%3 = linalg.indexed_generic {
indexing_maps = [#map0, #map1, #map1],
@@ -314,8 +314,8 @@ func @generic_op_indexed_generic_op_fusion(%arg0: tensor<?x?xi32>,
%arg1: tensor<?x?xi32>) -> tensor<?x?xi32> {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xi32>
- %1 = dim %arg0, %c1 : tensor<?x?xi32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xi32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xi32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xi32>
%3 = linalg.generic {
indexing_maps = [#map0, #map0, #map0],
@@ -364,8 +364,8 @@ func @indexed_generic_op_generic_op_fusion(%arg0: tensor<?x?xi32>,
%arg1: tensor<?x?xi32>) -> tensor<?x?xi32> {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xi32>
- %1 = dim %arg0, %c1 : tensor<?x?xi32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xi32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xi32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xi32>
%3 = linalg.indexed_generic {
indexing_maps = [#map0, #map0],
@@ -415,8 +415,8 @@ func @indexed_generic_op_generic_op_fusion(%arg0: tensor<?x?xi32>,
func @indexed_generic_op_fusion(%arg0: tensor<?x?xi32>) -> tensor<?x?xi32> {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : tensor<?x?xi32>
- %1 = dim %arg0, %c1 : tensor<?x?xi32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xi32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xi32>
%2 = linalg.init_tensor [%0, %1] : tensor<?x?xi32>
%3 = linalg.indexed_generic {
indexing_maps = [#map0, #map0],
@@ -468,7 +468,7 @@ func @indexed_generic_op_fusion(%arg0: tensor<?x?xi32>) -> tensor<?x?xi32> {
func @scalar_indexed_generic_fusion
(%arg0: tensor<5x1x1xf32>, %arg1 : tensor<i32>) -> tensor<10xf32>
{
- %c0 = constant 0 : index
+ %c0 = constant 0 : index
%cst = constant dense<1.000000e+00> : tensor<10xf32>
%0 = linalg.init_tensor [] : tensor<f32>
%1 = linalg.indexed_generic
diff --git a/mlir/test/Dialect/Linalg/fusion.mlir b/mlir/test/Dialect/Linalg/fusion.mlir
index 026807dcebd3..14fb995a6cfc 100644
--- a/mlir/test/Dialect/Linalg/fusion.mlir
+++ b/mlir/test/Dialect/Linalg/fusion.mlir
@@ -11,22 +11,22 @@ func @f1(%A: memref<?x?xf32, offset: 0, strides: [?, 1]>,
%c3 = constant 3 : index
%c2 = constant 2 : index
%c1 = constant 1 : index
- %0 = dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, 1]>
- %1 = dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, 1]>
- %2 = dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, 1]>
+ %0 = memref.dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, 1]>
+ %1 = memref.dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, 1]>
+ %2 = memref.dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, 1]>
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, 1]>,
memref<?x?xf32, offset: 0, strides: [?, 1]>)
outs(%C : memref<?x?xf32, offset: 0, strides: [?, 1]>)
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
- %5 = std.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, 1]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %B[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %B[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, 1]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %C[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %C[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, 1]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -62,19 +62,19 @@ func @f2(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%C: memref<?x?xf32, offset: 0, strides: [?, ?]>)
- %0 = dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
- %5 = std.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -87,9 +87,9 @@ func @f2(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
}
// CHECK-LABEL: func @f2
// CHECK: (%[[A:.*]]:{{.*}}, %[[B:.*]]:{{.*}}, %[[C:.*]]:{{.*}}, %[[D:.*]]:{{.*}}, %[[E:.*]]:{{.*}})
-// CHECK-DAG: %[[C_0:.*]] = dim %[[C]], %c0{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
-// CHECK-DAG: %[[C_1:.*]] = dim %[[C]], %c1{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
-// CHECK-DAG: %[[D_1:.*]] = dim %[[D]], %c1{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[C_0:.*]] = memref.dim %[[C]], %c0{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[C_1:.*]] = memref.dim %[[C]], %c1{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[D_1:.*]] = memref.dim %[[D]], %c1{{[_0-9]*}} : memref<?x?xf32, #[[$strided2D]]>
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_0]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[D_1]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_1]] step %{{.*}} {
@@ -114,19 +114,19 @@ func @f3(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%C : memref<?x?xf32, offset: 0, strides: [?, ?]>)
- %0 = dim %D, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %D, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
- %5 = std.subview %D[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %D[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %C[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %C[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -141,9 +141,9 @@ func @f3(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
// CHECK: (%[[A:.*]]:{{.*}}, %[[B:.*]]:{{.*}}, %[[C:.*]]:{{.*}}, %[[D:.*]]:{{.*}}, %[[E:.*]]:{{.*}})
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[D_0:.*]] = dim %[[D]], %[[C0]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[D_1:.*]] = dim %[[D]], %[[C1]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[C_1:.*]] = dim %[[C]], %[[C1]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[D_0:.*]] = memref.dim %[[D]], %[[C0]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[D_1:.*]] = memref.dim %[[D]], %[[C1]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[C_1:.*]] = memref.dim %[[C]], %[[C1]] : memref<?x?xf32, #[[$strided2D]]>
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[D_0]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_1]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[D_1]] step %{{.*}} {
@@ -171,19 +171,19 @@ func @f4(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%D : memref<?x?xf32, offset: 0, strides: [?, ?]>)
- %0 = dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
- %5 = std.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -198,9 +198,9 @@ func @f4(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
// CHECK: (%[[A:.*]]:{{.*}}, %[[B:.*]]:{{.*}}, %[[C:.*]]:{{.*}}, %[[D:.*]]:{{.*}}, %[[E:.*]]:{{.*}})
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[C_0:.*]] = dim %[[C]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[C_1:.*]] = dim %[[C]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[D_1:.*]] = dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[C_0:.*]] = memref.dim %[[C]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[C_1:.*]] = memref.dim %[[C]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[D_1:.*]] = memref.dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_0]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[D_1]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_1]] step %{{.*}} {
@@ -223,9 +223,9 @@ func @f5(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
%c4 = constant 4 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %0 = dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %D, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %D, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%C : memref<?x?xf32, offset: 0, strides: [?, ?]>)
@@ -235,13 +235,13 @@ func @f5(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
scf.for %arg5 = %c0 to %1 step %c2 {
scf.for %arg6 = %c0 to %0 step %c3 {
scf.for %arg7 = %c0 to %2 step %c4 {
- %5 = std.subview %D[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %D[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %7 = std.subview %B[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %B[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -256,19 +256,19 @@ func @f5(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
// CHECK: (%[[A:.*]]:{{.*}}, %[[B:.*]]:{{.*}}, %[[C:.*]]:{{.*}}, %[[D:.*]]:{{.*}}, %[[E:.*]]:{{.*}})
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
-// CHECK-DAG: %[[B_1:.*]] = dim %[[B]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK-DAG: %[[D_0:.*]] = dim %[[D]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK-DAG: %[[D_1:.*]] = dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK-DAG: %[[B_00:.*]] = subview %[[B]][0, 0]{{.*}}
+// CHECK-DAG: %[[B_1:.*]] = memref.dim %[[B]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[D_0:.*]] = memref.dim %[[D]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[D_1:.*]] = memref.dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK-DAG: %[[B_00:.*]] = memref.subview %[[B]][0, 0]{{.*}}
// CHECK: scf.for %[[I:.*]] = %{{.*}} to %[[D_0]] step %{{.*}} {
-// CHECK-DAG: %[[A_I0:.*]] = subview %[[A]][%[[I]], 0]
-// CHECK-DAG: %[[C_I0:.*]] = subview %[[C]][%[[I]], 0]
+// CHECK-DAG: %[[A_I0:.*]] = memref.subview %[[A]][%[[I]], 0]
+// CHECK-DAG: %[[C_I0:.*]] = memref.subview %[[C]][%[[I]], 0]
// CHECK: scf.for %[[J:.*]] = %{{.*}} to %[[B_1]] step %{{.*}} {
-// CHECK: %[[E_IJ:.*]] = subview %[[E]][%[[I]], %[[J]]]
+// CHECK: %[[E_IJ:.*]] = memref.subview %[[E]][%[[I]], %[[J]]]
// CHECK: scf.for %[[K:.*]] = %{{.*}} to %[[D_1]] step %{{.*}} {
-// CHECK-DAG: %[[D_IK:.*]] = subview %[[D]][%[[I]], %[[K]]]
-// CHECK-DAG: %[[B_0K:.*]] = subview %[[B]][0, %[[K]]]
-// CHECK-DAG: %[[B_KJ:.*]] = subview %[[B]][%[[K]], %[[J]]]
+// CHECK-DAG: %[[D_IK:.*]] = memref.subview %[[D]][%[[I]], %[[K]]]
+// CHECK-DAG: %[[B_0K:.*]] = memref.subview %[[B]][0, %[[K]]]
+// CHECK-DAG: %[[B_KJ:.*]] = memref.subview %[[B]][%[[K]], %[[J]]]
// CHECK: linalg.matmul ins(%[[A_I0]], %[[B_00]]{{.*}} outs(%[[C_I0]]
// CHECK: linalg.matmul ins(%[[C_I0]], %[[B_0K]]{{.*}} outs(%[[D_IK]]
// CHECK: linalg.matmul ins(%[[D_IK]], %[[B_KJ]]{{.*}} outs(%[[E_IJ]]
@@ -290,28 +290,28 @@ func @f6(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
%c4 = constant 4 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %0 = dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%C : memref<?x?xf32, offset: 0, strides: [?, ?]>)
linalg.matmul ins(%A, %C : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%E : memref<?x?xf32, offset: 0, strides: [?, ?]>)
- %1 = dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg5 = %c0 to %1 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %0 step %c4 {
%3 = affine.apply #map0(%arg5)
%4 = affine.apply #map1(%arg7)
- %5 = std.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
%6 = affine.apply #map2(%arg6)
- %7 = std.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -345,11 +345,11 @@ func @f7(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
%c4 = constant 4 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %0 = dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %2 = dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %3 = dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %4 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %C, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %3 = memref.dim %C, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %4 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
linalg.matmul ins(%A, %C : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%E : memref<?x?xf32, offset: 0, strides: [?, ?]>)
@@ -359,13 +359,13 @@ func @f7(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
- %7 = std.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %7 = memref.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %9 = std.subview %C[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %9 = memref.subview %C[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %10 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %10 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%7, %9 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -377,13 +377,13 @@ func @f7(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
scf.for %arg5 = %c0 to %3 step %c2 {
scf.for %arg6 = %c0 to %4 step %c3 {
scf.for %arg7 = %c0 to %2 step %c4 {
- %7 = std.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %7 = memref.subview %C[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %9 = std.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %9 = memref.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %10 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %10 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%7, %9 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -398,11 +398,11 @@ func @f7(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
// CHECK: (%[[A:.*]]:{{.*}}, %[[B:.*]]:{{.*}}, %[[C:.*]]:{{.*}}, %[[D:.*]]:{{.*}}, %[[E:.*]]:{{.*}})
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
-// CHECK: %[[A_0:.*]] = dim %[[A]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[A_1:.*]] = dim %[[A]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[C_1:.*]] = dim %[[C]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[C_0:.*]] = dim %[[C]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECK: %[[D_1:.*]] = dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[A_0:.*]] = memref.dim %[[A]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[A_1:.*]] = memref.dim %[[A]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[C_1:.*]] = memref.dim %[[C]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[C_0:.*]] = memref.dim %[[C]], %[[C0:.*]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[D_1:.*]] = memref.dim %[[D]], %[[C1:.*]] : memref<?x?xf32, #[[$strided2D]]>
// CHECK: linalg.matmul ins(%[[A]], %[[C]]{{.*}} outs(%[[E]]
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[A_0]] step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %[[C_1]] step %{{.*}} {
@@ -432,28 +432,28 @@ func @f8(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
%c4 = constant 4 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %0 = dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %A, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %A, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
linalg.matmul ins(%A, %C : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%D : memref<?x?xf32, offset: 0, strides: [?, ?]>)
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
outs(%C : memref<?x?xf32, offset: 0, strides: [?, ?]>)
- %2 = dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %2 = memref.dim %D, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg5 = %c0 to %0 step %c2 {
scf.for %arg6 = %c0 to %2 step %c3 {
scf.for %arg7 = %c0 to %1 step %c4 {
%3 = affine.apply #map0(%arg5)
%4 = affine.apply #map1(%arg7)
- %5 = std.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
+ %5 = memref.subview %A[%arg5, %arg7][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
%6 = affine.apply #map2(%arg6)
- %7 = std.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
+ %7 = memref.subview %D[%arg7, %arg6][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %8 = std.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
+ %8 = memref.subview %E[%arg5, %arg6][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%5, %7 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -497,17 +497,17 @@ func @pointwise(%A: memref<?x?xf32, offset: 0, strides: [?, ?]>,
%2 = addf %E, %arg5 : f32
linalg.yield %2 : f32
}
- %0 = dim %B, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %1 = dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %0 = memref.dim %B, %c0 : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %1 = memref.dim %B, %c1 : memref<?x?xf32, offset: 0, strides: [?, ?]>
scf.for %arg4 = %c0 to %0 step %c2 {
scf.for %arg5 = %c0 to %1 step %c3 {
- %4 = std.subview %B[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %4 = memref.subview %B[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %5 = std.subview %C[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %5 = memref.subview %C[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %6 = std.subview %D[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %6 = memref.subview %D[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.generic #pointwise_2d_trait
@@ -543,11 +543,11 @@ func @pointwise_no_view(%M: index, %N: index) {
%c0 = constant 0 : index
%c3 = constant 3 : index
%c2 = constant 2 : index
- %A = alloc (%M, %N): memref<?x?xf32>
- %B = alloc (%M, %N): memref<?x?xf32>
- %C = alloc (%M, %N): memref<?x?xf32>
- %D = alloc (%M, %N): memref<?x?xf32>
- %E = alloc (%M, %N): memref<?x?xf32>
+ %A = memref.alloc (%M, %N): memref<?x?xf32>
+ %B = memref.alloc (%M, %N): memref<?x?xf32>
+ %C = memref.alloc (%M, %N): memref<?x?xf32>
+ %D = memref.alloc (%M, %N): memref<?x?xf32>
+ %E = memref.alloc (%M, %N): memref<?x?xf32>
linalg.generic #pointwise_2d_trait
ins(%A, %A : memref<?x?xf32>, memref<?x?xf32>)
outs(%B : memref<?x?xf32>) {
@@ -555,17 +555,17 @@ func @pointwise_no_view(%M: index, %N: index) {
%2 = addf %e, %arg5 : f32
linalg.yield %2 : f32
}
- %0 = dim %B, %c0 : memref<?x?xf32>
- %1 = dim %B, %c1 : memref<?x?xf32>
+ %0 = memref.dim %B, %c0 : memref<?x?xf32>
+ %1 = memref.dim %B, %c1 : memref<?x?xf32>
scf.for %arg4 = %c0 to %0 step %c2 {
scf.for %arg5 = %c0 to %1 step %c3 {
- %4 = std.subview %B[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %4 = memref.subview %B[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %5 = std.subview %C[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %5 = memref.subview %C[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %6 = std.subview %D[%arg4, %arg5][%c2, %c3][%c1, %c1] :
+ %6 = memref.subview %D[%arg4, %arg5][%c2, %c3][%c1, %c1] :
memref<?x?xf32> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.generic #pointwise_2d_trait
@@ -601,7 +601,7 @@ func @fusion_of_three(%arg0: memref<100x10xf32>,
%arg2: memref<100x10xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = alloc() {temp = true} : memref<100x10xf32>
+ %0 = memref.alloc() {temp = true} : memref<100x10xf32>
linalg.generic {
indexing_maps = [#map0, #map1],
iterator_types = ["parallel", "parallel"]}
@@ -610,7 +610,7 @@ func @fusion_of_three(%arg0: memref<100x10xf32>,
^bb0(%arg3: f32, %arg4: f32): // no predecessors
linalg.yield %arg3 : f32
}
- %1 = alloc() {temp = true} : memref<100x10xf32>
+ %1 = memref.alloc() {temp = true} : memref<100x10xf32>
linalg.generic {
indexing_maps = [#map1, #map1, #map1],
iterator_types = ["parallel", "parallel"]}
@@ -620,16 +620,16 @@ func @fusion_of_three(%arg0: memref<100x10xf32>,
%2 = subf %arg3, %arg4 : f32
linalg.yield %2 : f32
}
- dealloc %0 : memref<100x10xf32>
- %2 = dim %1, %c0 : memref<100x10xf32>
- %3 = dim %1, %c1 : memref<100x10xf32>
- %4 = dim %arg2, %c0 : memref<100x10xf32>
- %5 = dim %arg2, %c1 : memref<100x10xf32>
+ memref.dealloc %0 : memref<100x10xf32>
+ %2 = memref.dim %1, %c0 : memref<100x10xf32>
+ %3 = memref.dim %1, %c1 : memref<100x10xf32>
+ %4 = memref.dim %arg2, %c0 : memref<100x10xf32>
+ %5 = memref.dim %arg2, %c1 : memref<100x10xf32>
scf.for %i = %c0 to %2 step %c1 {
scf.for %j = %c0 to %3 step %c1 {
- %6 = std.subview %1[%i, %j][%c1, %c1][%c1, %c1] :
+ %6 = memref.subview %1[%i, %j][%c1, %c1][%c1, %c1] :
memref<100x10xf32> to memref<?x?xf32, #map2>
- %7 = std.subview %arg2[%i, %j][%c1, %c1][%c1, %c1] :
+ %7 = memref.subview %arg2[%i, %j][%c1, %c1][%c1, %c1] :
memref<100x10xf32> to memref<?x?xf32, #map2>
linalg.generic {
indexing_maps = [#map1, #map1],
@@ -642,7 +642,7 @@ func @fusion_of_three(%arg0: memref<100x10xf32>,
}
}
}
- dealloc %1 : memref<100x10xf32>
+ memref.dealloc %1 : memref<100x10xf32>
return
}
// CHECK-LABEL: func @fusion
@@ -674,15 +674,15 @@ func @fill_and_conv(%arg0: memref<?x?x?x?xf32>, %arg1: memref<2x3x1x1xf32>, %arg
%c0 = constant 0 : index
%c2 = constant 2 : index
%c3 = constant 3 : index
- %4 = dim %arg1, %c0 : memref<2x3x1x1xf32>
- %5 = dim %arg1, %c1 : memref<2x3x1x1xf32>
- %6 = dim %arg0, %c0 : memref<?x?x?x?xf32>
- %7 = dim %arg0, %c1 : memref<?x?x?x?xf32>
- %8 = dim %arg0, %c3 : memref<?x?x?x?xf32>
- %9 = dim %arg2, %c0 : memref<?x?x?x?xf32>
- %10 = dim %arg2, %c1 : memref<?x?x?x?xf32>
- %11 = dim %arg2, %c2 : memref<?x?x?x?xf32>
- %12 = dim %arg2, %c3 : memref<?x?x?x?xf32>
+ %4 = memref.dim %arg1, %c0 : memref<2x3x1x1xf32>
+ %5 = memref.dim %arg1, %c1 : memref<2x3x1x1xf32>
+ %6 = memref.dim %arg0, %c0 : memref<?x?x?x?xf32>
+ %7 = memref.dim %arg0, %c1 : memref<?x?x?x?xf32>
+ %8 = memref.dim %arg0, %c3 : memref<?x?x?x?xf32>
+ %9 = memref.dim %arg2, %c0 : memref<?x?x?x?xf32>
+ %10 = memref.dim %arg2, %c1 : memref<?x?x?x?xf32>
+ %11 = memref.dim %arg2, %c2 : memref<?x?x?x?xf32>
+ %12 = memref.dim %arg2, %c3 : memref<?x?x?x?xf32>
%13 = linalg.range %c0 : %6 : %c2 : !linalg.range
%14 = linalg.range %c0 : %10 : %c3 : !linalg.range
scf.for %arg3 = %c0 to %6 step %c2 {
@@ -690,14 +690,14 @@ func @fill_and_conv(%arg0: memref<?x?x?x?xf32>, %arg1: memref<2x3x1x1xf32>, %arg
%15 = affine.min #map0(%c2, %c1, %arg3)
%16 = affine.apply #map2()[%7]
%17 = affine.min #map0(%16, %c4, %arg4)
- %18 = dim %arg0, %c2 : memref<?x?x?x?xf32>
- %19 = dim %arg0, %c3 : memref<?x?x?x?xf32>
- %20 = subview %arg0[%arg3, %arg4, %c0, %c0] [%15, %17, %18, %19] [%c1, %c1, %c1, %c1] : memref<?x?x?x?xf32> to memref<?x?x?x?xf32, #map1>
+ %18 = memref.dim %arg0, %c2 : memref<?x?x?x?xf32>
+ %19 = memref.dim %arg0, %c3 : memref<?x?x?x?xf32>
+ %20 = memref.subview %arg0[%arg3, %arg4, %c0, %c0] [%15, %17, %18, %19] [%c1, %c1, %c1, %c1] : memref<?x?x?x?xf32> to memref<?x?x?x?xf32, #map1>
%21 = affine.min #map0(%c2, %c1, %arg3)
%22 = affine.min #map0(%c3, %c4, %arg4)
- %23 = dim %arg2, %c2 : memref<?x?x?x?xf32>
- %24 = dim %arg2, %c3 : memref<?x?x?x?xf32>
- %25 = subview %arg2[%arg3, %arg4, %c0, %c0] [%21, %22, %23, %24] [%c1, %c1, %c1, %c1] : memref<?x?x?x?xf32> to memref<?x?x?x?xf32, #map1>
+ %23 = memref.dim %arg2, %c2 : memref<?x?x?x?xf32>
+ %24 = memref.dim %arg2, %c3 : memref<?x?x?x?xf32>
+ %25 = memref.subview %arg2[%arg3, %arg4, %c0, %c0] [%21, %22, %23, %24] [%c1, %c1, %c1, %c1] : memref<?x?x?x?xf32> to memref<?x?x?x?xf32, #map1>
linalg.conv(%arg1, %20, %25) {dilations = [1, 1], strides = [1, 1]} : memref<2x3x1x1xf32>, memref<?x?x?x?xf32, #map1>, memref<?x?x?x?xf32, #map1>
}
}
@@ -719,9 +719,9 @@ func @accept_
diff erent_alloc_ops(%dim: index, %s0 : index, %s1: index) {
%c3 = constant 3 : index
%c4 = constant 4 : index
- %A = alloca(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %B = alloca(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
- %C = alloc(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %A = memref.alloca(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %B = memref.alloca(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
+ %C = memref.alloc(%dim, %dim)[%s0, %s1] : memref<?x?xf32, offset: 0, strides: [?, ?]>
linalg.matmul ins(%A, %B : memref<?x?xf32, offset: 0, strides: [?, ?]>,
memref<?x?xf32, offset: 0, strides: [?, ?]>)
@@ -730,13 +730,13 @@ func @accept_
diff erent_alloc_ops(%dim: index, %s0 : index, %s1: index) {
scf.for %i = %c0 to %dim step %c2 {
scf.for %j = %c0 to %dim step %c3 {
scf.for %k = %c0 to %dim step %c4 {
- %0 = std.subview %A[%i, %k][%c2, %c4][%c1, %c1] :
+ %0 = memref.subview %A[%i, %k][%c2, %c4][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %1 = std.subview %B[%k, %j][%c4, %c3][%c1, %c1] :
+ %1 = memref.subview %B[%k, %j][%c4, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
- %2 = std.subview %C[%i, %j][%c2, %c3][%c1, %c1] :
+ %2 = memref.subview %C[%i, %j][%c2, %c3][%c1, %c1] :
memref<?x?xf32, offset: 0, strides: [?, ?]> to
memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul ins(%0, %1 : memref<?x?xf32, offset: ?, strides: [?, ?]>,
diff --git a/mlir/test/Dialect/Linalg/hoist-padding.mlir b/mlir/test/Dialect/Linalg/hoist-padding.mlir
index a6afe0cbbdfe..969974aa7e52 100644
--- a/mlir/test/Dialect/Linalg/hoist-padding.mlir
+++ b/mlir/test/Dialect/Linalg/hoist-padding.mlir
@@ -29,12 +29,12 @@ func @matmul_tensors(
%c0 = constant 0 : index
%c1 = constant 1 : index
- // CHECK-DAG: %[[dM:.*]] = dim %[[TA]], %[[C0]] : tensor<?x?xf32>
- // CHECK-DAG: %[[dK:.*]] = dim %[[TA]], %[[C1]] : tensor<?x?xf32>
- // CHECK-DAG: %[[dN:.*]] = dim %[[TB]], %[[C1]] : tensor<?x?xf32>
- %0 = dim %arg0, %c0 : tensor<?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?xf32>
- %2 = dim %arg1, %c1 : tensor<?x?xf32>
+ // CHECK-DAG: %[[dM:.*]] = memref.dim %[[TA]], %[[C0]] : tensor<?x?xf32>
+ // CHECK-DAG: %[[dK:.*]] = memref.dim %[[TA]], %[[C1]] : tensor<?x?xf32>
+ // CHECK-DAG: %[[dN:.*]] = memref.dim %[[TB]], %[[C1]] : tensor<?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?xf32>
+ %2 = memref.dim %arg1, %c1 : tensor<?x?xf32>
// CHECK: scf.for %[[I:[0-9a-z]+]] =
// First padded tensor is MxKx2x4 under loop M so Kx2x4
@@ -85,19 +85,19 @@ func @matmul_tensors(
%3 = scf.for %arg3 = %c0 to %0 step %c2 iter_args(%arg4 = %arg2) -> (tensor<?x?xf32>) {
%4 = scf.for %arg5 = %c0 to %2 step %c3 iter_args(%arg6 = %arg4) -> (tensor<?x?xf32>) {
%5 = scf.for %arg7 = %c0 to %1 step %c4 iter_args(%arg8 = %arg6) -> (tensor<?x?xf32>) {
- %6 = dim %arg0, %c0 : tensor<?x?xf32>
+ %6 = memref.dim %arg0, %c0 : tensor<?x?xf32>
%7 = affine.min #map0(%arg3)[%6]
- %8 = dim %arg0, %c1 : tensor<?x?xf32>
+ %8 = memref.dim %arg0, %c1 : tensor<?x?xf32>
%9 = affine.min #map1(%arg7)[%8]
%10 = subtensor %arg0[%arg3, %arg7] [%7, %9] [1, 1] : tensor<?x?xf32> to tensor<?x?xf32>
- %11 = dim %arg1, %c0 : tensor<?x?xf32>
+ %11 = memref.dim %arg1, %c0 : tensor<?x?xf32>
%12 = affine.min #map1(%arg7)[%11]
- %13 = dim %arg1, %c1 : tensor<?x?xf32>
+ %13 = memref.dim %arg1, %c1 : tensor<?x?xf32>
%14 = affine.min #map2(%arg5)[%13]
%15 = subtensor %arg1[%arg7, %arg5] [%12, %14] [1, 1] : tensor<?x?xf32> to tensor<?x?xf32>
- %16 = dim %arg8, %c0 : tensor<?x?xf32>
+ %16 = memref.dim %arg8, %c0 : tensor<?x?xf32>
%17 = affine.min #map3(%16, %arg3)
- %18 = dim %arg8, %c1 : tensor<?x?xf32>
+ %18 = memref.dim %arg8, %c1 : tensor<?x?xf32>
%19 = affine.min #map4(%18, %arg5)
%20 = subtensor %arg8[%arg3, %arg5] [%17, %19] [1, 1] : tensor<?x?xf32> to tensor<?x?xf32>
%21 = subi %c2, %7 : index
diff --git a/mlir/test/Dialect/Linalg/invalid.mlir b/mlir/test/Dialect/Linalg/invalid.mlir
index 084fd4abcf2a..28ff66abfe62 100644
--- a/mlir/test/Dialect/Linalg/invalid.mlir
+++ b/mlir/test/Dialect/Linalg/invalid.mlir
@@ -3,7 +3,7 @@
func @load_number_of_indices(%v : memref<f32>) {
// expected-error @+2 {{incorrect number of indices for load}}
%c0 = constant 0 : index
- load %v[%c0] : memref<f32>
+ memref.load %v[%c0] : memref<f32>
}
// -----
@@ -12,7 +12,7 @@ func @store_number_of_indices(%v : memref<f32>) {
// expected-error @+3 {{store index operand count not equal to memref rank}}
%c0 = constant 0 : index
%f0 = constant 0.0 : f32
- store %f0, %v[%c0] : memref<f32>
+ memref.store %f0, %v[%c0] : memref<f32>
}
// -----
diff --git a/mlir/test/Dialect/Linalg/loops.mlir b/mlir/test/Dialect/Linalg/loops.mlir
index 71fd7c7cf15a..5be3525854f9 100644
--- a/mlir/test/Dialect/Linalg/loops.mlir
+++ b/mlir/test/Dialect/Linalg/loops.mlir
@@ -34,9 +34,9 @@
func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %A = view %arg0[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
- %B = view %arg0[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
- %C = view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
+ %A = memref.view %arg0[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
+ %B = memref.view %arg0[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
+ %C = memref.view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
linalg.matmul ins(%A, %B: memref<?x?xf32>, memref<?x?xf32>)
outs(%C: memref<?x?xf32>)
return
@@ -45,16 +45,16 @@ func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECKLOOP-SAME: [[M:arg[0-9]+]]: index
// CHECKLOOP-SAME: [[N:arg[0-9]+]]: index
// CHECKLOOP-SAME: [[K:arg[0-9]+]]: index
-// CHECKLOOP: %[[A:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKLOOP: %[[B:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKLOOP: %[[C:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKLOOP: %[[A:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKLOOP: %[[B:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKLOOP: %[[C:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[M]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[N]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKLOOP-DAG: %[[a:.*]] = load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKLOOP-DAG: %[[b:.*]] = load %[[B]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKLOOP-DAG: %[[c:.*]] = load %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP-DAG: %[[c:.*]] = memref.load %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -62,15 +62,15 @@ func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECKPARALLEL-SAME: [[M:arg[0-9]+]]: index
// CHECKPARALLEL-SAME: [[N:arg[0-9]+]]: index
// CHECKPARALLEL-SAME: [[K:arg[0-9]+]]: index
-// CHECKPARALLEL: %[[A:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKPARALLEL: %[[B:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKPARALLEL: %[[C:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKPARALLEL: %[[A:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKPARALLEL: %[[B:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKPARALLEL: %[[C:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[M]], %[[N]]) step (%{{.*}}, %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKPARALLEL-DAG: %[[a:.*]] = load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKPARALLEL-DAG: %[[b:.*]] = load %[[B]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKPARALLEL-DAG: %[[c:.*]] = load %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL-DAG: %[[c:.*]] = memref.load %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[C]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -79,9 +79,9 @@ func @matmul(%arg0: memref<?xi8>, %M: index, %N: index, %K: index) {
func @matvec(%arg0: memref<?xi8>, %M: index, %N: index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %2 = view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
- %3 = view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
- %4 = view %arg0[%c0][%N] : memref<?xi8> to memref<?xf32>
+ %2 = memref.view %arg0[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
+ %3 = memref.view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
+ %4 = memref.view %arg0[%c0][%N] : memref<?xi8> to memref<?xf32>
linalg.matvec ins(%2, %3: memref<?x?xf32>, memref<?xf32>)
outs(%4 : memref<?xf32>)
return
@@ -89,30 +89,30 @@ func @matvec(%arg0: memref<?xi8>, %M: index, %N: index) {
// CHECKLOOP-LABEL: func @matvec(%{{.*}}: memref<?xi8>,
// CHECKLOOP-SAME: [[M:arg[0-9]+]]: index
// CHECKLOOP-SAME: [[K:arg[0-9]+]]: index
-// CHECKLOOP: %[[A:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKLOOP: %[[B:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKLOOP: %[[C:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKLOOP: %[[A:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKLOOP: %[[B:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKLOOP: %[[C:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[M]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKLOOP-DAG: %[[a:.*]] = load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKLOOP-DAG: %[[b:.*]] = load %[[B]][%{{.*}}] : memref<?xf32>
+// CHECKLOOP-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}] : memref<?xf32>
// CHECKLOOP-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKLOOP-DAG: %[[c:.*]] = load %[[C]][%{{.*}}] : memref<?xf32>
+// CHECKLOOP-DAG: %[[c:.*]] = memref.load %[[C]][%{{.*}}] : memref<?xf32>
// CHECKLOOP-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %[[C]][%{{.*}}] : memref<?xf32>
// CHECKPARALLEL-LABEL: func @matvec(%{{.*}}: memref<?xi8>,
// CHECKPARALLEL-SAME: [[M:arg[0-9]+]]: index
// CHECKPARALLEL-SAME: [[K:arg[0-9]+]]: index
-// CHECKPARALLEL: %[[A:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECKPARALLEL: %[[B:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKPARALLEL: %[[C:.*]] = std.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKPARALLEL: %[[A:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECKPARALLEL: %[[B:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKPARALLEL: %[[C:.*]] = memref.view %{{.*}}[{{.*}}] : memref<?xi8> to memref<?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}) = (%{{.*}}) to (%[[M]]) step (%{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKPARALLEL-DAG: %[[a:.*]] = load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKPARALLEL-DAG: %[[b:.*]] = load %[[B]][%{{.*}}] : memref<?xf32>
+// CHECKPARALLEL-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}] : memref<?xf32>
// CHECKPARALLEL-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKPARALLEL-DAG: %[[c:.*]] = load %[[C]][%{{.*}}] : memref<?xf32>
+// CHECKPARALLEL-DAG: %[[c:.*]] = memref.load %[[C]][%{{.*}}] : memref<?xf32>
// CHECKPARALLEL-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[C]][%{{.*}}] : memref<?xf32>
@@ -120,36 +120,36 @@ func @matvec(%arg0: memref<?xi8>, %M: index, %N: index) {
func @dot(%arg0: memref<?xi8>, %M: index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %1 = view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
- %2 = view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
- %3 = view %arg0[%c0][] : memref<?xi8> to memref<f32>
+ %1 = memref.view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
+ %2 = memref.view %arg0[%c0][%M] : memref<?xi8> to memref<?xf32>
+ %3 = memref.view %arg0[%c0][] : memref<?xi8> to memref<f32>
linalg.dot ins(%1, %2 : memref<?xf32>, memref<?xf32>)
outs(%3 : memref<f32>)
return
}
// CHECKLOOP-LABEL: func @dot(%{{.*}}: memref<?xi8>,
// CHECKLOOP-SAME: [[K:arg[0-9]+]]: index
-// CHECKLOOP: %[[A:.*]] = std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKLOOP: %[[B:.*]] = std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKLOOP: %[[C:.*]] = std.view %{{.*}}[{{.*}}][] : memref<?xi8> to memref<f32>
+// CHECKLOOP: %[[A:.*]] = memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKLOOP: %[[B:.*]] = memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKLOOP: %[[C:.*]] = memref.view %{{.*}}[{{.*}}][] : memref<?xi8> to memref<f32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKLOOP-DAG: %[[a:.*]] = load %[[A]][%{{.*}}] : memref<?xf32>
-// CHECKLOOP-DAG: %[[b:.*]] = load %[[B]][%{{.*}}] : memref<?xf32>
+// CHECKLOOP-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}] : memref<?xf32>
+// CHECKLOOP-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}] : memref<?xf32>
// CHECKLOOP-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKLOOP-DAG: %[[c:.*]] = load %[[C]][] : memref<f32>
+// CHECKLOOP-DAG: %[[c:.*]] = memref.load %[[C]][] : memref<f32>
// CHECKLOOP-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %[[C]][] : memref<f32>
// CHECKPARALLEL-LABEL: func @dot(%{{.*}}: memref<?xi8>,
// CHECKPARALLEL-SAME: [[K:arg[0-9]+]]: index
-// CHECKPARALLEL: %[[A:.*]] = std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKPARALLEL: %[[B:.*]] = std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
-// CHECKPARALLEL: %[[C:.*]] = std.view %{{.*}}[{{.*}}][] : memref<?xi8> to memref<f32>
+// CHECKPARALLEL: %[[A:.*]] = memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKPARALLEL: %[[B:.*]] = memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?xf32>
+// CHECKPARALLEL: %[[C:.*]] = memref.view %{{.*}}[{{.*}}][] : memref<?xi8> to memref<f32>
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKPARALLEL-DAG: %[[a:.*]] = load %[[A]][%{{.*}}] : memref<?xf32>
-// CHECKPARALLEL-DAG: %[[b:.*]] = load %[[B]][%{{.*}}] : memref<?xf32>
+// CHECKPARALLEL-DAG: %[[a:.*]] = memref.load %[[A]][%{{.*}}] : memref<?xf32>
+// CHECKPARALLEL-DAG: %[[b:.*]] = memref.load %[[B]][%{{.*}}] : memref<?xf32>
// CHECKPARALLEL-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKPARALLEL-DAG: %[[c:.*]] = load %[[C]][] : memref<f32>
+// CHECKPARALLEL-DAG: %[[c:.*]] = memref.load %[[C]][] : memref<f32>
// CHECKPARALLEL-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[C]][] : memref<f32>
@@ -162,23 +162,23 @@ func @dot_view(%arg0: memref<?xf32, offset: ?, strides: [1]>, %arg1: memref<?xf3
}
// CHECKLOOP-LABEL: func @dot_view(
// CHECKLOOP: %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<f32>) {
-// CHECKLOOP: %[[K:.*]] = dim %arg0, %c0 : memref<?xf32, #[[$strided1D]]>
+// CHECKLOOP: %[[K:.*]] = memref.dim %arg0, %c0 : memref<?xf32, #[[$strided1D]]>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKLOOP-DAG: %[[a:.*]] = load %arg0[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
-// CHECKLOOP-DAG: %[[b:.*]] = load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKLOOP-DAG: %[[a:.*]] = memref.load %arg0[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKLOOP-DAG: %[[b:.*]] = memref.load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
// CHECKLOOP-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKLOOP-DAG: %[[c:.*]] = load %{{.*}}[] : memref<f32>
+// CHECKLOOP-DAG: %[[c:.*]] = memref.load %{{.*}}[] : memref<f32>
// CHECKLOOP-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %{{.*}}[] : memref<f32>
// CHECKPARALLEL-LABEL: func @dot_view(
// CHECKPARALLEL: %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<f32>) {
-// CHECKPARALLEL: %[[K:.*]] = dim %arg0, %c0 : memref<?xf32, #[[$strided1D]]>
+// CHECKPARALLEL: %[[K:.*]] = memref.dim %arg0, %c0 : memref<?xf32, #[[$strided1D]]>
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKPARALLEL-DAG: %[[a:.*]] = load %arg0[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
-// CHECKPARALLEL-DAG: %[[b:.*]] = load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKPARALLEL-DAG: %[[a:.*]] = memref.load %arg0[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKPARALLEL-DAG: %[[b:.*]] = memref.load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
// CHECKPARALLEL-DAG: %[[inc:.*]] = mulf %[[a]], %[[b]] : f32
-// CHECKPARALLEL-DAG: %[[c:.*]] = load %{{.*}}[] : memref<f32>
+// CHECKPARALLEL-DAG: %[[c:.*]] = memref.load %{{.*}}[] : memref<f32>
// CHECKPARALLEL-DAG: %[[res:.*]] = addf %[[c]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %{{.*}}[] : memref<f32>
@@ -229,13 +229,13 @@ func @copy_view(%arg0: memref<?xf32, offset: ?, strides: [1]>, %arg1: memref<?xf
// CHECKLOOP-LABEL: func @copy_view(
// CHECKLOOP: %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<?xf32, #[[$strided1D]]>) {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECKLOOP: %[[L:.*]] = load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKLOOP: %[[L:.*]] = memref.load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
// CHECKLOOP: store %[[L]], %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
// CHECKPARALLEL-LABEL: func @copy_view(
// CHECKPARALLEL: %{{.*}}: memref<?xf32, #[[$strided1D]]>, %{{.*}}: memref<?xf32, #[[$strided1D]]>) {
// CHECKPARALLEL: scf.parallel (%{{.*}}) = (%{{.*}}) to (%{{.*}}) step (%{{.*}}) {
-// CHECKPARALLEL: %[[L:.*]] = load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
+// CHECKPARALLEL: %[[L:.*]] = memref.load %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
// CHECKPARALLEL: store %[[L]], %{{.*}}[%{{.*}}] : memref<?xf32, #[[$strided1D]]>
func @copy_view0(%arg0: memref<f32>, %arg1: memref<f32>) {
@@ -243,11 +243,11 @@ func @copy_view0(%arg0: memref<f32>, %arg1: memref<f32>) {
return
}
// CHECKLOOP-LABEL: func @copy_view0(%{{.*}}: memref<f32>, %{{.*}}: memref<f32>) {
-// CHECKLOOP: %{{.*}} = load %{{.*}}[] : memref<f32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[] : memref<f32>
// CHECKLOOP: store %{{.*}}, %{{.*}}[] : memref<f32>
// CHECKPARALLEL-LABEL: func @copy_view0(%{{.*}}: memref<f32>, %{{.*}}: memref<f32>) {
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[] : memref<f32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[] : memref<f32>
// CHECKPARALLEL: store %{{.*}}, %{{.*}}[] : memref<f32>
func @copy_view3(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>) {
@@ -261,13 +261,13 @@ func @copy_view3(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1:
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECKLOOP: %[[L:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[L:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKLOOP: store %[[L]], %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL-LABEL: func @copy_view3
// CHECKPARALLEL: (%{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>) {
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}, %{{.*}}) to (%{{.*}}, %{{.*}}, %{{.*}}) step (%{{.*}}, %{{.*}}, %{{.*}}) {
-// CHECKPARALLEL: %[[L:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[L:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL: store %[[L]], %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
func @conv_view3(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg2: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>) {
@@ -276,39 +276,39 @@ func @conv_view3(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1:
}
// CHECKLOOP-LABEL: func @conv_view3(
// CHECKLOOP: %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>) {
-// CHECKLOOP: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %[[Q:.*]] = dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %[[K:.*]] = dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[Q:.*]] = memref.dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[K:.*]] = memref.dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[B]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[X0]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[K]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[Q]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[Z0]] step %{{.*}} {
// CHECKLOOP: %[[SUM:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %[[SUM]], %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[SUM]], %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKLOOP: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKLOOP: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKLOOP: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL-LABEL: func @conv_view3(
// CHECKPARALLEL: %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>, %{{.*}}: memref<?x?x?xf32, #[[$strided3D]]>) {
-// CHECKPARALLEL: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %[[Q:.*]] = dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %[[K:.*]] = dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[Q:.*]] = memref.dim %arg0, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[K:.*]] = memref.dim %arg0, %c2 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}, %{{.*}}) to (%[[B]], %[[X0]], %[[K]]) step (%{{.*}}, %{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Q]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Z0]] step %{{.*}} {
// CHECKPARALLEL: %[[SUM:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %[[SUM]], %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[SUM]], %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKPARALLEL: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?xf32, #[[$strided3D]]>
@@ -318,13 +318,13 @@ func @conv_view4(%arg0: memref<?x?x?x?xf32, offset: ?, strides: [?, ?, ?, 1]>, %
}
// CHECKLOOP-LABEL: func @conv_view4(
// CHECKLOOP: %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>, %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>, %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>) {
-// CHECKLOOP: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[Z1:.*]] = dim %arg0, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[Q:.*]] = dim %arg0, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[K:.*]] = dim %arg0, %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %[[X1:.*]] = dim %arg2, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[Z1:.*]] = memref.dim %arg0, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[Q:.*]] = memref.dim %arg0, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[K:.*]] = memref.dim %arg0, %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %[[X1:.*]] = memref.dim %arg2, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[B]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[X0]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[X1]] step %{{.*}} {
@@ -334,32 +334,32 @@ func @conv_view4(%arg0: memref<?x?x?x?xf32, offset: ?, strides: [?, ?, ?, 1]>, %
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[Z1]] step %{{.*}} {
// CHECKLOOP: %[[SUM0:.*]] = affine.apply #[[$stride2Dilation4]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[SUM1:.*]] = affine.apply #[[$stride3Dilation5]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %[[SUM0]], %[[SUM1]], %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[SUM0]], %[[SUM1]], %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKLOOP: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKLOOP: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKLOOP: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKPARALLEL-LABEL: func @conv_view4(
// CHECKPARALLEL: %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>, %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>, %{{.*}}: memref<?x?x?x?xf32, #[[$strided4D]]>) {
-// CHECKPARALLEL: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[Z1:.*]] = dim %arg0, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[Q:.*]] = dim %arg0, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[K:.*]] = dim %arg0, %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %[[X1:.*]] = dim %arg2, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[Z1:.*]] = memref.dim %arg0, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[Q:.*]] = memref.dim %arg0, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[K:.*]] = memref.dim %arg0, %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %[[X1:.*]] = memref.dim %arg2, %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) to (%[[B]], %[[X0]], %[[X1]], %[[K]]) step (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Q]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Z0]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Z1]] step %{{.*}} {
// CHECKPARALLEL: %[[SUM0:.*]] = affine.apply #[[$stride2Dilation4]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[SUM1:.*]] = affine.apply #[[$stride3Dilation5]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %[[SUM0]], %[[SUM1]], %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[SUM0]], %[[SUM1]], %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKPARALLEL: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
// CHECKPARALLEL: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKPARALLEL: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32, #[[$strided4D]]>
@@ -375,13 +375,13 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32>,
// CHECKLOOP-LABEL: func @conv_padding
// CHECKLOOP: %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>) {
// CHECKLOOP: %[[ZERO:.*]] = constant 0.000000e+00 : f32
-// CHECKLOOP: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[Z1:.*]] = dim %arg0, %c1 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[Q:.*]] = dim %arg0, %c2 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[K:.*]] = dim %arg0, %c3 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?x?xf32>
-// CHECKLOOP: %[[X1:.*]] = dim %arg2, %c2 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[Z1:.*]] = memref.dim %arg0, %c1 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[Q:.*]] = memref.dim %arg0, %c2 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[K:.*]] = memref.dim %arg0, %c3 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?x?xf32>
+// CHECKLOOP: %[[X1:.*]] = memref.dim %arg2, %c2 : memref<?x?x?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[B]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[X0]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[X1]] step %{{.*}} {
@@ -393,24 +393,24 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32>,
// CHECKLOOP: %[[SUM1:.*]] = affine.apply #{{.*}}(%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[SUM0]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[SUM1]])
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
// CHECKLOOP: %{{.*}} = select %{{.*}}, %{{.*}}, %{{.*}} : f32
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECKLOOP: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECKLOOP: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKLOOP: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECKPARALLEL-LABEL: func @conv_padding
// CHECKPARALLEL: %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>, %{{.*}}: memref<?x?x?x?xf32>) {
// CHECKPARALLEL: %[[ZERO:.*]] = constant 0.000000e+00 : f32
-// CHECKPARALLEL: %[[Z0:.*]] = dim %arg0, %c0 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[Z1:.*]] = dim %arg0, %c1 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[Q:.*]] = dim %arg0, %c2 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[K:.*]] = dim %arg0, %c3 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[B:.*]] = dim %arg1, %c0 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[X0:.*]] = dim %arg2, %c1 : memref<?x?x?x?xf32>
-// CHECKPARALLEL: %[[X1:.*]] = dim %arg2, %c2 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[Z0:.*]] = memref.dim %arg0, %c0 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[Z1:.*]] = memref.dim %arg0, %c1 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[Q:.*]] = memref.dim %arg0, %c2 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[K:.*]] = memref.dim %arg0, %c3 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[B:.*]] = memref.dim %arg1, %c0 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[X0:.*]] = memref.dim %arg2, %c1 : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %[[X1:.*]] = memref.dim %arg2, %c2 : memref<?x?x?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) to (%[[B]], %[[X0]], %[[X1]], %[[K]]) step (%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Q]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[Z0]] step %{{.*}} {
@@ -419,11 +419,11 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32>,
// CHECKPARALLEL: %[[SUM1:.*]] = affine.apply #{{.*}}(%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[SUM0]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[SUM1]])
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %[[IDX]], %[[IDY]], %{{.*}}] : memref<?x?x?x?xf32>
// CHECKPARALLEL: %{{.*}} = select %{{.*}}, %{{.*}}, %{{.*}} : f32
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECKPARALLEL: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
// CHECKPARALLEL: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
// CHECKPARALLEL: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}, %{{.*}}, %{{.*}}] : memref<?x?x?x?xf32>
@@ -435,33 +435,33 @@ func @pooling_max(%arg0: memref<?x?xf32>,
return
}
// CHECKLOOP-LABEL: func @pooling_max
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %{{.*}}, %{{.*}} : f32
// CHECKLOOP: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-LABEL: func @pooling_max
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %{{.*}}, %{{.*}} : f32
// CHECKPARALLEL: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -474,20 +474,20 @@ func @pooling_max_padding(%arg0: memref<?x?xf32>,
}
// CHECKLOOP-LABEL: func @pooling_max_padding
// CHECKLOOP: %[[PAD:.*]] = constant 0xFF800000 : f32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
// CHECKLOOP: %[[CMP:.*]] = cmpf ogt, %[[RHS]], %[[SEL]] : f32
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : f32
@@ -495,19 +495,19 @@ func @pooling_max_padding(%arg0: memref<?x?xf32>,
// CHECKPARALLEL-LABEL: func @pooling_max_padding
// CHECKPARALLEL: %[[PAD:.*]] = constant 0xFF800000 : f32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
// CHECKPARALLEL: %[[CMP:.*]] = cmpf ogt, %[[RHS]], %[[SEL]] : f32
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : f32
@@ -522,20 +522,20 @@ func @pooling_max_padding_i32(%arg0: memref<?x?xi32>,
}
// CHECKLOOP-LABEL: func @pooling_max_padding_i32
// CHECKLOOP: %[[PAD:.*]] = constant -2147483648 : i32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
// CHECKLOOP: %[[CMP:.*]] = cmpi sgt, %[[RHS]], %[[SEL]] : i32
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : i32
@@ -543,19 +543,19 @@ func @pooling_max_padding_i32(%arg0: memref<?x?xi32>,
// CHECKPARALLEL-LABEL: func @pooling_max_padding_i32
// CHECKPARALLEL: %[[PAD:.*]] = constant -2147483648 : i32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
// CHECKPARALLEL: %[[CMP:.*]] = cmpi sgt, %[[RHS]], %[[SEL]] : i32
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : i32
@@ -569,33 +569,33 @@ func @pooling_min(%arg0: memref<?x?xf32>,
return
}
// CHECKLOOP-LABEL: func @pooling_min
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKLOOP: %{{.*}} = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %{{.*}} = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %{{.*}}, %{{.*}} : f32
// CHECKLOOP: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-LABEL: func @pooling_min
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
-// CHECKPARALLEL: %{{.*}} = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %{{.*}} = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %{{.*}}, %{{.*}} : f32
// CHECKPARALLEL: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -608,20 +608,20 @@ func @pooling_min_padding(%arg0: memref<?x?xf32>,
}
// CHECKLOOP-LABEL: func @pooling_min_padding
// CHECKLOOP: %[[PAD:.*]] = constant 0x7F800000 : f32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
// CHECKLOOP: %[[CMP:.*]] = cmpf olt, %[[RHS]], %[[SEL]] : f32
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : f32
@@ -629,19 +629,19 @@ func @pooling_min_padding(%arg0: memref<?x?xf32>,
// CHECKPARALLEL-LABEL: func @pooling_min_padding
// CHECKPARALLEL: %[[PAD:.*]] = constant 0x7F800000 : f32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
// CHECKPARALLEL: %[[CMP:.*]] = cmpf olt, %[[RHS]], %[[SEL]] : f32
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : f32
@@ -656,20 +656,20 @@ func @pooling_min_padding_i32(%arg0: memref<?x?xi32>,
}
// CHECKLOOP-LABEL: func @pooling_min_padding_i32
// CHECKLOOP: %[[PAD:.*]] = constant 2147483647 : i32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
// CHECKLOOP: %[[CMP:.*]] = cmpi slt, %[[RHS]], %[[SEL]] : i32
// CHECKLOOP: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : i32
@@ -677,19 +677,19 @@ func @pooling_min_padding_i32(%arg0: memref<?x?xi32>,
// CHECKPARALLEL-LABEL: func @pooling_min_padding_i32
// CHECKPARALLEL: %[[PAD:.*]] = constant 2147483647 : i32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride1Dilation1Padding2]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
// CHECKPARALLEL: %[[CMP:.*]] = cmpi slt, %[[RHS]], %[[SEL]] : i32
// CHECKPARALLEL: %[[RES:.*]] = select %{{.*}}, %[[RHS]], %[[SEL]] : i32
@@ -703,33 +703,33 @@ func @pooling_sum(%arg0: memref<?x?xf32>,
return
}
// CHECKLOOP-LABEL: func @pooling_sum
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKLOOP: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP: %[[RES:.*]] = addf %[[LHS]], %[[RHS]] : f32
// CHECKLOOP: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-LABEL: func @pooling_sum
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
// CHECKPARALLEL: %[[IX:.*]] = affine.apply #[[$stride2Dilation1]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1]](%{{.*}}, %{{.*}})
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%[[IX]], %[[IY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL: %[[RES:.*]] = addf %[[LHS]], %[[RHS]] : f32
// CHECKPARALLEL: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -742,10 +742,10 @@ func @pooling_sum_padding(%arg0: memref<?x?xf32>,
}
// CHECKLOOP-LABEL: func @pooling_sum_padding
// CHECKLOOP: %[[PAD:.*]] = constant 0.000000e+00 : f32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
@@ -754,18 +754,18 @@ func @pooling_sum_padding(%arg0: memref<?x?xf32>,
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKLOOP: %[[RES:.*]] = addf %[[RHS]], %[[SEL]] : f32
// CHECKLOOP: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL-LABEL: func @pooling_sum_padding
// CHECKPARALLEL: %[[PAD:.*]] = constant 0.000000e+00 : f32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xf32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xf32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xf32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
@@ -773,9 +773,9 @@ func @pooling_sum_padding(%arg0: memref<?x?xf32>,
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : f32
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECKPARALLEL: %[[RES:.*]] = addf %[[RHS]], %[[SEL]] : f32
// CHECKPARALLEL: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
@@ -788,10 +788,10 @@ func @pooling_sum_padding_i32(%arg0: memref<?x?xi32>,
}
// CHECKLOOP-LABEL: func @pooling_sum_padding_i32
// CHECKLOOP: %[[PAD:.*]] = constant 0 : i32
-// CHECKLOOP: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKLOOP: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKLOOP: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKLOOP: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKLOOP: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OX]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[OY]] step %{{.*}} {
// CHECKLOOP: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
@@ -800,18 +800,18 @@ func @pooling_sum_padding_i32(%arg0: memref<?x?xi32>,
// CHECKLOOP: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
// CHECKLOOP: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKLOOP: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKLOOP: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKLOOP: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKLOOP: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
-// CHECKLOOP: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKLOOP: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKLOOP: %[[RES:.*]] = addi %[[RHS]], %[[SEL]] : i32
// CHECKLOOP: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKPARALLEL-LABEL: func @pooling_sum_padding_i32
// CHECKPARALLEL: %[[PAD:.*]] = constant 0 : i32
-// CHECKPARALLEL: %[[WX:.*]] = dim %arg1, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[WY:.*]] = dim %arg1, %c1 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OX:.*]] = dim %arg2, %c0 : memref<?x?xi32>
-// CHECKPARALLEL: %[[OY:.*]] = dim %arg2, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WX:.*]] = memref.dim %arg1, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[WY:.*]] = memref.dim %arg1, %c1 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OX:.*]] = memref.dim %arg2, %c0 : memref<?x?xi32>
+// CHECKPARALLEL: %[[OY:.*]] = memref.dim %arg2, %c1 : memref<?x?xi32>
// CHECKPARALLEL: scf.parallel (%{{.*}}, %{{.*}}) = (%{{.*}}, %{{.*}}) to (%[[OX]], %[[OY]]) step (%{{.*}}, %{{.*}}) {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WX]] step %{{.*}} {
// CHECKPARALLEL: scf.for %{{.*}} = %{{.*}} to %[[WY]] step %{{.*}} {
@@ -819,9 +819,9 @@ func @pooling_sum_padding_i32(%arg0: memref<?x?xi32>,
// CHECKPARALLEL: %[[IY:.*]] = affine.apply #[[$stride1Dilation1Padding1]](%{{.*}}, %{{.*}})
// CHECKPARALLEL: %[[IDX:.*]] = affine.max #[[$clampMinMap]](%[[IX]])
// CHECKPARALLEL: %[[IDY:.*]] = affine.max #[[$clampMinMap]](%[[IY]])
-// CHECKPARALLEL: %[[LHS:.*]] = load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
+// CHECKPARALLEL: %[[LHS:.*]] = memref.load %{{.*}}[%[[IDX]], %[[IDY]]] : memref<?x?xi32>
// CHECKPARALLEL: %[[SEL:.*]] = select %{{.*}}, %[[PAD]], %[[LHS]] : i32
-// CHECKPARALLEL: %[[RHS:.*]] = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
+// CHECKPARALLEL: %[[RHS:.*]] = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
// CHECKPARALLEL: %[[RES:.*]] = addi %[[RHS]], %[[SEL]] : i32
// CHECKPARALLEL: store %[[RES]], %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xi32>
@@ -854,9 +854,9 @@ func @generic_region(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>, %arg1:
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[j:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[k:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECKLOOP: %[[b:.*]] = load %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKLOOP: %[[c:.*]] = load %{{.*}}[%[[i]], %[[k]], %[[j]]] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[a:.*]] = memref.load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECKLOOP: %[[b:.*]] = memref.load %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKLOOP: %[[c:.*]] = memref.load %{{.*}}[%[[i]], %[[k]], %[[j]]] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKLOOP: %[[d:.*]] = mulf %[[a]], %[[b]] : f32
// CHECKLOOP: %[[e:.*]] = addf %[[c]], %[[d]] : f32
// CHECKLOOP: store %[[d]], %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
@@ -864,9 +864,9 @@ func @generic_region(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>, %arg1:
// CHECKPARALLEL-LABEL: @generic_region
// CHECKPARALLEL: scf.parallel (%[[i:[a-zA-Z0-9_]*]], %[[j:[a-zA-Z0-9_]*]], %[[k:[a-zA-Z0-9_]*]])
-// CHECKPARALLEL: %[[a:.*]] = load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32, #[[$strided2D]]>
-// CHECKPARALLEL: %[[b:.*]] = load %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
-// CHECKPARALLEL: %[[c:.*]] = load %{{.*}}[%[[i]], %[[k]], %[[j]]] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[a:.*]] = memref.load %{{.*}}[%[[i]], %[[j]]] : memref<?x?xf32, #[[$strided2D]]>
+// CHECKPARALLEL: %[[b:.*]] = memref.load %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
+// CHECKPARALLEL: %[[c:.*]] = memref.load %{{.*}}[%[[i]], %[[k]], %[[j]]] : memref<?x?x?xf32, #[[$strided3D]]>
// CHECKPARALLEL: %[[d:.*]] = mulf %[[a]], %[[b]] : f32
// CHECKPARALLEL: %[[e:.*]] = addf %[[c]], %[[d]] : f32
// CHECKPARALLEL: store %[[d]], %{{.*}}[%[[i]], %[[j]], %[[k]]] : memref<?x?x?xf32, #[[$strided3D]]>
@@ -906,9 +906,9 @@ func @indexed_generic_region(
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[j:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[k:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %{{.*}}[%[[i]], %[[j]]]
-// CHECKLOOP: %[[b:.*]] = load %{{.*}}[%[[i]], %[[j]], %[[k]]]
-// CHECKLOOP: %[[c:.*]] = load %{{.*}}[%[[i]], %[[k]], %[[j]]]
+// CHECKLOOP: %[[a:.*]] = memref.load %{{.*}}[%[[i]], %[[j]]]
+// CHECKLOOP: %[[b:.*]] = memref.load %{{.*}}[%[[i]], %[[j]], %[[k]]]
+// CHECKLOOP: %[[c:.*]] = memref.load %{{.*}}[%[[i]], %[[k]], %[[j]]]
// CHECKLOOP: %[[result_1:.*]] = mulf %[[a]], %[[b]] : f32
// CHECKLOOP: %[[ij:.*]] = addi %[[i]], %[[j]] : index
// CHECKLOOP: %[[ijk:.*]] = addi %[[ij]], %[[k]] : index
@@ -920,9 +920,9 @@ func @indexed_generic_region(
// CHECKPARALLEL-LABEL: @indexed_generic_region
// CHECKPARALLEL: scf.parallel (%[[i:[a-zA-Z0-9_]*]], %[[j:[a-zA-Z0-9_]*]], %[[k:[a-zA-Z0-9_]*]])
-// CHECKPARALLEL: %[[a:.*]] = load %{{.*}}[%[[i]], %[[j]]]
-// CHECKPARALLEL: %[[b:.*]] = load %{{.*}}[%[[i]], %[[j]], %[[k]]]
-// CHECKPARALLEL: %[[c:.*]] = load %{{.*}}[%[[i]], %[[k]], %[[j]]]
+// CHECKPARALLEL: %[[a:.*]] = memref.load %{{.*}}[%[[i]], %[[j]]]
+// CHECKPARALLEL: %[[b:.*]] = memref.load %{{.*}}[%[[i]], %[[j]], %[[k]]]
+// CHECKPARALLEL: %[[c:.*]] = memref.load %{{.*}}[%[[i]], %[[k]], %[[j]]]
// CHECKPARALLEL: %[[result_1:.*]] = mulf %[[a]], %[[b]] : f32
// CHECKPARALLEL: %[[ij:.*]] = addi %[[i]], %[[j]] : index
// CHECKPARALLEL: %[[ijk:.*]] = addi %[[ij]], %[[k]] : index
@@ -963,14 +963,14 @@ func @generic_op_zero_rank(%arg0: memref<f32>, %arg1: memref<3x4xf32>)
// CHECKLOOP-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<3x4xf32>
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[j:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %[[ARG0]][]
+// CHECKLOOP: %[[a:.*]] = memref.load %[[ARG0]][]
// CHECKLOOP: store %[[a]], %[[ARG1]][%[[i]], %[[j]]]
// CHECKPARALLEL-LABEL: @generic_op_zero_rank
// CHECKPARALLEL-SAME: %[[ARG0:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<3x4xf32>
// CHECKPARALLEL: scf.parallel (%[[i:[a-zA-Z0-9_]*]], %[[j:[a-zA-Z0-9_]*]])
-// CHECKPARALLEL: %[[a:.*]] = load %[[ARG0]][]
+// CHECKPARALLEL: %[[a:.*]] = memref.load %[[ARG0]][]
// CHECKPARALLEL: store %[[a]], %[[ARG1]][%[[i]], %[[j]]]
func @indexed_generic_op_zero_rank(%arg0: memref<i32>, %arg1: memref<3x4xi32>)
@@ -992,7 +992,7 @@ func @indexed_generic_op_zero_rank(%arg0: memref<i32>, %arg1: memref<3x4xi32>)
// CHECKLOOP-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<3x4xi32>
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
// CHECKLOOP: scf.for %[[j:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %[[ARG0]][
+// CHECKLOOP: %[[a:.*]] = memref.load %[[ARG0]][
// CHECKLOOP: %[[ij:.*]] = addi %[[i]], %[[j]] : index
// CHECKLOOP: %[[ij_int:.*]] = index_cast %[[ij]] : index to i32
// CHECKLOOP: %[[result:.*]] = addi %[[a]], %[[ij_int]] : i32
@@ -1002,7 +1002,7 @@ func @indexed_generic_op_zero_rank(%arg0: memref<i32>, %arg1: memref<3x4xi32>)
// CHECKPARALLEL-SAME: %[[ARG0:[a-zA-Z0-9_]*]]: memref<i32>
// CHECKPARALLEL-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<3x4xi32>
// CHECKPARALLEL: scf.parallel (%[[i:[a-zA-Z0-9_]*]], %[[j:[a-zA-Z0-9_]*]])
-// CHECKPARALLEL: %[[a:.*]] = load %[[ARG0]][
+// CHECKPARALLEL: %[[a:.*]] = memref.load %[[ARG0]][
// CHECKPARALLEL: %[[ij:.*]] = addi %[[i]], %[[j]] : index
// CHECKPARALLEL: %[[ij_int:.*]] = index_cast %[[ij]] : index to i32
// CHECKPARALLEL: %[[result:.*]] = addi %[[a]], %[[ij_int]] : i32
@@ -1036,8 +1036,8 @@ func @generic_op_1D_reduce(%arg0: memref<?xf32>, %arg1: memref<f32>)
// CHECKLOOP-SAME: %[[ARG0:[a-zA-Z0-9_]*]]: memref<?xf32>
// CHECKLOOP-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %[[ARG0]][%[[i]]]
-// CHECKLOOP: %[[b:.*]] = load %[[ARG1]][]
+// CHECKLOOP: %[[a:.*]] = memref.load %[[ARG0]][%[[i]]]
+// CHECKLOOP: %[[b:.*]] = memref.load %[[ARG1]][]
// CHECKLOOP: %[[c:.*]] = addf %[[a]], %[[b]] : f32
// CHECKLOOP: store %[[c]], %[[ARG1]][]
@@ -1045,8 +1045,8 @@ func @generic_op_1D_reduce(%arg0: memref<?xf32>, %arg1: memref<f32>)
// CHECKPARALLEL-SAME: %[[ARG0:[a-zA-Z0-9_]*]]: memref<?xf32>
// CHECKPARALLEL-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL: scf.for %[[i:.*]] = {{.*}}
-// CHECKPARALLEL: %[[a:.*]] = load %[[ARG0]][%[[i]]]
-// CHECKPARALLEL: %[[b:.*]] = load %[[ARG1]][]
+// CHECKPARALLEL: %[[a:.*]] = memref.load %[[ARG0]][%[[i]]]
+// CHECKPARALLEL: %[[b:.*]] = memref.load %[[ARG1]][]
// CHECKPARALLEL: %[[c:.*]] = addf %[[a]], %[[b]] : f32
// CHECKPARALLEL: store %[[c]], %[[ARG1]][]
@@ -1086,9 +1086,9 @@ func @indexed_generic_op_1D_reduce(%arg0: memref<?xf32>,
// CHECKLOOP-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKLOOP-SAME: %[[ARG2:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKLOOP: scf.for %[[i:.*]] = {{.*}}
-// CHECKLOOP: %[[a:.*]] = load %[[ARG0]][%[[i]]]
-// CHECKLOOP: %[[b:.*]] = load %[[ARG1]][]
-// CHECKLOOP: %[[c:.*]] = load %[[ARG2]][]
+// CHECKLOOP: %[[a:.*]] = memref.load %[[ARG0]][%[[i]]]
+// CHECKLOOP: %[[b:.*]] = memref.load %[[ARG1]][]
+// CHECKLOOP: %[[c:.*]] = memref.load %[[ARG2]][]
// CHECKLOOP: %[[d:.*]] = select %{{.*}}, %[[b]], %[[c]]
// CHECKLOOP: %[[e:.*]] = addf %[[a]], %[[d]]
// CHECKLOOP: store %[[e]], %[[ARG2]][]
@@ -1098,9 +1098,9 @@ func @indexed_generic_op_1D_reduce(%arg0: memref<?xf32>,
// CHECKPARALLEL-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL-SAME: %[[ARG2:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL: scf.for %[[i:.*]] = {{.*}}
-// CHECKPARALLEL: %[[a:.*]] = load %[[ARG0]][%[[i]]]
-// CHECKPARALLEL: %[[b:.*]] = load %[[ARG1]][]
-// CHECKPARALLEL: %[[c:.*]] = load %[[ARG2]][]
+// CHECKPARALLEL: %[[a:.*]] = memref.load %[[ARG0]][%[[i]]]
+// CHECKPARALLEL: %[[b:.*]] = memref.load %[[ARG1]][]
+// CHECKPARALLEL: %[[c:.*]] = memref.load %[[ARG2]][]
// CHECKPARALLEL: %[[d:.*]] = select %{{.*}}, %[[b]], %[[c]]
// CHECKPARALLEL: %[[e:.*]] = addf %[[a]], %[[d]]
// CHECKPARALLEL: store %[[e]], %[[ARG2]][]
@@ -1164,8 +1164,8 @@ func @scalar_code(%arg0: memref<f32>, %arg1 : memref<f32>, %arg2 : memref<f32>,
// CHECKLOOP-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKLOOP-SAME: %[[ARG2:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKLOOP-NOT: scf.for
-// CHECKLOOP: load %[[ARG0]][]
-// CHECKLOOP: load %[[ARG1]][]
+// CHECKLOOP: memref.load %[[ARG0]][]
+// CHECKLOOP: memref.load %[[ARG1]][]
// CHECKLOOP: scf.if
// CHECKLOOP: scf.yield
// CHECKLOOP: else
@@ -1177,8 +1177,8 @@ func @scalar_code(%arg0: memref<f32>, %arg1 : memref<f32>, %arg2 : memref<f32>,
// CHECKPARALLEL-SAME: %[[ARG1:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL-SAME: %[[ARG2:[a-zA-Z0-9_]*]]: memref<f32>
// CHECKPARALLEL-NOT: scf.for
-// CHECKPARALLEL: load %[[ARG0]][]
-// CHECKPARALLEL: load %[[ARG1]][]
+// CHECKPARALLEL: memref.load %[[ARG0]][]
+// CHECKPARALLEL: memref.load %[[ARG1]][]
// CHECKPARALLEL: scf.if
// CHECKPARALLEL: scf.yield
// CHECKPARALLEL: else
@@ -1197,17 +1197,17 @@ func @named_batch_matmul(%A: memref<?x?x?xf32>, %B: memref<?x?x?xf32>, %C: memre
// CHECKLOOP-SAME: %[[mA:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECKLOOP-SAME: %[[mB:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECKLOOP-SAME: %[[mC:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
-// CHECKLOOP: %[[B:.*]] = dim %[[mA]], %c0 : memref<?x?x?xf32>
-// CHECKLOOP: %[[M:.*]] = dim %[[mA]], %c1 : memref<?x?x?xf32>
-// CHECKLOOP: %[[K:.*]] = dim %[[mA]], %c2 : memref<?x?x?xf32>
-// CHECKLOOP: %[[N:.*]] = dim %[[mB]], %c2 : memref<?x?x?xf32>
+// CHECKLOOP: %[[B:.*]] = memref.dim %[[mA]], %c0 : memref<?x?x?xf32>
+// CHECKLOOP: %[[M:.*]] = memref.dim %[[mA]], %c1 : memref<?x?x?xf32>
+// CHECKLOOP: %[[K:.*]] = memref.dim %[[mA]], %c2 : memref<?x?x?xf32>
+// CHECKLOOP: %[[N:.*]] = memref.dim %[[mB]], %c2 : memref<?x?x?xf32>
// CHECKLOOP: scf.for %[[b:.*]] = %{{.*}} to %[[B]] step %{{.*}} {
// CHECKLOOP: scf.for %[[m:.*]] = %{{.*}} to %[[M]] step %{{.*}} {
// CHECKLOOP: scf.for %[[n:.*]] = %{{.*}} to %[[N]] step %{{.*}} {
// CHECKLOOP: scf.for %[[k:.*]] = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKLOOP: %[[va:.*]] = load %[[mA]][%[[b]], %[[m]], %[[k]]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[vb:.*]] = load %[[mB]][%[[b]], %[[k]], %[[n]]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[vc:.*]] = load %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[va:.*]] = memref.load %[[mA]][%[[b]], %[[m]], %[[k]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[vb:.*]] = memref.load %[[mB]][%[[b]], %[[k]], %[[n]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[vc:.*]] = memref.load %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
// CHECKLOOP: %[[inc:.*]] = mulf %[[va]], %[[vb]] : f32
// CHECKLOOP: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
@@ -1216,15 +1216,15 @@ func @named_batch_matmul(%A: memref<?x?x?xf32>, %B: memref<?x?x?xf32>, %C: memre
// CHECKPARALLEL-SAME: %[[mA:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECKPARALLEL-SAME: %[[mB:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
// CHECKPARALLEL-SAME: %[[mC:[a-zA-Z0-9]+]]: memref<?x?x?xf32>
-// CHECKPARALLEL: %[[B:.*]] = dim %[[mA]], %c0 : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[M:.*]] = dim %[[mA]], %c1 : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[K:.*]] = dim %[[mA]], %c2 : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[N:.*]] = dim %[[mB]], %c2 : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[B:.*]] = memref.dim %[[mA]], %c0 : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[M:.*]] = memref.dim %[[mA]], %c1 : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[K:.*]] = memref.dim %[[mA]], %c2 : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[N:.*]] = memref.dim %[[mB]], %c2 : memref<?x?x?xf32>
// CHECKPARALLEL: scf.parallel (%[[b:.*]], %[[m:.*]], %[[n:.*]]) = ({{.*}}) to (%[[B]], %[[M]], %[[N]]) step ({{.*}}) {
// CHECKPARALLEL: scf.for %[[k:.*]] = %{{.*}} to %[[K]] step %{{.*}} {
-// CHECKPARALLEL: %[[va:.*]] = load %[[mA]][%[[b]], %[[m]], %[[k]]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[vb:.*]] = load %[[mB]][%[[b]], %[[k]], %[[n]]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[vc:.*]] = load %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[va:.*]] = memref.load %[[mA]][%[[b]], %[[m]], %[[k]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[vb:.*]] = memref.load %[[mB]][%[[b]], %[[k]], %[[n]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[vc:.*]] = memref.load %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
// CHECKPARALLEL: %[[inc:.*]] = mulf %[[va]], %[[vb]] : f32
// CHECKPARALLEL: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[mC]][%[[b]], %[[m]], %[[n]]] : memref<?x?x?xf32>
@@ -1242,14 +1242,14 @@ func @conv1d_no_symbols(%in : memref<?xf32>, %filter : memref<?xf32>, %out : mem
// CHECKLOOP-SAME: %[[arg2:[a-zA-Z0-9]+]]: memref<?xf32>
// CHECKLOOP: %[[c0:.*]] = constant 0 : index
// CHECKLOOP: %[[c1:.*]] = constant 1 : index
-// CHECKLOOP: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?xf32>
-// CHECKLOOP: %[[dim1:.*]] = dim %[[arg2]], %[[c0]] : memref<?xf32>
+// CHECKLOOP: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?xf32>
+// CHECKLOOP: %[[dim1:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?xf32>
// CHECKLOOP: scf.for %[[b:.*]] = %[[c0]] to %[[dim1]] step %[[c1]] {
// CHECKLOOP: scf.for %[[m:.*]] = %[[c0]] to %[[dim0]] step %[[c1]] {
// CHECKLOOP: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[b]], %[[m]])
-// CHECKLOOP: %[[vb:.*]] = load %[[arg0]][%[[aff]]] : memref<?xf32>
-// CHECKLOOP: %[[va:.*]] = load %[[arg1]][%[[m]]] : memref<?xf32>
-// CHECKLOOP: %[[vc:.*]] = load %[[arg2]][%[[b]]] : memref<?xf32>
+// CHECKLOOP: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]]] : memref<?xf32>
+// CHECKLOOP: %[[va:.*]] = memref.load %[[arg1]][%[[m]]] : memref<?xf32>
+// CHECKLOOP: %[[vc:.*]] = memref.load %[[arg2]][%[[b]]] : memref<?xf32>
// CHECKLOOP: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKLOOP: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKLOOP: store %[[res]], %[[arg2]][%[[b]]] : memref<?xf32>
@@ -1260,14 +1260,14 @@ func @conv1d_no_symbols(%in : memref<?xf32>, %filter : memref<?xf32>, %out : mem
// CHECKPARALLEL-SAME: %[[arg2:[a-zA-Z0-9]+]]: memref<?xf32>
// CHECKPARALLEL: %[[c0:.*]] = constant 0 : index
// CHECKPARALLEL: %[[c1:.*]] = constant 1 : index
-// CHECKPARALLEL: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?xf32>
-// CHECKPARALLEL: %[[dim1:.*]] = dim %[[arg2]], %[[c0]] : memref<?xf32>
+// CHECKPARALLEL: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?xf32>
+// CHECKPARALLEL: %[[dim1:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?xf32>
// CHECKPARALLEL: scf.parallel (%[[b:.*]]) = (%[[c0]]) to (%[[dim1]]) step (%[[c1]]) {
// CHECKPARALLEL: scf.for %[[m:.*]] = %[[c0]] to %[[dim0]] step %[[c1]] {
// CHECKPARALLEL: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[b]], %[[m]])
-// CHECKPARALLEL: %[[vb:.*]] = load %[[arg0]][%[[aff]]] : memref<?xf32>
-// CHECKPARALLEL: %[[va:.*]] = load %[[arg1]][%[[m]]] : memref<?xf32>
-// CHECKPARALLEL: %[[vc:.*]] = load %[[arg2]][%[[b]]] : memref<?xf32>
+// CHECKPARALLEL: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]]] : memref<?xf32>
+// CHECKPARALLEL: %[[va:.*]] = memref.load %[[arg1]][%[[m]]] : memref<?xf32>
+// CHECKPARALLEL: %[[vc:.*]] = memref.load %[[arg2]][%[[b]]] : memref<?xf32>
// CHECKPARALLEL: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKPARALLEL: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[arg2]][%[[b]]] : memref<?xf32>
@@ -1284,20 +1284,20 @@ func @conv2d_no_symbols(%in : memref<?x?xf32>, %filter : memref<?x?xf32>, %out :
// CHECKLOOP-SAME: %[[arg2:[a-zA-Z0-9]+]]: memref<?x?xf32>
// CHECKLOOP: %[[c0:.*]] = constant 0 : index
// CHECKLOOP: %[[c1:.*]] = constant 1 : index
-// CHECKLOOP: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?x?xf32>
-// CHECKLOOP: %[[dim1:.*]] = dim %[[arg1]], %[[c1]] : memref<?x?xf32>
-// CHECKLOOP: %[[dim2:.*]] = dim %[[arg2]], %[[c0]] : memref<?x?xf32>
-// CHECKLOOP: %[[dim3:.*]] = dim %[[arg2]], %[[c1]] : memref<?x?xf32>
+// CHECKLOOP: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?x?xf32>
+// CHECKLOOP: %[[dim1:.*]] = memref.dim %[[arg1]], %[[c1]] : memref<?x?xf32>
+// CHECKLOOP: %[[dim2:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?x?xf32>
+// CHECKLOOP: %[[dim3:.*]] = memref.dim %[[arg2]], %[[c1]] : memref<?x?xf32>
// CHECKLOOP: scf.for %[[arg3:.*]] = %[[c0]] to %[[dim2]] step %[[c1]] {
// CHECKLOOP: scf.for %[[arg4:.*]] = %[[c0]] to %[[dim3]] step %[[c1]] {
// CHECKLOOP: scf.for %[[arg5:.*]] = %[[c0]] to %[[dim0]] step %[[c1]] {
// CHECKLOOP: scf.for %[[arg6:.*]] = %[[c0]] to %[[dim1]] step %[[c1]] {
// CHECKLOOP: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg3]], %[[arg5]])
// CHECKLOOP: %[[aff2:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg4]], %[[arg6]])
-// CHECKLOOP: %[[vb:.*]] = load %[[arg0]][%[[aff]], %[[aff2]]] : memref<?x?xf32>
+// CHECKLOOP: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]], %[[aff2]]] : memref<?x?xf32>
-// CHECKLOOP: %[[va:.*]] = load %[[arg1]][%[[arg5]], %[[arg6]]] : memref<?x?xf32>
-// CHECKLOOP: %[[vc:.*]] = load %[[arg2]][%[[arg3]], %[[arg4]]] : memref<?x?xf32>
+// CHECKLOOP: %[[va:.*]] = memref.load %[[arg1]][%[[arg5]], %[[arg6]]] : memref<?x?xf32>
+// CHECKLOOP: %[[vc:.*]] = memref.load %[[arg2]][%[[arg3]], %[[arg4]]] : memref<?x?xf32>
// CHECKLOOP: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKLOOP: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
@@ -1309,18 +1309,18 @@ func @conv2d_no_symbols(%in : memref<?x?xf32>, %filter : memref<?x?xf32>, %out :
// CHECKPARALLEL-SAME: %[[arg2:[a-zA-Z0-9]+]]: memref<?x?xf32>
// CHECKPARALLEL: %[[c0:.*]] = constant 0 : index
// CHECKPARALLEL: %[[c1:.*]] = constant 1 : index
-// CHECKPARALLEL: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[dim1:.*]] = dim %[[arg1]], %[[c1]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[dim2:.*]] = dim %[[arg2]], %[[c0]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[dim3:.*]] = dim %[[arg2]], %[[c1]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[dim1:.*]] = memref.dim %[[arg1]], %[[c1]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[dim2:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[dim3:.*]] = memref.dim %[[arg2]], %[[c1]] : memref<?x?xf32>
// CHECKPARALLEL: scf.parallel (%[[arg3:.*]], %[[arg4:.*]]) = (%[[c0]], %[[c0]]) to (%[[dim2]], %[[dim3]]) step (%[[c1]], %[[c1]]) {
// CHECKPARALLEL: scf.for %[[arg5:.*]] = %[[c0]] to %[[dim0]] step %[[c1]] {
// CHECKPARALLEL: scf.for %[[arg6:.*]] = %[[c0]] to %[[dim1]] step %[[c1]] {
// CHECKPARALLEL: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg3]], %[[arg5]])
// CHECKPARALLEL: %[[aff2:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg4]], %[[arg6]])
-// CHECKPARALLEL: %[[vb:.*]] = load %[[arg0]][%[[aff]], %[[aff2]]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[va:.*]] = load %[[arg1]][%[[arg5]], %[[arg6]]] : memref<?x?xf32>
-// CHECKPARALLEL: %[[vc:.*]] = load %[[arg2]][%[[arg3]], %[[arg4]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]], %[[aff2]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[va:.*]] = memref.load %[[arg1]][%[[arg5]], %[[arg6]]] : memref<?x?xf32>
+// CHECKPARALLEL: %[[vc:.*]] = memref.load %[[arg2]][%[[arg3]], %[[arg4]]] : memref<?x?xf32>
// CHECKPARALLEL: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKPARALLEL: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[arg2]][%[[arg3]], %[[arg4]]] : memref<?x?xf32>
@@ -1339,12 +1339,12 @@ func @conv3d_no_symbols(%in : memref<?x?x?xf32>, %filter : memref<?x?x?xf32>, %o
// CHECKLOOP: %[[c2:.*]] = constant 2 : index
// CHECKLOOP: %[[c0:.*]] = constant 0 : index
// CHECKLOOP: %[[c1:.*]] = constant 1 : index
-// CHECKLOOP: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[dim1:.*]] = dim %[[arg1]], %[[c1]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[dim2:.*]] = dim %[[arg1]], %[[c2]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[dim3:.*]] = dim %[[arg2]], %[[c0]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[dim4:.*]] = dim %[[arg2]], %[[c1]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[dim5:.*]] = dim %[[arg2]], %[[c2]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim1:.*]] = memref.dim %[[arg1]], %[[c1]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim2:.*]] = memref.dim %[[arg1]], %[[c2]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim3:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim4:.*]] = memref.dim %[[arg2]], %[[c1]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[dim5:.*]] = memref.dim %[[arg2]], %[[c2]] : memref<?x?x?xf32>
// CHECKLOOP: scf.for %[[arg3:.*]] = %[[c0]] to %[[dim3]] step %[[c1]] {
// CHECKLOOP: scf.for %[[arg4:.*]] = %[[c0]] to %[[dim4]] step %[[c1]] {
// CHECKLOOP: scf.for %[[arg5:.*]] = %[[c0]] to %[[dim5]] step %[[c1]] {
@@ -1354,10 +1354,10 @@ func @conv3d_no_symbols(%in : memref<?x?x?xf32>, %filter : memref<?x?x?xf32>, %o
// CHECKLOOP: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg3]], %[[arg6]])
// CHECKLOOP: %[[aff2:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg4]], %[[arg7]])
// CHECKLOOP: %[[aff3:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg5]], %[[arg8]])
-// CHECKLOOP: %[[vb:.*]] = load %[[arg0]][%[[aff]], %[[aff2]], %[[aff3]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]], %[[aff2]], %[[aff3]]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[va:.*]] = load %[[arg1]][%[[arg6]], %[[arg7]], %[[arg8]]] : memref<?x?x?xf32>
-// CHECKLOOP: %[[vc:.*]] = load %[[arg2]][%[[arg3]], %[[arg4]], %[[arg5]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[va:.*]] = memref.load %[[arg1]][%[[arg6]], %[[arg7]], %[[arg8]]] : memref<?x?x?xf32>
+// CHECKLOOP: %[[vc:.*]] = memref.load %[[arg2]][%[[arg3]], %[[arg4]], %[[arg5]]] : memref<?x?x?xf32>
// CHECKLOOP: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKLOOP: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
@@ -1370,12 +1370,12 @@ func @conv3d_no_symbols(%in : memref<?x?x?xf32>, %filter : memref<?x?x?xf32>, %o
// CHECKPARALLEL: %[[c2:.*]] = constant 2 : index
// CHECKPARALLEL: %[[c0:.*]] = constant 0 : index
// CHECKPARALLEL: %[[c1:.*]] = constant 1 : index
-// CHECKPARALLEL: %[[dim0:.*]] = dim %[[arg1]], %[[c0]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[dim1:.*]] = dim %[[arg1]], %[[c1]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[dim2:.*]] = dim %[[arg1]], %[[c2]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[dim3:.*]] = dim %[[arg2]], %[[c0]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[dim4:.*]] = dim %[[arg2]], %[[c1]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[dim5:.*]] = dim %[[arg2]], %[[c2]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim0:.*]] = memref.dim %[[arg1]], %[[c0]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim1:.*]] = memref.dim %[[arg1]], %[[c1]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim2:.*]] = memref.dim %[[arg1]], %[[c2]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim3:.*]] = memref.dim %[[arg2]], %[[c0]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim4:.*]] = memref.dim %[[arg2]], %[[c1]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[dim5:.*]] = memref.dim %[[arg2]], %[[c2]] : memref<?x?x?xf32>
// CHECKPARALLEL: scf.parallel (%[[arg3:.*]], %[[arg4:.*]], %[[arg5:.*]]) = (%[[c0]], %[[c0]], %[[c0]]) to (%[[dim3]], %[[dim4]], %[[dim5]]) step (%[[c1]], %[[c1]], %[[c1]]) {
// CHECKPARALLEL: scf.for %[[arg6:.*]] = %[[c0]] to %[[dim0]] step %[[c1]] {
// CHECKPARALLEL: scf.for %[[arg7:.*]] = %[[c0]] to %[[dim1]] step %[[c1]] {
@@ -1383,9 +1383,9 @@ func @conv3d_no_symbols(%in : memref<?x?x?xf32>, %filter : memref<?x?x?xf32>, %o
// CHECKPARALLEL: %[[aff:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg3]], %[[arg6]])
// CHECKPARALLEL: %[[aff2:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg4]], %[[arg7]])
// CHECKPARALLEL: %[[aff3:.*]] = affine.apply #[[$stride1Dilation1]](%[[arg5]], %[[arg8]])
-// CHECKPARALLEL: %[[vb:.*]] = load %[[arg0]][%[[aff]], %[[aff2]], %[[aff3]]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[va:.*]] = load %[[arg1]][%[[arg6]], %[[arg7]], %[[arg8]]] : memref<?x?x?xf32>
-// CHECKPARALLEL: %[[vc:.*]] = load %[[arg2]][%[[arg3]], %[[arg4]], %[[arg5]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[vb:.*]] = memref.load %[[arg0]][%[[aff]], %[[aff2]], %[[aff3]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[va:.*]] = memref.load %[[arg1]][%[[arg6]], %[[arg7]], %[[arg8]]] : memref<?x?x?xf32>
+// CHECKPARALLEL: %[[vc:.*]] = memref.load %[[arg2]][%[[arg3]], %[[arg4]], %[[arg5]]] : memref<?x?x?xf32>
// CHECKPARALLEL: %[[inc:.*]] = mulf %[[vb]], %[[va]] : f32
// CHECKPARALLEL: %[[res:.*]] = addf %[[vc]], %[[inc]] : f32
// CHECKPARALLEL: store %[[res]], %[[arg2]][%[[arg3]], %[[arg4]], %[[arg5]]] : memref<?x?x?xf32>
diff --git a/mlir/test/Dialect/Linalg/parallel-loops.mlir b/mlir/test/Dialect/Linalg/parallel-loops.mlir
index 8d365af6a5a3..500566251da5 100644
--- a/mlir/test/Dialect/Linalg/parallel-loops.mlir
+++ b/mlir/test/Dialect/Linalg/parallel-loops.mlir
@@ -21,8 +21,8 @@ func @linalg_generic_sum(%lhs: memref<2x2xf32>,
// CHECK-DAG: %[[C0:.*]] = constant 0
// CHECK-DAG: %[[C1:.*]] = constant 1
// CHECK: scf.parallel (%[[I:.*]], %[[J:.*]]) = {{.*}}
-// CHECK: %[[LHS_ELEM:.*]] = load %[[LHS]][%[[I]], %[[J]]]
-// CHECK: %[[RHS_ELEM:.*]] = load %[[RHS]][%[[I]], %[[J]]]
+// CHECK: %[[LHS_ELEM:.*]] = memref.load %[[LHS]][%[[I]], %[[J]]]
+// CHECK: %[[RHS_ELEM:.*]] = memref.load %[[RHS]][%[[I]], %[[J]]]
// CHECK: %[[SUM:.*]] = addf %[[LHS_ELEM]], %[[RHS_ELEM]] : f32
// CHECK: store %[[SUM]], %{{.*}}[%[[I]], %[[J]]]
// CHECK: scf.yield
@@ -50,14 +50,14 @@ func @lower_outer_parallel(%A: memref<?x?x?x?xf32>, %B: memref<?x?x?xf32>) {
// CHECK-LABEL: @lower_outer_parallel
// CHECK-DAG: %[[C0:.*]] = constant 0
// CHECK-DAG: %[[C1:.*]] = constant 1
-// CHECK-DAG: %[[D0:.*]] = dim %{{.*}}, %c0
-// CHECK-DAG: %[[D1:.*]] = dim %{{.*}}, %c1
-// CHECK-DAG: %[[D2:.*]] = dim %{{.*}}, %c2
-// CHECK-DAG: %[[D3:.*]] = dim %{{.*}}, %c3
+// CHECK-DAG: %[[D0:.*]] = memref.dim %{{.*}}, %c0
+// CHECK-DAG: %[[D1:.*]] = memref.dim %{{.*}}, %c1
+// CHECK-DAG: %[[D2:.*]] = memref.dim %{{.*}}, %c2
+// CHECK-DAG: %[[D3:.*]] = memref.dim %{{.*}}, %c3
// CHECK: scf.parallel (%[[IV0:.*]], %[[IV1:.*]]) = (%[[C0]], %[[C0]]) to (%[[D0]], %[[D1]]) step (%[[C1]], %[[C1]])
// CHECK: scf.for %[[IV2:.*]] = %[[C0]] to %[[D2]] step %[[C1]]
// CHECK: scf.parallel (%[[IV3:.*]]) = (%[[C0]]) to (%[[D3]]) step (%[[C1]])
-// CHECK: load %{{.*}}[%[[IV0]], %[[IV1]], %[[IV2]], %[[IV3]]]
+// CHECK: memref.load %{{.*}}[%[[IV0]], %[[IV1]], %[[IV2]], %[[IV3]]]
// CHECK: store %{{.*}}, %{{.*}}[%[[IV0]], %[[IV1]], %[[IV3]]]
// -----
@@ -83,15 +83,15 @@ func @lower_mixed_parallel(%A: memref<?x?x?x?x?x?xf32>, %B: memref<?x?x?x?xf32>)
// CHECK-LABEL: @lower_mixed_parallel
// CHECK-DAG: %[[C0:.*]] = constant 0
// CHECK-DAG: %[[C1:.*]] = constant 1
-// CHECK-DAG: %[[D0:.*]] = dim %{{.*}}, %c0
-// CHECK-DAG: %[[D1:.*]] = dim %{{.*}}, %c1
-// CHECK-DAG: %[[D2:.*]] = dim %{{.*}}, %c2
-// CHECK-DAG: %[[D3:.*]] = dim %{{.*}}, %c3
-// CHECK-DAG: %[[D4:.*]] = dim %{{.*}}, %c4
-// CHECK-DAG: %[[D5:.*]] = dim %{{.*}}, %c5
+// CHECK-DAG: %[[D0:.*]] = memref.dim %{{.*}}, %c0
+// CHECK-DAG: %[[D1:.*]] = memref.dim %{{.*}}, %c1
+// CHECK-DAG: %[[D2:.*]] = memref.dim %{{.*}}, %c2
+// CHECK-DAG: %[[D3:.*]] = memref.dim %{{.*}}, %c3
+// CHECK-DAG: %[[D4:.*]] = memref.dim %{{.*}}, %c4
+// CHECK-DAG: %[[D5:.*]] = memref.dim %{{.*}}, %c5
// CHECK: scf.parallel (%[[IV0:.*]], %[[IV1:.*]]) = (%[[C0]], %[[C0]]) to (%[[D0]], %[[D1]]) step (%[[C1]], %[[C1]])
// CHECK: scf.for %[[IV2:.*]] = %[[C0]] to %[[D2]] step %[[C1]]
// CHECK: scf.parallel (%[[IV3:.*]], %[[IV4:.*]]) = (%[[C0]], %[[C0]]) to (%[[D3]], %[[D4]]) step (%[[C1]], %[[C1]])
// CHECK: scf.for %[[IV5:.*]] = %[[C0]] to %[[D5]] step %[[C1]]
-// CHECK: load %{{.*}}[%[[IV0]], %[[IV1]], %[[IV2]], %[[IV3]], %[[IV4]], %[[IV5]]]
+// CHECK: memref.load %{{.*}}[%[[IV0]], %[[IV1]], %[[IV2]], %[[IV3]], %[[IV4]], %[[IV5]]]
// CHECK: store %{{.*}}, %{{.*}}[%[[IV0]], %[[IV1]], %[[IV4]], %[[IV3]]]
diff --git a/mlir/test/Dialect/Linalg/promote.mlir b/mlir/test/Dialect/Linalg/promote.mlir
index 838699301639..af8cd6867641 100644
--- a/mlir/test/Dialect/Linalg/promote.mlir
+++ b/mlir/test/Dialect/Linalg/promote.mlir
@@ -14,18 +14,18 @@ func @matmul_f32(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %3 = view %A[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
- %4 = view %A[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
- %5 = view %A[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
- %6 = dim %3, %c0 : memref<?x?xf32>
- %7 = dim %3, %c1 : memref<?x?xf32>
- %8 = dim %4, %c1 : memref<?x?xf32>
+ %3 = memref.view %A[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
+ %4 = memref.view %A[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
+ %5 = memref.view %A[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
+ %6 = memref.dim %3, %c0 : memref<?x?xf32>
+ %7 = memref.dim %3, %c1 : memref<?x?xf32>
+ %8 = memref.dim %4, %c1 : memref<?x?xf32>
scf.for %arg4 = %c0 to %6 step %c2 {
scf.for %arg5 = %c0 to %8 step %c3 {
scf.for %arg6 = %c0 to %7 step %c4 {
- %11 = std.subview %3[%arg4, %arg6][%c2, %c4][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
- %14 = std.subview %4[%arg6, %arg5][%c4, %c3][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
- %17 = std.subview %5[%arg4, %arg5][%c2, %c3][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %11 = memref.subview %3[%arg4, %arg6][%c2, %c4][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %14 = memref.subview %4[%arg6, %arg5][%c4, %c3][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %17 = memref.subview %5[%arg4, %arg5][%c2, %c3][1, 1] : memref<?x?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
linalg.matmul
ins(%11, %14: memref<?x?xf32, offset: ?, strides: [?, 1]>,
memref<?x?xf32, offset: ?, strides: [?, 1]>)
@@ -40,27 +40,27 @@ func @matmul_f32(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECK: %[[vA:.*]] = subview {{.*}} : memref<?x?xf32>
-// CHECK: %[[vB:.*]] = subview {{.*}} : memref<?x?xf32>
-// CHECK: %[[vC:.*]] = subview {{.*}} : memref<?x?xf32>
+// CHECK: %[[vA:.*]] = memref.subview {{.*}} : memref<?x?xf32>
+// CHECK: %[[vB:.*]] = memref.subview {{.*}} : memref<?x?xf32>
+// CHECK: %[[vC:.*]] = memref.subview {{.*}} : memref<?x?xf32>
///
-// CHECK: %[[tmpA:.*]] = alloc() : memref<32xi8>
-// ALLOCA: %[[tmpA:.*]] = alloca() : memref<32xi8>
-// CHECK: %[[fullA:.*]] = std.view %[[tmpA]][{{.*}}][{{.*}}] : memref<32xi8> to memref<?x?xf32>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[partialA:.*]] = subview %[[fullA]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[tmpA:.*]] = memref.alloc() : memref<32xi8>
+// ALLOCA: %[[tmpA:.*]] = memref.alloca() : memref<32xi8>
+// CHECK: %[[fullA:.*]] = memref.view %[[tmpA]][{{.*}}][{{.*}}] : memref<32xi8> to memref<?x?xf32>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[partialA:.*]] = memref.subview %[[fullA]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
///
-// CHECK: %[[tmpB:.*]] = alloc() : memref<48xi8>
-// ALLOCA: %[[tmpB:.*]] = alloca() : memref<48xi8>
-// CHECK: %[[fullB:.*]] = std.view %[[tmpB]][{{.*}}][{{.*}}] : memref<48xi8> to memref<?x?xf32>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[partialB:.*]] = subview %[[fullB]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[tmpB:.*]] = memref.alloc() : memref<48xi8>
+// ALLOCA: %[[tmpB:.*]] = memref.alloca() : memref<48xi8>
+// CHECK: %[[fullB:.*]] = memref.view %[[tmpB]][{{.*}}][{{.*}}] : memref<48xi8> to memref<?x?xf32>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[partialB:.*]] = memref.subview %[[fullB]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
///
-// CHECK: %[[tmpC:.*]] = alloc() : memref<24xi8>
-// ALLOCA: %[[tmpC:.*]] = alloca() : memref<24xi8>
-// CHECK: %[[fullC:.*]] = std.view %[[tmpC]][{{.*}}][{{.*}}] : memref<24xi8> to memref<?x?xf32>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[partialC:.*]] = subview %[[fullC]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
+// CHECK: %[[tmpC:.*]] = memref.alloc() : memref<24xi8>
+// ALLOCA: %[[tmpC:.*]] = memref.alloca() : memref<24xi8>
+// CHECK: %[[fullC:.*]] = memref.view %[[tmpC]][{{.*}}][{{.*}}] : memref<24xi8> to memref<?x?xf32>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[partialC:.*]] = memref.subview %[[fullC]]{{.*}} : memref<?x?xf32> to memref<?x?xf32, #[[$strided2D]]>
// CHECK: linalg.copy(%[[vA]], %[[partialA]]) : memref<?x?xf32, #[[$strided2D]]>, memref<?x?xf32, #[[$strided2D]]>
// CHECK: linalg.copy(%[[vB]], %[[partialB]]) : memref<?x?xf32, #[[$strided2D]]>, memref<?x?xf32, #[[$strided2D]]>
@@ -72,12 +72,12 @@ func @matmul_f32(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECK: memref<?x?xf32, #[[$strided2D]]>,
// CHECK: memref<?x?xf32, #[[$strided2D]]>
//
-// CHECK: dealloc %[[tmpA]] : memref<32xi8>
-// CHECK: dealloc %[[tmpB]] : memref<48xi8>
-// CHECK: dealloc %[[tmpC]] : memref<24xi8>
-// ALLOCA-NOT: dealloc %[[tmpA]] : memref<32xi8>
-// ALLOCA-NOT: dealloc %[[tmpB]] : memref<48xi8>
-// ALLOCA-NOT: dealloc %[[tmpC]] : memref<24xi8>
+// CHECK: memref.dealloc %[[tmpA]] : memref<32xi8>
+// CHECK: memref.dealloc %[[tmpB]] : memref<48xi8>
+// CHECK: memref.dealloc %[[tmpC]] : memref<24xi8>
+// ALLOCA-NOT: memref.dealloc %[[tmpA]] : memref<32xi8>
+// ALLOCA-NOT: memref.dealloc %[[tmpB]] : memref<48xi8>
+// ALLOCA-NOT: memref.dealloc %[[tmpC]] : memref<24xi8>
// -----
@@ -87,18 +87,18 @@ func @matmul_f64(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %3 = view %A[%c0][%M, %K] : memref<?xi8> to memref<?x?xf64>
- %4 = view %A[%c0][%K, %N] : memref<?xi8> to memref<?x?xf64>
- %5 = view %A[%c0][%M, %N] : memref<?xi8> to memref<?x?xf64>
- %6 = dim %3, %c0 : memref<?x?xf64>
- %7 = dim %3, %c1 : memref<?x?xf64>
- %8 = dim %4, %c1 : memref<?x?xf64>
+ %3 = memref.view %A[%c0][%M, %K] : memref<?xi8> to memref<?x?xf64>
+ %4 = memref.view %A[%c0][%K, %N] : memref<?xi8> to memref<?x?xf64>
+ %5 = memref.view %A[%c0][%M, %N] : memref<?xi8> to memref<?x?xf64>
+ %6 = memref.dim %3, %c0 : memref<?x?xf64>
+ %7 = memref.dim %3, %c1 : memref<?x?xf64>
+ %8 = memref.dim %4, %c1 : memref<?x?xf64>
scf.for %arg4 = %c0 to %6 step %c2 {
scf.for %arg5 = %c0 to %8 step %c3 {
scf.for %arg6 = %c0 to %7 step %c4 {
- %11 = std.subview %3[%arg4, %arg6][%c2, %c4][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
- %14 = std.subview %4[%arg6, %arg5][%c4, %c3][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
- %17 = std.subview %5[%arg4, %arg5][%c2, %c3][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
+ %11 = memref.subview %3[%arg4, %arg6][%c2, %c4][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
+ %14 = memref.subview %4[%arg6, %arg5][%c4, %c3][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
+ %17 = memref.subview %5[%arg4, %arg5][%c2, %c3][1, 1] : memref<?x?xf64> to memref<?x?xf64, offset: ?, strides: [?, 1]>
linalg.matmul
ins(%11, %14: memref<?x?xf64, offset: ?, strides: [?, 1]>,
memref<?x?xf64, offset: ?, strides: [?, 1]>)
@@ -113,24 +113,24 @@ func @matmul_f64(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECK: %[[vA_f64:.*]] = subview {{.*}} : memref<?x?xf64>
-// CHECK: %[[vB_f64:.*]] = subview {{.*}} : memref<?x?xf64>
-// CHECK: %[[vC_f64:.*]] = subview {{.*}} : memref<?x?xf64>
+// CHECK: %[[vA_f64:.*]] = memref.subview {{.*}} : memref<?x?xf64>
+// CHECK: %[[vB_f64:.*]] = memref.subview {{.*}} : memref<?x?xf64>
+// CHECK: %[[vC_f64:.*]] = memref.subview {{.*}} : memref<?x?xf64>
///
-// CHECK: %[[tmpA_f64:.*]] = alloc() : memref<64xi8>
-// CHECK: %[[fullA_f64:.*]] = std.view %[[tmpA_f64]][{{.*}}][{{.*}}] : memref<64xi8> to memref<?x?xf64>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
-// CHECK: %[[partialA_f64:.*]] = subview %[[fullA_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
+// CHECK: %[[tmpA_f64:.*]] = memref.alloc() : memref<64xi8>
+// CHECK: %[[fullA_f64:.*]] = memref.view %[[tmpA_f64]][{{.*}}][{{.*}}] : memref<64xi8> to memref<?x?xf64>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
+// CHECK: %[[partialA_f64:.*]] = memref.subview %[[fullA_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
///
-// CHECK: %[[tmpB_f64:.*]] = alloc() : memref<96xi8>
-// CHECK: %[[fullB_f64:.*]] = std.view %[[tmpB_f64]][{{.*}}][{{.*}}] : memref<96xi8> to memref<?x?xf64>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
-// CHECK: %[[partialB_f64:.*]] = subview %[[fullB_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
+// CHECK: %[[tmpB_f64:.*]] = memref.alloc() : memref<96xi8>
+// CHECK: %[[fullB_f64:.*]] = memref.view %[[tmpB_f64]][{{.*}}][{{.*}}] : memref<96xi8> to memref<?x?xf64>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
+// CHECK: %[[partialB_f64:.*]] = memref.subview %[[fullB_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
///
-// CHECK: %[[tmpC_f64:.*]] = alloc() : memref<48xi8>
-// CHECK: %[[fullC_f64:.*]] = std.view %[[tmpC_f64]][{{.*}}][{{.*}}] : memref<48xi8> to memref<?x?xf64>
-// DYNAMIC: std.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
-// CHECK: %[[partialC_f64:.*]] = subview %[[fullC_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
+// CHECK: %[[tmpC_f64:.*]] = memref.alloc() : memref<48xi8>
+// CHECK: %[[fullC_f64:.*]] = memref.view %[[tmpC_f64]][{{.*}}][{{.*}}] : memref<48xi8> to memref<?x?xf64>
+// DYNAMIC: memref.view %{{.*}}[{{.*}}][{{.*}}] : memref<?xi8> to memref<?x?xf64>
+// CHECK: %[[partialC_f64:.*]] = memref.subview %[[fullC_f64]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf64> to memref<?x?xf64, #[[$strided2D]]>
// CHECK: linalg.copy(%[[vA_f64]], %[[partialA_f64]]) : memref<?x?xf64, #[[$strided2D]]>, memref<?x?xf64, #[[$strided2D]]>
// CHECK: linalg.copy(%[[vB_f64]], %[[partialB_f64]]) : memref<?x?xf64, #[[$strided2D]]>, memref<?x?xf64, #[[$strided2D]]>
@@ -142,6 +142,6 @@ func @matmul_f64(%A: memref<?xi8>, %M: index, %N: index, %K: index) {
// CHECK: memref<?x?xf64, #[[$strided2D]]>,
// CHECK: memref<?x?xf64, #[[$strided2D]]>
//
-// CHECK: dealloc %[[tmpA_f64]] : memref<64xi8>
-// CHECK: dealloc %[[tmpB_f64]] : memref<96xi8>
-// CHECK: dealloc %[[tmpC_f64]] : memref<48xi8>
+// CHECK: memref.dealloc %[[tmpA_f64]] : memref<64xi8>
+// CHECK: memref.dealloc %[[tmpB_f64]] : memref<96xi8>
+// CHECK: memref.dealloc %[[tmpC_f64]] : memref<48xi8>
diff --git a/mlir/test/Dialect/Linalg/promotion_options.mlir b/mlir/test/Dialect/Linalg/promotion_options.mlir
index 0f38c904eb5a..397d76e4761a 100644
--- a/mlir/test/Dialect/Linalg/promotion_options.mlir
+++ b/mlir/test/Dialect/Linalg/promotion_options.mlir
@@ -16,13 +16,13 @@ func @gemm(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: scf.for
// CHECK: scf.for
// CHECK: scf.for
-// CHECK: %[[T7:.+]] = subview %[[ARG0]]
-// CHECK: %[[T12:.+]] = subview %[[ARG1]]
-// CHECK: %[[T17:.+]] = subview %[[ARG2]]
-// CHECK: %[[T18:.+]] = alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 3>
-// CHECK: %[[T19:.+]] = subview %[[T18]]
-// CHECK: %[[T20:.+]] = alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 3>
-// CHECK: %[[T21:.+]] = subview %[[T20]]
+// CHECK: %[[T7:.+]] = memref.subview %[[ARG0]]
+// CHECK: %[[T12:.+]] = memref.subview %[[ARG1]]
+// CHECK: %[[T17:.+]] = memref.subview %[[ARG2]]
+// CHECK: %[[T18:.+]] = memref.alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 3>
+// CHECK: %[[T19:.+]] = memref.subview %[[T18]]
+// CHECK: %[[T20:.+]] = memref.alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 3>
+// CHECK: %[[T21:.+]] = memref.subview %[[T20]]
// CHECK: linalg.fill(%[[T19]], %[[C42]])
// CHECK: linalg.copy(%[[T7]], %[[T19]])
// CHECK: linalg.fill(%[[T21]], %[[C42]])
@@ -30,5 +30,5 @@ func @gemm(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: linalg.matmul ins(%[[T19]], %[[T12]]{{.*}} outs(%[[T21]]
// CHECK-NOT: linalg.fill
// CHECK: linalg.copy(%[[T21]], %[[T17]])
-// CHECK: dealloc %[[T18]]
-// CHECK: dealloc %[[T20]]
+// CHECK: memref.dealloc %[[T18]]
+// CHECK: memref.dealloc %[[T20]]
diff --git a/mlir/test/Dialect/Linalg/reshape_fusion.mlir b/mlir/test/Dialect/Linalg/reshape_fusion.mlir
index 50269e36751b..fbaf47c9ac4d 100644
--- a/mlir/test/Dialect/Linalg/reshape_fusion.mlir
+++ b/mlir/test/Dialect/Linalg/reshape_fusion.mlir
@@ -512,7 +512,7 @@ func @unit_dim_reshape_expansion_full
affine_map<(d0, d1, d2, d3, d4, d5) -> (d3, d4)>,
affine_map<(d0, d1, d2, d3, d4, d5) -> (d5)>]
: tensor<1x?x1x2x1x4xf32> into tensor<?x2x4xf32>
- %1 = dim %arg0, %c1 : tensor<1x?x1x2x1x4xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<1x?x1x2x1x4xf32>
%2 = linalg.init_tensor [%1, 2, 4] : tensor<?x2x4xf32>
%3 = linalg.generic
{indexing_maps = [affine_map<(d0, d1, d2) -> (d0, d1, d2)>,
diff --git a/mlir/test/Dialect/Linalg/roundtrip.mlir b/mlir/test/Dialect/Linalg/roundtrip.mlir
index d034ff48c09f..ab2547952cfb 100644
--- a/mlir/test/Dialect/Linalg/roundtrip.mlir
+++ b/mlir/test/Dialect/Linalg/roundtrip.mlir
@@ -110,22 +110,22 @@ func @range(%arg0: index, %arg1: index, %arg2: index) {
func @views(%arg0: index, %arg1: index, %arg2: index, %arg3: index, %arg4: index) {
%c0 = constant 0 : index
%0 = muli %arg0, %arg0 : index
- %1 = alloc (%0) : memref<?xi8>
+ %1 = memref.alloc (%0) : memref<?xi8>
%2 = linalg.range %arg0:%arg1:%arg2 : !linalg.range
- %3 = view %1[%c0][%arg0, %arg0] : memref<?xi8> to memref<?x?xf32>
- %4 = view %1[%c0][%arg0, %arg0] : memref<?xi8> to memref<?x?xvector<4x4xf32>>
- dealloc %1 : memref<?xi8>
+ %3 = memref.view %1[%c0][%arg0, %arg0] : memref<?xi8> to memref<?x?xf32>
+ %4 = memref.view %1[%c0][%arg0, %arg0] : memref<?xi8> to memref<?x?xvector<4x4xf32>>
+ memref.dealloc %1 : memref<?xi8>
return
}
// CHECK-LABEL: func @views
// CHECK: muli %{{.*}}, %{{.*}} : index
-// CHECK-NEXT: alloc(%{{.*}}) : memref<?xi8>
+// CHECK-NEXT: memref.alloc(%{{.*}}) : memref<?xi8>
// CHECK-NEXT: range
-// CHECK-NEXT: std.view %{{.*}}[%{{.*}}][%{{.*}}] :
+// CHECK-NEXT: memref.view %{{.*}}[%{{.*}}][%{{.*}}] :
// CHECK-SAME: memref<?xi8> to memref<?x?xf32>
-// CHECK-NEXT: view %{{.*}}[%{{.*}}][%{{.*}}] :
+// CHECK-NEXT: memref.view %{{.*}}[%{{.*}}][%{{.*}}] :
// CHECK-SAME: memref<?xi8> to memref<?x?xvector<4x4xf32>>
-// CHECK-NEXT: dealloc %{{.*}} : memref<?xi8>
+// CHECK-NEXT: memref.dealloc %{{.*}} : memref<?xi8>
// -----
@@ -172,11 +172,11 @@ func @fill_view(%arg0: memref<?xf32, offset: ?, strides: [1]>, %arg1: f32) {
// -----
func @transpose(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>) {
- %0 = transpose %arg0 (i, j, k) -> (k, j, i) : memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]> to memref<?x?x?xf32, affine_map<(d0, d1, d2)[s0, s1, s2] -> (d2 * s1 + s0 + d1 * s2 + d0)>>
+ %0 = memref.transpose %arg0 (i, j, k) -> (k, j, i) : memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]> to memref<?x?x?xf32, affine_map<(d0, d1, d2)[s0, s1, s2] -> (d2 * s1 + s0 + d1 * s2 + d0)>>
return
}
// CHECK-LABEL: func @transpose
-// CHECK: transpose %{{.*}} ([[i:.*]], [[j:.*]], [[k:.*]]) -> ([[k]], [[j]], [[i]]) :
+// CHECK: memref.transpose %{{.*}} ([[i:.*]], [[j:.*]], [[k:.*]]) -> ([[k]], [[j]], [[i]]) :
// CHECK-SAME: memref<?x?x?xf32, #[[$strided3D]]> to memref<?x?x?xf32, #[[$strided3DT]]>
// -----
@@ -847,9 +847,9 @@ func @tiled_loop_reduction(%input_3d: tensor<16x24x32xf32>,
%c2 = constant 2 : index
%c4 = constant 4 : index
%c8 = constant 8 : index
- %X = dim %input_3d, %c0 : tensor<16x24x32xf32>
- %Y = dim %input_3d, %c1 : tensor<16x24x32xf32>
- %Z = dim %input_3d, %c2 : tensor<16x24x32xf32>
+ %X = memref.dim %input_3d, %c0 : tensor<16x24x32xf32>
+ %Y = memref.dim %input_3d, %c1 : tensor<16x24x32xf32>
+ %Z = memref.dim %input_3d, %c2 : tensor<16x24x32xf32>
%result = linalg.tiled_loop (%i, %j, %k)
= (%c0, %c0, %c0) to (%X, %Y, %Z) step (%c2, %c4, %c8)
ins(%input_3d, %input_2d: tensor<16x24x32xf32>, tensor<16x32xf32>)
diff --git a/mlir/test/Dialect/Linalg/sparse_1d.mlir b/mlir/test/Dialect/Linalg/sparse_1d.mlir
index bd0d7ee859d8..81e62eb515d6 100644
--- a/mlir/test/Dialect/Linalg/sparse_1d.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_1d.mlir
@@ -21,16 +21,16 @@
// CHECK: %[[VAL_3:.*]] = constant 32 : index
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant 1 : index
-// CHECK: %[[VAL_6:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_8:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_6:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_7]], %[[VAL_8]]) : memref<32xf32>, memref<32xf32>
// CHECK: scf.for %[[VAL_9:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<32xf32>
// CHECK: %[[VAL_11:.*]] = addf %[[VAL_10]], %[[VAL_1]] : f32
-// CHECK: store %[[VAL_11]], %[[VAL_8]]{{\[}}%[[VAL_9]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_11]], %[[VAL_8]]{{\[}}%[[VAL_9]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_12:.*]] = tensor_load %[[VAL_8]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.tensor_load %[[VAL_8]] : memref<32xf32>
// CHECK: return %[[VAL_12]] : tensor<32xf32>
// CHECK: }
func @add_d(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -51,16 +51,16 @@ func @add_d(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<
// CHECK: %[[VAL_3:.*]] = constant 32 : index
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant 1 : index
-// CHECK: %[[VAL_6:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_8:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_6:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_7]], %[[VAL_8]]) : memref<32xf32>, memref<32xf32>
// CHECK: scf.for %[[VAL_9:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<32xf32>
// CHECK: %[[VAL_11:.*]] = mulf %[[VAL_10]], %[[VAL_1]] : f32
-// CHECK: store %[[VAL_11]], %[[VAL_8]]{{\[}}%[[VAL_9]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_11]], %[[VAL_8]]{{\[}}%[[VAL_9]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_12:.*]] = tensor_load %[[VAL_8]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.tensor_load %[[VAL_8]] : memref<32xf32>
// CHECK: return %[[VAL_12]] : tensor<32xf32>
// CHECK: }
func @mul_d(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -98,25 +98,25 @@ func @mul_d(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
// CHECK: %[[VAL_14:.*]]:2 = scf.while (%[[VAL_15:.*]] = %[[VAL_12]], %[[VAL_16:.*]] = %[[VAL_4]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_17:.*]] = cmpi ult, %[[VAL_15]], %[[VAL_13]] : index
// CHECK: scf.condition(%[[VAL_17]]) %[[VAL_15]], %[[VAL_16]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_18:.*]]: index, %[[VAL_19:.*]]: index):
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]] = cmpi eq, %[[VAL_20]], %[[VAL_19]] : index
// CHECK: scf.if %[[VAL_21]] {
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_18]]] : memref<?xf32>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_18]]] : memref<?xf32>
// CHECK: %[[VAL_23:.*]] = addf %[[VAL_22]], %[[VAL_1]] : f32
-// CHECK: store %[[VAL_23]], %[[VAL_11]]{{\[}}%[[VAL_19]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_23]], %[[VAL_11]]{{\[}}%[[VAL_19]]] : memref<32xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
-// CHECK: store %[[VAL_1]], %[[VAL_11]]{{\[}}%[[VAL_19]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_1]], %[[VAL_11]]{{\[}}%[[VAL_19]]] : memref<32xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -127,9 +127,9 @@ func @mul_d(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<
// CHECK: scf.yield %[[VAL_26]], %[[VAL_27]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_28:.*]] = %[[VAL_29:.*]]#1 to %[[VAL_3]] step %[[VAL_6]] {
-// CHECK: store %[[VAL_1]], %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_1]], %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_30:.*]] = tensor_load %[[VAL_11]] : memref<32xf32>
+// CHECK: %[[VAL_30:.*]] = memref.tensor_load %[[VAL_11]] : memref<32xf32>
// CHECK: return %[[VAL_30]] : tensor<32xf32>
// CHECK: }
func @add_s(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -151,23 +151,23 @@ func @add_s(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<
// CHECK: %[[VAL_4:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_2]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_5:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_2]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_8:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_7]], %[[VAL_8]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_9:.*]] = load %[[VAL_4]]{{\[}}%[[VAL_2]]] : memref<?xindex>
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_4]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_9:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_2]]] : memref<?xindex>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_3]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_11:.*]] = %[[VAL_9]] to %[[VAL_10]] step %[[VAL_3]] {
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_11]]] : memref<?xindex>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_11]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
// CHECK: %[[VAL_15:.*]] = addf %[[VAL_13]], %[[VAL_14]] : f32
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xf32>
// CHECK: %[[VAL_18:.*]] = addf %[[VAL_16]], %[[VAL_17]] : f32
// CHECK: %[[VAL_19:.*]] = addf %[[VAL_15]], %[[VAL_18]] : f32
-// CHECK: store %[[VAL_19]], %[[VAL_8]]{{\[}}%[[VAL_12]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_19]], %[[VAL_8]]{{\[}}%[[VAL_12]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_20:.*]] = tensor_load %[[VAL_8]] : memref<32xf32>
+// CHECK: %[[VAL_20:.*]] = memref.tensor_load %[[VAL_8]] : memref<32xf32>
// CHECK: return %[[VAL_20]] : tensor<32xf32>
// CHECK: }
func @repeated_add_s(%arga: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -192,18 +192,18 @@ func @repeated_add_s(%arga: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32x
// CHECK: %[[VAL_5:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_9:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_9:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_8]], %[[VAL_9]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_10]] to %[[VAL_11]] step %[[VAL_4]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_12]]] : memref<?xf32>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_12]]] : memref<?xf32>
// CHECK: %[[VAL_15:.*]] = mulf %[[VAL_14]], %[[VAL_1]] : f32
-// CHECK: store %[[VAL_15]], %[[VAL_9]]{{\[}}%[[VAL_13]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_15]], %[[VAL_9]]{{\[}}%[[VAL_13]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_16:.*]] = tensor_load %[[VAL_9]] : memref<32xf32>
+// CHECK: %[[VAL_16:.*]] = memref.tensor_load %[[VAL_9]] : memref<32xf32>
// CHECK: return %[[VAL_16]] : tensor<32xf32>
// CHECK: }
func @mul_s(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -239,18 +239,18 @@ func @mul_s(%arga: tensor<32xf32>, %argb: f32, %argx: tensor<32xf32>) -> tensor<
// CHECK: %[[VAL_3:.*]] = constant 32 : index
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant 1 : index
-// CHECK: %[[VAL_6:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_9:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_6:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_9:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_8]], %[[VAL_9]]) : memref<32xf32>, memref<32xf32>
// CHECK: scf.for %[[VAL_10:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_10]]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_10]]] : memref<32xf32>
// CHECK: %[[VAL_13:.*]] = addf %[[VAL_11]], %[[VAL_12]] : f32
-// CHECK: store %[[VAL_13]], %[[VAL_9]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_13]], %[[VAL_9]]{{\[}}%[[VAL_10]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_14:.*]] = tensor_load %[[VAL_9]] : memref<32xf32>
+// CHECK: %[[VAL_14:.*]] = memref.tensor_load %[[VAL_9]] : memref<32xf32>
// CHECK: return %[[VAL_14]] : tensor<32xf32>
// CHECK: }
func @add_dd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -271,18 +271,18 @@ func @add_dd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_3:.*]] = constant 32 : index
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant 1 : index
-// CHECK: %[[VAL_6:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_9:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_6:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_9:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_8]], %[[VAL_9]]) : memref<32xf32>, memref<32xf32>
// CHECK: scf.for %[[VAL_10:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_10]]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_10]]] : memref<32xf32>
// CHECK: %[[VAL_13:.*]] = mulf %[[VAL_11]], %[[VAL_12]] : f32
-// CHECK: store %[[VAL_13]], %[[VAL_9]]{{\[}}%[[VAL_10]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_13]], %[[VAL_9]]{{\[}}%[[VAL_10]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_14:.*]] = tensor_load %[[VAL_9]] : memref<32xf32>
+// CHECK: %[[VAL_14:.*]] = memref.tensor_load %[[VAL_9]] : memref<32xf32>
// CHECK: return %[[VAL_14]] : tensor<32xf32>
// CHECK: }
func @mul_dd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -319,31 +319,31 @@ func @mul_dd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant true
// CHECK: %[[VAL_6:.*]] = constant 1 : index
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
// CHECK: %[[VAL_15:.*]]:2 = scf.while (%[[VAL_16:.*]] = %[[VAL_13]], %[[VAL_17:.*]] = %[[VAL_4]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_18:.*]] = cmpi ult, %[[VAL_16]], %[[VAL_14]] : index
// CHECK: scf.condition(%[[VAL_18]]) %[[VAL_16]], %[[VAL_17]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_19:.*]]: index, %[[VAL_20:.*]]: index):
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]] = cmpi eq, %[[VAL_21]], %[[VAL_20]] : index
// CHECK: scf.if %[[VAL_22]] {
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<32xf32>
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_19]]] : memref<?xf32>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_19]]] : memref<?xf32>
// CHECK: %[[VAL_25:.*]] = addf %[[VAL_23]], %[[VAL_24]] : f32
-// CHECK: store %[[VAL_25]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_25]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<32xf32>
-// CHECK: store %[[VAL_26]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_26]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -354,10 +354,10 @@ func @mul_dd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.yield %[[VAL_29]], %[[VAL_30]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_31:.*]] = %[[VAL_32:.*]]#1 to %[[VAL_3]] step %[[VAL_6]] {
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_31]]] : memref<32xf32>
-// CHECK: store %[[VAL_33]], %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<32xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_31]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_33]], %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_34:.*]] = tensor_load %[[VAL_12]] : memref<32xf32>
+// CHECK: %[[VAL_34:.*]] = memref.tensor_load %[[VAL_12]] : memref<32xf32>
// CHECK: return %[[VAL_34]] : tensor<32xf32>
// CHECK: }
func @add_ds(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -377,23 +377,23 @@ func @add_ds(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK-SAME: %[[VAL_2:.*2]]: tensor<32xf32>) -> tensor<32xf32> {
// CHECK: %[[VAL_3:.*]] = constant 0 : index
// CHECK: %[[VAL_4:.*]] = constant 1 : index
-// CHECK: %[[VAL_5:.*]] = tensor_to_memref %[[VAL_0]] : memref<32xf32>
+// CHECK: %[[VAL_5:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32xf32>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_10:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_9]], %[[VAL_10]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_11]] to %[[VAL_12]] step %[[VAL_4]] {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_14]]] : memref<32xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_13]]] : memref<?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_14]]] : memref<32xf32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_13]]] : memref<?xf32>
// CHECK: %[[VAL_17:.*]] = mulf %[[VAL_15]], %[[VAL_16]] : f32
-// CHECK: store %[[VAL_17]], %[[VAL_10]]{{\[}}%[[VAL_14]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_17]], %[[VAL_10]]{{\[}}%[[VAL_14]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_18:.*]] = tensor_load %[[VAL_10]] : memref<32xf32>
+// CHECK: %[[VAL_18:.*]] = memref.tensor_load %[[VAL_10]] : memref<32xf32>
// CHECK: return %[[VAL_18]] : tensor<32xf32>
// CHECK: }
func @mul_ds(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -433,28 +433,28 @@ func @mul_ds(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
// CHECK: %[[VAL_15:.*]]:2 = scf.while (%[[VAL_16:.*]] = %[[VAL_13]], %[[VAL_17:.*]] = %[[VAL_4]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_18:.*]] = cmpi ult, %[[VAL_16]], %[[VAL_14]] : index
// CHECK: scf.condition(%[[VAL_18]]) %[[VAL_16]], %[[VAL_17]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_19:.*]]: index, %[[VAL_20:.*]]: index):
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]] = cmpi eq, %[[VAL_21]], %[[VAL_20]] : index
// CHECK: scf.if %[[VAL_22]] {
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xf32>
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xf32>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: %[[VAL_25:.*]] = addf %[[VAL_23]], %[[VAL_24]] : f32
-// CHECK: store %[[VAL_25]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_25]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<32xf32>
-// CHECK: store %[[VAL_26]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_26]], %[[VAL_12]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -465,10 +465,10 @@ func @mul_ds(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.yield %[[VAL_29]], %[[VAL_30]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_31:.*]] = %[[VAL_32:.*]]#1 to %[[VAL_3]] step %[[VAL_6]] {
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_31]]] : memref<32xf32>
-// CHECK: store %[[VAL_33]], %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<32xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_31]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_33]], %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_34:.*]] = tensor_load %[[VAL_12]] : memref<32xf32>
+// CHECK: %[[VAL_34:.*]] = memref.tensor_load %[[VAL_12]] : memref<32xf32>
// CHECK: return %[[VAL_34]] : tensor<32xf32>
// CHECK: }
func @add_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -491,20 +491,20 @@ func @add_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_5:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_10:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_9]], %[[VAL_10]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_11]] to %[[VAL_12]] step %[[VAL_4]] {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_14]]] : memref<32xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xf32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_14]]] : memref<32xf32>
// CHECK: %[[VAL_17:.*]] = mulf %[[VAL_15]], %[[VAL_16]] : f32
-// CHECK: store %[[VAL_17]], %[[VAL_10]]{{\[}}%[[VAL_14]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_17]], %[[VAL_10]]{{\[}}%[[VAL_14]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_18:.*]] = tensor_load %[[VAL_10]] : memref<32xf32>
+// CHECK: %[[VAL_18:.*]] = memref.tensor_load %[[VAL_10]] : memref<32xf32>
// CHECK: return %[[VAL_18]] : tensor<32xf32>
// CHECK: }
func @mul_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -545,13 +545,13 @@ func @mul_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]]:2 = scf.while (%[[VAL_18:.*]] = %[[VAL_13]], %[[VAL_19:.*]] = %[[VAL_15]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_20:.*]] = cmpi ult, %[[VAL_18]], %[[VAL_14]] : index
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_16]] : index
@@ -559,28 +559,28 @@ func @mul_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.condition(%[[VAL_22]]) %[[VAL_18]], %[[VAL_19]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_23:.*]]: index, %[[VAL_24:.*]]: index):
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
// CHECK: %[[VAL_27:.*]] = cmpi ult, %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_28:.*]] = select %[[VAL_27]], %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_29:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_28]] : index
// CHECK: %[[VAL_31:.*]] = and %[[VAL_29]], %[[VAL_30]] : i1
// CHECK: scf.if %[[VAL_31]] {
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_34:.*]] = addf %[[VAL_32]], %[[VAL_33]] : f32
-// CHECK: store %[[VAL_34]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_34]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
// CHECK: } else {
// CHECK: %[[VAL_35:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
-// CHECK: store %[[VAL_36]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: memref.store %[[VAL_36]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
// CHECK: } else {
// CHECK: %[[VAL_37:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_28]] : index
// CHECK: scf.if %[[VAL_37]] {
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
-// CHECK: store %[[VAL_38]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: memref.store %[[VAL_38]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -594,16 +594,16 @@ func @mul_sd(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.yield %[[VAL_41]], %[[VAL_44]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_45:.*]] = %[[VAL_46:.*]]#0 to %[[VAL_14]] step %[[VAL_4]] {
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_45]]] : memref<?xindex>
-// CHECK: %[[VAL_48:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_45]]] : memref<?xf32>
-// CHECK: store %[[VAL_48]], %[[VAL_12]]{{\[}}%[[VAL_47]]] : memref<32xf32>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_45]]] : memref<?xindex>
+// CHECK: %[[VAL_48:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_45]]] : memref<?xf32>
+// CHECK: memref.store %[[VAL_48]], %[[VAL_12]]{{\[}}%[[VAL_47]]] : memref<32xf32>
// CHECK: }
// CHECK: scf.for %[[VAL_49:.*]] = %[[VAL_50:.*]]#1 to %[[VAL_16]] step %[[VAL_4]] {
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_49]]] : memref<?xindex>
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_49]]] : memref<?xf32>
-// CHECK: store %[[VAL_52]], %[[VAL_12]]{{\[}}%[[VAL_51]]] : memref<32xf32>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: memref.store %[[VAL_52]], %[[VAL_12]]{{\[}}%[[VAL_51]]] : memref<32xf32>
// CHECK: }
-// CHECK: %[[VAL_53:.*]] = tensor_load %[[VAL_12]] : memref<32xf32>
+// CHECK: %[[VAL_53:.*]] = memref.tensor_load %[[VAL_12]] : memref<32xf32>
// CHECK: return %[[VAL_53]] : tensor<32xf32>
// CHECK: }
func @add_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -629,13 +629,13 @@ func @add_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_3]] : tensor<32xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32xf32>, memref<32xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]]:2 = scf.while (%[[VAL_18:.*]] = %[[VAL_13]], %[[VAL_19:.*]] = %[[VAL_15]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_20:.*]] = cmpi ult, %[[VAL_18]], %[[VAL_14]] : index
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_16]] : index
@@ -643,18 +643,18 @@ func @add_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.condition(%[[VAL_22]]) %[[VAL_18]], %[[VAL_19]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_23:.*]]: index, %[[VAL_24:.*]]: index):
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
// CHECK: %[[VAL_27:.*]] = cmpi ult, %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_28:.*]] = select %[[VAL_27]], %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_29:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_28]] : index
// CHECK: %[[VAL_31:.*]] = and %[[VAL_29]], %[[VAL_30]] : i1
// CHECK: scf.if %[[VAL_31]] {
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_34:.*]] = mulf %[[VAL_32]], %[[VAL_33]] : f32
-// CHECK: store %[[VAL_34]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
+// CHECK: memref.store %[[VAL_34]], %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<32xf32>
// CHECK: } else {
// CHECK: }
// CHECK: %[[VAL_35:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
@@ -665,7 +665,7 @@ func @add_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_40:.*]] = select %[[VAL_38]], %[[VAL_39]], %[[VAL_24]] : index
// CHECK: scf.yield %[[VAL_37]], %[[VAL_40]] : index, index
// CHECK: }
-// CHECK: %[[VAL_41:.*]] = tensor_load %[[VAL_12]] : memref<32xf32>
+// CHECK: %[[VAL_41:.*]] = memref.tensor_load %[[VAL_12]] : memref<32xf32>
// CHECK: return %[[VAL_41]] : tensor<32xf32>
// CHECK: }
func @mul_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>) -> tensor<32xf32> {
@@ -707,13 +707,13 @@ func @mul_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<16xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_3]] : memref<16xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<16xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_3]] : memref<16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<16xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<16xf32>, memref<16xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: %[[VAL_18:.*]]:2 = scf.while (%[[VAL_19:.*]] = %[[VAL_14]], %[[VAL_20:.*]] = %[[VAL_16]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_15]] : index
// CHECK: %[[VAL_22:.*]] = cmpi ult, %[[VAL_20]], %[[VAL_17]] : index
@@ -721,32 +721,32 @@ func @mul_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.condition(%[[VAL_23]]) %[[VAL_19]], %[[VAL_20]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_24:.*]]: index, %[[VAL_25:.*]]: index):
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_24]]] : memref<?xindex>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
// CHECK: %[[VAL_28:.*]] = cmpi ult, %[[VAL_27]], %[[VAL_26]] : index
// CHECK: %[[VAL_29:.*]] = select %[[VAL_28]], %[[VAL_27]], %[[VAL_26]] : index
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_29]] : index
// CHECK: %[[VAL_31:.*]] = cmpi eq, %[[VAL_27]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = and %[[VAL_30]], %[[VAL_31]] : i1
// CHECK: scf.if %[[VAL_32]] {
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_34:.*]] = mulf %[[VAL_33]], %[[VAL_2]] : f32
-// CHECK: %[[VAL_35:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_35:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
// CHECK: %[[VAL_36:.*]] = mulf %[[VAL_35]], %[[VAL_2]] : f32
// CHECK: %[[VAL_37:.*]] = addf %[[VAL_34]], %[[VAL_36]] : f32
-// CHECK: store %[[VAL_37]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_37]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: %[[VAL_38:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_29]] : index
// CHECK: scf.if %[[VAL_38]] {
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_40:.*]] = mulf %[[VAL_39]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_40]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_40]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: %[[VAL_41:.*]] = cmpi eq, %[[VAL_27]], %[[VAL_29]] : index
// CHECK: scf.if %[[VAL_41]] {
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
// CHECK: %[[VAL_43:.*]] = mulf %[[VAL_42]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_43]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_43]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -760,18 +760,18 @@ func @mul_ss(%arga: tensor<32xf32>, %argb: tensor<32xf32>, %argx: tensor<32xf32>
// CHECK: scf.yield %[[VAL_46]], %[[VAL_49]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_50:.*]] = %[[VAL_51:.*]]#0 to %[[VAL_15]] step %[[VAL_5]] {
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_50]]] : memref<?xindex>
-// CHECK: %[[VAL_53:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_50]]] : memref<?xf32>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_50]]] : memref<?xindex>
+// CHECK: %[[VAL_53:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_50]]] : memref<?xf32>
// CHECK: %[[VAL_54:.*]] = mulf %[[VAL_53]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_54]], %[[VAL_13]]{{\[}}%[[VAL_52]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_54]], %[[VAL_13]]{{\[}}%[[VAL_52]]] : memref<16xf32>
// CHECK: }
// CHECK: scf.for %[[VAL_55:.*]] = %[[VAL_56:.*]]#1 to %[[VAL_17]] step %[[VAL_5]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_55]]] : memref<?xindex>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_55]]] : memref<?xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_55]]] : memref<?xindex>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_55]]] : memref<?xf32>
// CHECK: %[[VAL_59:.*]] = mulf %[[VAL_58]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_59]], %[[VAL_13]]{{\[}}%[[VAL_57]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_59]], %[[VAL_13]]{{\[}}%[[VAL_57]]] : memref<16xf32>
// CHECK: }
-// CHECK: %[[VAL_60:.*]] = tensor_load %[[VAL_13]] : memref<16xf32>
+// CHECK: %[[VAL_60:.*]] = memref.tensor_load %[[VAL_13]] : memref<16xf32>
// CHECK: return %[[VAL_60]] : tensor<16xf32>
// CHECK: }
func @two_way_inv(%arga: tensor<16xf32>, %argb: tensor<16xf32>, %argc: f32, %argx: tensor<16xf32>) -> tensor<16xf32> {
@@ -800,13 +800,13 @@ func @two_way_inv(%arga: tensor<16xf32>, %argb: tensor<16xf32>, %argc: f32, %arg
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<16xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_3]] : memref<16xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<16xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_3]] : memref<16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<16xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<16xf32>, memref<16xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: %[[VAL_18:.*]]:2 = scf.while (%[[VAL_19:.*]] = %[[VAL_14]], %[[VAL_20:.*]] = %[[VAL_16]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_15]] : index
// CHECK: %[[VAL_22:.*]] = cmpi ult, %[[VAL_20]], %[[VAL_17]] : index
@@ -814,31 +814,31 @@ func @two_way_inv(%arga: tensor<16xf32>, %argb: tensor<16xf32>, %argc: f32, %arg
// CHECK: scf.condition(%[[VAL_23]]) %[[VAL_19]], %[[VAL_20]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_24:.*]]: index, %[[VAL_25:.*]]: index):
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_24]]] : memref<?xindex>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
// CHECK: %[[VAL_28:.*]] = cmpi ult, %[[VAL_27]], %[[VAL_26]] : index
// CHECK: %[[VAL_29:.*]] = select %[[VAL_28]], %[[VAL_27]], %[[VAL_26]] : index
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_29]] : index
// CHECK: %[[VAL_31:.*]] = cmpi eq, %[[VAL_27]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = and %[[VAL_30]], %[[VAL_31]] : i1
// CHECK: scf.if %[[VAL_32]] {
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
// CHECK: %[[VAL_35:.*]] = addf %[[VAL_33]], %[[VAL_34]] : f32
// CHECK: %[[VAL_36:.*]] = mulf %[[VAL_35]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_36]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_36]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: %[[VAL_37:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_29]] : index
// CHECK: scf.if %[[VAL_37]] {
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_39:.*]] = mulf %[[VAL_38]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_39]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_39]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: %[[VAL_40:.*]] = cmpi eq, %[[VAL_27]], %[[VAL_29]] : index
// CHECK: scf.if %[[VAL_40]] {
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
// CHECK: %[[VAL_42:.*]] = mulf %[[VAL_41]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_42]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_42]], %[[VAL_13]]{{\[}}%[[VAL_29]]] : memref<16xf32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -852,18 +852,18 @@ func @two_way_inv(%arga: tensor<16xf32>, %argb: tensor<16xf32>, %argc: f32, %arg
// CHECK: scf.yield %[[VAL_45]], %[[VAL_48]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_49:.*]] = %[[VAL_50:.*]]#0 to %[[VAL_15]] step %[[VAL_5]] {
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_49]]] : memref<?xindex>
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: %[[VAL_53:.*]] = mulf %[[VAL_52]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_53]], %[[VAL_13]]{{\[}}%[[VAL_51]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_53]], %[[VAL_13]]{{\[}}%[[VAL_51]]] : memref<16xf32>
// CHECK: }
// CHECK: scf.for %[[VAL_54:.*]] = %[[VAL_55:.*]]#1 to %[[VAL_17]] step %[[VAL_5]] {
-// CHECK: %[[VAL_56:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_54]]] : memref<?xindex>
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_54]]] : memref<?xf32>
+// CHECK: %[[VAL_56:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_54]]] : memref<?xindex>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_54]]] : memref<?xf32>
// CHECK: %[[VAL_58:.*]] = mulf %[[VAL_57]], %[[VAL_2]] : f32
-// CHECK: store %[[VAL_58]], %[[VAL_13]]{{\[}}%[[VAL_56]]] : memref<16xf32>
+// CHECK: memref.store %[[VAL_58]], %[[VAL_13]]{{\[}}%[[VAL_56]]] : memref<16xf32>
// CHECK: }
-// CHECK: %[[VAL_59:.*]] = tensor_load %[[VAL_13]] : memref<16xf32>
+// CHECK: %[[VAL_59:.*]] = memref.tensor_load %[[VAL_13]] : memref<16xf32>
// CHECK: return %[[VAL_59]] : tensor<16xf32>
// CHECK: }
func @two_way_inv_alt(%arga: tensor<16xf32>,
@@ -900,19 +900,19 @@ func @two_way_inv_alt(%arga: tensor<16xf32>,
// CHECK: %[[VAL_3:.*]] = constant 1 : index
// CHECK: %[[VAL_4:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_2]] : tensor<?xf32> to memref<?xindex>
// CHECK: %[[VAL_5:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<?xf32> to memref<?xf32>
-// CHECK: %[[VAL_6:.*]] = tensor_to_memref %[[VAL_1]] : memref<f32>
-// CHECK: %[[VAL_7:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_6:.*]] = memref.buffer_cast %[[VAL_1]] : memref<f32>
+// CHECK: %[[VAL_7:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_6]], %[[VAL_7]]) : memref<f32>, memref<f32>
-// CHECK: %[[VAL_8:.*]] = load %[[VAL_4]]{{\[}}%[[VAL_2]]] : memref<?xindex>
-// CHECK: %[[VAL_9:.*]] = load %[[VAL_4]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_7]][] : memref<f32>
+// CHECK: %[[VAL_8:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_2]]] : memref<?xindex>
+// CHECK: %[[VAL_9:.*]] = memref.load %[[VAL_4]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_7]][] : memref<f32>
// CHECK: %[[VAL_11:.*]] = scf.for %[[VAL_12:.*]] = %[[VAL_8]] to %[[VAL_9]] step %[[VAL_3]] iter_args(%[[VAL_13:.*]] = %[[VAL_10]]) -> (f32) {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_12]]] : memref<?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_12]]] : memref<?xf32>
// CHECK: %[[VAL_15:.*]] = addf %[[VAL_13]], %[[VAL_14]] : f32
// CHECK: scf.yield %[[VAL_15]] : f32
// CHECK: }
-// CHECK: store %[[VAL_16:.*]], %[[VAL_7]][] : memref<f32>
-// CHECK: %[[VAL_17:.*]] = tensor_load %[[VAL_7]] : memref<f32>
+// CHECK: memref.store %[[VAL_16:.*]], %[[VAL_7]][] : memref<f32>
+// CHECK: %[[VAL_17:.*]] = memref.tensor_load %[[VAL_7]] : memref<f32>
// CHECK: return %[[VAL_17]] : tensor<f32>
// CHECK: }
func @sum_reduction(%arga: tensor<?xf32>, %argx: tensor<f32>) -> tensor<f32> {
@@ -953,13 +953,13 @@ func @sum_reduction(%arga: tensor<?xf32>, %argx: tensor<f32>) -> tensor<f32> {
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_3]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_3]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<16xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<f32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<f32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<f32>, memref<f32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]]:2 = scf.while (%[[VAL_18:.*]] = %[[VAL_13]], %[[VAL_19:.*]] = %[[VAL_15]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_20:.*]] = cmpi ult, %[[VAL_18]], %[[VAL_14]] : index
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_16]] : index
@@ -967,34 +967,34 @@ func @sum_reduction(%arga: tensor<?xf32>, %argx: tensor<f32>) -> tensor<f32> {
// CHECK: scf.condition(%[[VAL_22]]) %[[VAL_18]], %[[VAL_19]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_23:.*]]: index, %[[VAL_24:.*]]: index):
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xindex>
// CHECK: %[[VAL_27:.*]] = cmpi ult, %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_28:.*]] = select %[[VAL_27]], %[[VAL_26]], %[[VAL_25]] : index
// CHECK: %[[VAL_29:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_28]] : index
// CHECK: %[[VAL_31:.*]] = and %[[VAL_29]], %[[VAL_30]] : i1
// CHECK: scf.if %[[VAL_31]] {
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_12]][] : memref<f32>
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_35:.*]] = addf %[[VAL_33]], %[[VAL_34]] : f32
// CHECK: %[[VAL_36:.*]] = addf %[[VAL_32]], %[[VAL_35]] : f32
-// CHECK: store %[[VAL_36]], %[[VAL_12]][] : memref<f32>
+// CHECK: memref.store %[[VAL_36]], %[[VAL_12]][] : memref<f32>
// CHECK: } else {
// CHECK: %[[VAL_37:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_28]] : index
// CHECK: scf.if %[[VAL_37]] {
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_12]][] : memref<f32>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_23]]] : memref<?xf32>
// CHECK: %[[VAL_40:.*]] = addf %[[VAL_38]], %[[VAL_39]] : f32
-// CHECK: store %[[VAL_40]], %[[VAL_12]][] : memref<f32>
+// CHECK: memref.store %[[VAL_40]], %[[VAL_12]][] : memref<f32>
// CHECK: } else {
// CHECK: %[[VAL_41:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_28]] : index
// CHECK: scf.if %[[VAL_41]] {
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_12]][] : memref<f32>
-// CHECK: %[[VAL_43:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_43:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: %[[VAL_44:.*]] = addf %[[VAL_42]], %[[VAL_43]] : f32
-// CHECK: store %[[VAL_44]], %[[VAL_12]][] : memref<f32>
+// CHECK: memref.store %[[VAL_44]], %[[VAL_12]][] : memref<f32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1007,19 +1007,19 @@ func @sum_reduction(%arga: tensor<?xf32>, %argx: tensor<f32>) -> tensor<f32> {
// CHECK: %[[VAL_50:.*]] = select %[[VAL_48]], %[[VAL_49]], %[[VAL_24]] : index
// CHECK: scf.yield %[[VAL_47]], %[[VAL_50]] : index, index
// CHECK: }
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_12]][] : memref<f32>
// CHECK: %[[VAL_52:.*]] = scf.for %[[VAL_53:.*]] = %[[VAL_54:.*]]#0 to %[[VAL_14]] step %[[VAL_4]] iter_args(%[[VAL_55:.*]] = %[[VAL_51]]) -> (f32) {
-// CHECK: %[[VAL_56:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_53]]] : memref<?xf32>
+// CHECK: %[[VAL_56:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_53]]] : memref<?xf32>
// CHECK: %[[VAL_57:.*]] = addf %[[VAL_55]], %[[VAL_56]] : f32
// CHECK: scf.yield %[[VAL_57]] : f32
// CHECK: }
// CHECK: %[[VAL_58:.*]] = scf.for %[[VAL_59:.*]] = %[[VAL_60:.*]]#1 to %[[VAL_16]] step %[[VAL_4]] iter_args(%[[VAL_61:.*]] = %[[VAL_62:.*]]) -> (f32) {
-// CHECK: %[[VAL_63:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_59]]] : memref<?xf32>
+// CHECK: %[[VAL_63:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_59]]] : memref<?xf32>
// CHECK: %[[VAL_64:.*]] = addf %[[VAL_61]], %[[VAL_63]] : f32
// CHECK: scf.yield %[[VAL_64]] : f32
// CHECK: }
-// CHECK: store %[[VAL_65:.*]], %[[VAL_12]][] : memref<f32>
-// CHECK: %[[VAL_66:.*]] = tensor_load %[[VAL_12]] : memref<f32>
+// CHECK: memref.store %[[VAL_65:.*]], %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_66:.*]] = memref.tensor_load %[[VAL_12]] : memref<f32>
// CHECK: return %[[VAL_66]] : tensor<f32>
// CHECK: }
func @sum_reduction_ss(%arga: tensor<16xf32>,
@@ -1065,18 +1065,18 @@ func @sum_reduction_ss(%arga: tensor<16xf32>,
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<16xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<f32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<f32>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_pointers %[[VAL_2]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_indices %[[VAL_2]], %[[VAL_4]] : tensor<16xf32> to memref<?xindex>
// CHECK: %[[VAL_12:.*]] = linalg.sparse_values %[[VAL_2]] : tensor<16xf32> to memref<?xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_3]] : memref<f32>
-// CHECK: %[[VAL_14:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_3]] : memref<f32>
+// CHECK: %[[VAL_14:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_13]], %[[VAL_14]]) : memref<f32>, memref<f32>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_9]][] : memref<f32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_9]][] : memref<f32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: %[[VAL_20:.*]]:2 = scf.while (%[[VAL_21:.*]] = %[[VAL_16]], %[[VAL_22:.*]] = %[[VAL_18]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_23:.*]] = cmpi ult, %[[VAL_21]], %[[VAL_17]] : index
// CHECK: %[[VAL_24:.*]] = cmpi ult, %[[VAL_22]], %[[VAL_19]] : index
@@ -1084,36 +1084,36 @@ func @sum_reduction_ss(%arga: tensor<16xf32>,
// CHECK: scf.condition(%[[VAL_25]]) %[[VAL_21]], %[[VAL_22]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_26:.*]]: index, %[[VAL_27:.*]]: index):
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_26]]] : memref<?xindex>
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_30:.*]] = cmpi ult, %[[VAL_29]], %[[VAL_28]] : index
// CHECK: %[[VAL_31:.*]] = select %[[VAL_30]], %[[VAL_29]], %[[VAL_28]] : index
// CHECK: %[[VAL_32:.*]] = cmpi eq, %[[VAL_28]], %[[VAL_31]] : index
// CHECK: %[[VAL_33:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_31]] : index
// CHECK: %[[VAL_34:.*]] = and %[[VAL_32]], %[[VAL_33]] : i1
// CHECK: scf.if %[[VAL_34]] {
-// CHECK: %[[VAL_35:.*]] = load %[[VAL_14]][] : memref<f32>
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_26]]] : memref<?xf32>
+// CHECK: %[[VAL_35:.*]] = memref.load %[[VAL_14]][] : memref<f32>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_26]]] : memref<?xf32>
// CHECK: %[[VAL_37:.*]] = mulf %[[VAL_36]], %[[VAL_15]] : f32
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
// CHECK: %[[VAL_39:.*]] = addf %[[VAL_37]], %[[VAL_38]] : f32
// CHECK: %[[VAL_40:.*]] = addf %[[VAL_35]], %[[VAL_39]] : f32
-// CHECK: store %[[VAL_40]], %[[VAL_14]][] : memref<f32>
+// CHECK: memref.store %[[VAL_40]], %[[VAL_14]][] : memref<f32>
// CHECK: } else {
// CHECK: %[[VAL_41:.*]] = cmpi eq, %[[VAL_28]], %[[VAL_31]] : index
// CHECK: scf.if %[[VAL_41]] {
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_14]][] : memref<f32>
-// CHECK: %[[VAL_43:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_26]]] : memref<?xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_14]][] : memref<f32>
+// CHECK: %[[VAL_43:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_26]]] : memref<?xf32>
// CHECK: %[[VAL_44:.*]] = mulf %[[VAL_43]], %[[VAL_15]] : f32
// CHECK: %[[VAL_45:.*]] = addf %[[VAL_42]], %[[VAL_44]] : f32
-// CHECK: store %[[VAL_45]], %[[VAL_14]][] : memref<f32>
+// CHECK: memref.store %[[VAL_45]], %[[VAL_14]][] : memref<f32>
// CHECK: } else {
// CHECK: %[[VAL_46:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_31]] : index
// CHECK: scf.if %[[VAL_46]] {
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_14]][] : memref<f32>
-// CHECK: %[[VAL_48:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_14]][] : memref<f32>
+// CHECK: %[[VAL_48:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
// CHECK: %[[VAL_49:.*]] = addf %[[VAL_47]], %[[VAL_48]] : f32
-// CHECK: store %[[VAL_49]], %[[VAL_14]][] : memref<f32>
+// CHECK: memref.store %[[VAL_49]], %[[VAL_14]][] : memref<f32>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1126,20 +1126,20 @@ func @sum_reduction_ss(%arga: tensor<16xf32>,
// CHECK: %[[VAL_55:.*]] = select %[[VAL_53]], %[[VAL_54]], %[[VAL_27]] : index
// CHECK: scf.yield %[[VAL_52]], %[[VAL_55]] : index, index
// CHECK: }
-// CHECK: %[[VAL_56:.*]] = load %[[VAL_14]][] : memref<f32>
+// CHECK: %[[VAL_56:.*]] = memref.load %[[VAL_14]][] : memref<f32>
// CHECK: %[[VAL_57:.*]] = scf.for %[[VAL_58:.*]] = %[[VAL_59:.*]]#0 to %[[VAL_17]] step %[[VAL_5]] iter_args(%[[VAL_60:.*]] = %[[VAL_56]]) -> (f32) {
-// CHECK: %[[VAL_61:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_58]]] : memref<?xf32>
+// CHECK: %[[VAL_61:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_58]]] : memref<?xf32>
// CHECK: %[[VAL_62:.*]] = mulf %[[VAL_61]], %[[VAL_15]] : f32
// CHECK: %[[VAL_63:.*]] = addf %[[VAL_60]], %[[VAL_62]] : f32
// CHECK: scf.yield %[[VAL_63]] : f32
// CHECK: }
// CHECK: %[[VAL_64:.*]] = scf.for %[[VAL_65:.*]] = %[[VAL_66:.*]]#1 to %[[VAL_19]] step %[[VAL_5]] iter_args(%[[VAL_67:.*]] = %[[VAL_68:.*]]) -> (f32) {
-// CHECK: %[[VAL_69:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_65]]] : memref<?xf32>
+// CHECK: %[[VAL_69:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_65]]] : memref<?xf32>
// CHECK: %[[VAL_70:.*]] = addf %[[VAL_67]], %[[VAL_69]] : f32
// CHECK: scf.yield %[[VAL_70]] : f32
// CHECK: }
-// CHECK: store %[[VAL_71:.*]], %[[VAL_14]][] : memref<f32>
-// CHECK: %[[VAL_72:.*]] = tensor_load %[[VAL_14]] : memref<f32>
+// CHECK: memref.store %[[VAL_71:.*]], %[[VAL_14]][] : memref<f32>
+// CHECK: %[[VAL_72:.*]] = memref.tensor_load %[[VAL_14]] : memref<f32>
// CHECK: return %[[VAL_72]] : tensor<f32>
// CHECK: }
func @sum_reduction_inv(%arga: tensor<16xf32>,
@@ -1188,22 +1188,22 @@ func @sum_reduction_inv(%arga: tensor<16xf32>,
// CHECK: %[[VAL_5:.*]] = constant 0 : index
// CHECK: %[[VAL_6:.*]] = constant true
// CHECK: %[[VAL_7:.*]] = constant 1 : index
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_0]] : memref<?xf64>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_0]] : memref<?xf64>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_5]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_5]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<?xf64> to memref<?xf64>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_2]] : memref<?xf64>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_2]] : memref<?xf64>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_pointers %[[VAL_3]], %[[VAL_5]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_indices %[[VAL_3]], %[[VAL_5]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_15:.*]] = linalg.sparse_values %[[VAL_3]] : tensor<?xf64> to memref<?xf64>
-// CHECK: %[[VAL_16:.*]] = dim %[[VAL_4]], %[[VAL_5]] : tensor<?xf64>
-// CHECK: %[[VAL_17:.*]] = tensor_to_memref %[[VAL_4]] : memref<?xf64>
-// CHECK: %[[VAL_18:.*]] = alloc(%[[VAL_16]]) : memref<?xf64>
+// CHECK: %[[VAL_16:.*]] = memref.dim %[[VAL_4]], %[[VAL_5]] : tensor<?xf64>
+// CHECK: %[[VAL_17:.*]] = memref.buffer_cast %[[VAL_4]] : memref<?xf64>
+// CHECK: %[[VAL_18:.*]] = memref.alloc(%[[VAL_16]]) : memref<?xf64>
// CHECK: linalg.copy(%[[VAL_17]], %[[VAL_18]]) : memref<?xf64>, memref<?xf64>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_7]]] : memref<?xindex>
// CHECK: %[[VAL_23:.*]]:3 = scf.while (%[[VAL_24:.*]] = %[[VAL_19]], %[[VAL_25:.*]] = %[[VAL_21]], %[[VAL_26:.*]] = %[[VAL_5]]) : (index, index, index) -> (index, index, index) {
// CHECK: %[[VAL_27:.*]] = cmpi ult, %[[VAL_24]], %[[VAL_20]] : index
// CHECK: %[[VAL_28:.*]] = cmpi ult, %[[VAL_25]], %[[VAL_22]] : index
@@ -1211,44 +1211,44 @@ func @sum_reduction_inv(%arga: tensor<16xf32>,
// CHECK: scf.condition(%[[VAL_29]]) %[[VAL_24]], %[[VAL_25]], %[[VAL_26]] : index, index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_30:.*]]: index, %[[VAL_31:.*]]: index, %[[VAL_32:.*]]: index):
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_30]]] : memref<?xindex>
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_31]]] : memref<?xindex>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_30]]] : memref<?xindex>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_31]]] : memref<?xindex>
// CHECK: %[[VAL_35:.*]] = cmpi eq, %[[VAL_33]], %[[VAL_32]] : index
// CHECK: %[[VAL_36:.*]] = cmpi eq, %[[VAL_34]], %[[VAL_32]] : index
// CHECK: %[[VAL_37:.*]] = and %[[VAL_35]], %[[VAL_36]] : i1
// CHECK: scf.if %[[VAL_37]] {
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf64>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf64>
// CHECK: %[[VAL_40:.*]] = addf %[[VAL_38]], %[[VAL_39]] : f64
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xf64>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xf64>
// CHECK: %[[VAL_43:.*]] = addf %[[VAL_41]], %[[VAL_42]] : f64
// CHECK: %[[VAL_44:.*]] = addf %[[VAL_40]], %[[VAL_43]] : f64
-// CHECK: store %[[VAL_44]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_44]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: } else {
// CHECK: %[[VAL_45:.*]] = cmpi eq, %[[VAL_33]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_45]] {
-// CHECK: %[[VAL_46:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf64>
+// CHECK: %[[VAL_46:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf64>
// CHECK: %[[VAL_48:.*]] = addf %[[VAL_46]], %[[VAL_47]] : f64
-// CHECK: %[[VAL_49:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_49:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_50:.*]] = addf %[[VAL_48]], %[[VAL_49]] : f64
-// CHECK: store %[[VAL_50]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_50]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: } else {
// CHECK: %[[VAL_51:.*]] = cmpi eq, %[[VAL_34]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_51]] {
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_53:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_54:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xf64>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_53:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_54:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xf64>
// CHECK: %[[VAL_55:.*]] = addf %[[VAL_53]], %[[VAL_54]] : f64
// CHECK: %[[VAL_56:.*]] = addf %[[VAL_52]], %[[VAL_55]] : f64
-// CHECK: store %[[VAL_56]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_56]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_59:.*]] = addf %[[VAL_57]], %[[VAL_58]] : f64
-// CHECK: store %[[VAL_59]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_59]], %[[VAL_18]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1268,21 +1268,21 @@ func @sum_reduction_inv(%arga: tensor<16xf32>,
// CHECK: scf.condition(%[[VAL_71]]) %[[VAL_68]], %[[VAL_70]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_72:.*]]: index, %[[VAL_73:.*]]: index):
-// CHECK: %[[VAL_74:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_72]]] : memref<?xindex>
+// CHECK: %[[VAL_74:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_72]]] : memref<?xindex>
// CHECK: %[[VAL_75:.*]] = cmpi eq, %[[VAL_74]], %[[VAL_73]] : index
// CHECK: scf.if %[[VAL_75]] {
-// CHECK: %[[VAL_76:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_73]]] : memref<?xf64>
-// CHECK: %[[VAL_77:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_72]]] : memref<?xf64>
+// CHECK: %[[VAL_76:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: %[[VAL_77:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_72]]] : memref<?xf64>
// CHECK: %[[VAL_78:.*]] = addf %[[VAL_76]], %[[VAL_77]] : f64
-// CHECK: %[[VAL_79:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: %[[VAL_79:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_73]]] : memref<?xf64>
// CHECK: %[[VAL_80:.*]] = addf %[[VAL_78]], %[[VAL_79]] : f64
-// CHECK: store %[[VAL_80]], %[[VAL_18]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_80]], %[[VAL_18]]{{\[}}%[[VAL_73]]] : memref<?xf64>
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
-// CHECK: %[[VAL_81:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_73]]] : memref<?xf64>
-// CHECK: %[[VAL_82:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: %[[VAL_81:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: %[[VAL_82:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_73]]] : memref<?xf64>
// CHECK: %[[VAL_83:.*]] = addf %[[VAL_81]], %[[VAL_82]] : f64
-// CHECK: store %[[VAL_83]], %[[VAL_18]]{{\[}}%[[VAL_73]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_83]], %[[VAL_18]]{{\[}}%[[VAL_73]]] : memref<?xf64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1297,21 +1297,21 @@ func @sum_reduction_inv(%arga: tensor<16xf32>,
// CHECK: scf.condition(%[[VAL_93]]) %[[VAL_89]], %[[VAL_91]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_94:.*]]: index, %[[VAL_95:.*]]: index):
-// CHECK: %[[VAL_96:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_94]]] : memref<?xindex>
+// CHECK: %[[VAL_96:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_94]]] : memref<?xindex>
// CHECK: %[[VAL_97:.*]] = cmpi eq, %[[VAL_96]], %[[VAL_95]] : index
// CHECK: scf.if %[[VAL_97]] {
-// CHECK: %[[VAL_98:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_95]]] : memref<?xf64>
-// CHECK: %[[VAL_99:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_95]]] : memref<?xf64>
-// CHECK: %[[VAL_100:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_94]]] : memref<?xf64>
+// CHECK: %[[VAL_98:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: %[[VAL_99:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: %[[VAL_100:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_94]]] : memref<?xf64>
// CHECK: %[[VAL_101:.*]] = addf %[[VAL_99]], %[[VAL_100]] : f64
// CHECK: %[[VAL_102:.*]] = addf %[[VAL_98]], %[[VAL_101]] : f64
-// CHECK: store %[[VAL_102]], %[[VAL_18]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_102]], %[[VAL_18]]{{\[}}%[[VAL_95]]] : memref<?xf64>
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
-// CHECK: %[[VAL_103:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_95]]] : memref<?xf64>
-// CHECK: %[[VAL_104:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: %[[VAL_103:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: %[[VAL_104:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_95]]] : memref<?xf64>
// CHECK: %[[VAL_105:.*]] = addf %[[VAL_103]], %[[VAL_104]] : f64
-// CHECK: store %[[VAL_105]], %[[VAL_18]]{{\[}}%[[VAL_95]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_105]], %[[VAL_18]]{{\[}}%[[VAL_95]]] : memref<?xf64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1322,12 +1322,12 @@ func @sum_reduction_inv(%arga: tensor<16xf32>,
// CHECK: scf.yield %[[VAL_108]], %[[VAL_109]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_110:.*]] = %[[VAL_111:.*]]#1 to %[[VAL_16]] step %[[VAL_7]] {
-// CHECK: %[[VAL_112:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_110]]] : memref<?xf64>
-// CHECK: %[[VAL_113:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_110]]] : memref<?xf64>
+// CHECK: %[[VAL_112:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_110]]] : memref<?xf64>
+// CHECK: %[[VAL_113:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_110]]] : memref<?xf64>
// CHECK: %[[VAL_114:.*]] = addf %[[VAL_112]], %[[VAL_113]] : f64
-// CHECK: store %[[VAL_114]], %[[VAL_18]]{{\[}}%[[VAL_110]]] : memref<?xf64>
+// CHECK: memref.store %[[VAL_114]], %[[VAL_18]]{{\[}}%[[VAL_110]]] : memref<?xf64>
// CHECK: }
-// CHECK: %[[VAL_115:.*]] = tensor_load %[[VAL_18]] : memref<?xf64>
+// CHECK: %[[VAL_115:.*]] = memref.tensor_load %[[VAL_18]] : memref<?xf64>
// CHECK: return %[[VAL_115]] : tensor<?xf64>
// CHECK: }
func @four_tensors_op(%arga: tensor<?xf64>,
@@ -1380,15 +1380,15 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_2]], %[[VAL_4]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_2]], %[[VAL_4]] : tensor<?xf64> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_2]] : tensor<?xf64> to memref<?xf64>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_3]] : memref<f64>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<f64>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_3]] : memref<f64>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<f64>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<f64>, memref<f64>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: %[[VAL_23:.*]]:3 = scf.while (%[[VAL_24:.*]] = %[[VAL_17]], %[[VAL_25:.*]] = %[[VAL_19]], %[[VAL_26:.*]] = %[[VAL_21]]) : (index, index, index) -> (index, index, index) {
// CHECK: %[[VAL_27:.*]] = cmpi ult, %[[VAL_24]], %[[VAL_18]] : index
// CHECK: %[[VAL_28:.*]] = cmpi ult, %[[VAL_25]], %[[VAL_20]] : index
@@ -1398,11 +1398,11 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: scf.condition(%[[VAL_31]]) %[[VAL_24]], %[[VAL_25]], %[[VAL_26]] : index, index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_32:.*]]: index, %[[VAL_33:.*]]: index, %[[VAL_34:.*]]: index):
-// CHECK: %[[VAL_35:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_32]]] : memref<?xindex>
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_33]]] : memref<?xindex>
+// CHECK: %[[VAL_35:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_32]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_33]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = cmpi ult, %[[VAL_36]], %[[VAL_35]] : index
// CHECK: %[[VAL_38:.*]] = select %[[VAL_37]], %[[VAL_36]], %[[VAL_35]] : index
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_34]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_34]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = cmpi ult, %[[VAL_39]], %[[VAL_38]] : index
// CHECK: %[[VAL_41:.*]] = select %[[VAL_40]], %[[VAL_39]], %[[VAL_38]] : index
// CHECK: %[[VAL_42:.*]] = cmpi eq, %[[VAL_35]], %[[VAL_41]] : index
@@ -1411,68 +1411,68 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: %[[VAL_45:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_41]] : index
// CHECK: %[[VAL_46:.*]] = and %[[VAL_44]], %[[VAL_45]] : i1
// CHECK: scf.if %[[VAL_46]] {
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_48:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_48:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_49:.*]] = addf %[[VAL_47]], %[[VAL_48]] : f64
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
// CHECK: %[[VAL_51:.*]] = addf %[[VAL_49]], %[[VAL_50]] : f64
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
// CHECK: %[[VAL_53:.*]] = addf %[[VAL_51]], %[[VAL_52]] : f64
-// CHECK: store %[[VAL_53]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_53]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_54:.*]] = cmpi eq, %[[VAL_36]], %[[VAL_41]] : index
// CHECK: %[[VAL_55:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_41]] : index
// CHECK: %[[VAL_56:.*]] = and %[[VAL_54]], %[[VAL_55]] : i1
// CHECK: scf.if %[[VAL_56]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
// CHECK: %[[VAL_59:.*]] = addf %[[VAL_57]], %[[VAL_58]] : f64
-// CHECK: %[[VAL_60:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
+// CHECK: %[[VAL_60:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
// CHECK: %[[VAL_61:.*]] = addf %[[VAL_59]], %[[VAL_60]] : f64
-// CHECK: store %[[VAL_61]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_61]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_62:.*]] = cmpi eq, %[[VAL_35]], %[[VAL_41]] : index
// CHECK: %[[VAL_63:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_41]] : index
// CHECK: %[[VAL_64:.*]] = and %[[VAL_62]], %[[VAL_63]] : i1
// CHECK: scf.if %[[VAL_64]] {
-// CHECK: %[[VAL_65:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_66:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_65:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_66:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_67:.*]] = addf %[[VAL_65]], %[[VAL_66]] : f64
-// CHECK: %[[VAL_68:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
+// CHECK: %[[VAL_68:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
// CHECK: %[[VAL_69:.*]] = addf %[[VAL_67]], %[[VAL_68]] : f64
-// CHECK: store %[[VAL_69]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_69]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_70:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_41]] : index
// CHECK: scf.if %[[VAL_70]] {
-// CHECK: %[[VAL_71:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_72:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
+// CHECK: %[[VAL_71:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_72:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_34]]] : memref<?xf64>
// CHECK: %[[VAL_73:.*]] = addf %[[VAL_71]], %[[VAL_72]] : f64
-// CHECK: store %[[VAL_73]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_73]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_74:.*]] = cmpi eq, %[[VAL_35]], %[[VAL_41]] : index
// CHECK: %[[VAL_75:.*]] = cmpi eq, %[[VAL_36]], %[[VAL_41]] : index
// CHECK: %[[VAL_76:.*]] = and %[[VAL_74]], %[[VAL_75]] : i1
// CHECK: scf.if %[[VAL_76]] {
-// CHECK: %[[VAL_77:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_78:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_77:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_78:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_79:.*]] = addf %[[VAL_77]], %[[VAL_78]] : f64
-// CHECK: %[[VAL_80:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
+// CHECK: %[[VAL_80:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
// CHECK: %[[VAL_81:.*]] = addf %[[VAL_79]], %[[VAL_80]] : f64
-// CHECK: store %[[VAL_81]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_81]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_82:.*]] = cmpi eq, %[[VAL_36]], %[[VAL_41]] : index
// CHECK: scf.if %[[VAL_82]] {
-// CHECK: %[[VAL_83:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_84:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
+// CHECK: %[[VAL_83:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_84:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xf64>
// CHECK: %[[VAL_85:.*]] = addf %[[VAL_83]], %[[VAL_84]] : f64
-// CHECK: store %[[VAL_85]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_85]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_86:.*]] = cmpi eq, %[[VAL_35]], %[[VAL_41]] : index
// CHECK: scf.if %[[VAL_86]] {
-// CHECK: %[[VAL_87:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_88:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
+// CHECK: %[[VAL_87:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_88:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_32]]] : memref<?xf64>
// CHECK: %[[VAL_89:.*]] = addf %[[VAL_87]], %[[VAL_88]] : f64
-// CHECK: store %[[VAL_89]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_89]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1499,34 +1499,34 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: scf.condition(%[[VAL_105]]) %[[VAL_100]], %[[VAL_102]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_106:.*]]: index, %[[VAL_107:.*]]: index):
-// CHECK: %[[VAL_108:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_106]]] : memref<?xindex>
-// CHECK: %[[VAL_109:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_107]]] : memref<?xindex>
+// CHECK: %[[VAL_108:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_106]]] : memref<?xindex>
+// CHECK: %[[VAL_109:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_107]]] : memref<?xindex>
// CHECK: %[[VAL_110:.*]] = cmpi ult, %[[VAL_109]], %[[VAL_108]] : index
// CHECK: %[[VAL_111:.*]] = select %[[VAL_110]], %[[VAL_109]], %[[VAL_108]] : index
// CHECK: %[[VAL_112:.*]] = cmpi eq, %[[VAL_108]], %[[VAL_111]] : index
// CHECK: %[[VAL_113:.*]] = cmpi eq, %[[VAL_109]], %[[VAL_111]] : index
// CHECK: %[[VAL_114:.*]] = and %[[VAL_112]], %[[VAL_113]] : i1
// CHECK: scf.if %[[VAL_114]] {
-// CHECK: %[[VAL_115:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_116:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_106]]] : memref<?xf64>
+// CHECK: %[[VAL_115:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_116:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_106]]] : memref<?xf64>
// CHECK: %[[VAL_117:.*]] = addf %[[VAL_115]], %[[VAL_116]] : f64
-// CHECK: %[[VAL_118:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_107]]] : memref<?xf64>
+// CHECK: %[[VAL_118:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_107]]] : memref<?xf64>
// CHECK: %[[VAL_119:.*]] = addf %[[VAL_117]], %[[VAL_118]] : f64
-// CHECK: store %[[VAL_119]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_119]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_120:.*]] = cmpi eq, %[[VAL_109]], %[[VAL_111]] : index
// CHECK: scf.if %[[VAL_120]] {
-// CHECK: %[[VAL_121:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_122:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_107]]] : memref<?xf64>
+// CHECK: %[[VAL_121:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_122:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_107]]] : memref<?xf64>
// CHECK: %[[VAL_123:.*]] = addf %[[VAL_121]], %[[VAL_122]] : f64
-// CHECK: store %[[VAL_123]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_123]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_124:.*]] = cmpi eq, %[[VAL_108]], %[[VAL_111]] : index
// CHECK: scf.if %[[VAL_124]] {
-// CHECK: %[[VAL_125:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_126:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_106]]] : memref<?xf64>
+// CHECK: %[[VAL_125:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_126:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_106]]] : memref<?xf64>
// CHECK: %[[VAL_127:.*]] = addf %[[VAL_125]], %[[VAL_126]] : f64
-// CHECK: store %[[VAL_127]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_127]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1546,34 +1546,34 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: scf.condition(%[[VAL_141]]) %[[VAL_135]], %[[VAL_137]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_142:.*]]: index, %[[VAL_143:.*]]: index):
-// CHECK: %[[VAL_144:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_142]]] : memref<?xindex>
-// CHECK: %[[VAL_145:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_143]]] : memref<?xindex>
+// CHECK: %[[VAL_144:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_142]]] : memref<?xindex>
+// CHECK: %[[VAL_145:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_143]]] : memref<?xindex>
// CHECK: %[[VAL_146:.*]] = cmpi ult, %[[VAL_145]], %[[VAL_144]] : index
// CHECK: %[[VAL_147:.*]] = select %[[VAL_146]], %[[VAL_145]], %[[VAL_144]] : index
// CHECK: %[[VAL_148:.*]] = cmpi eq, %[[VAL_144]], %[[VAL_147]] : index
// CHECK: %[[VAL_149:.*]] = cmpi eq, %[[VAL_145]], %[[VAL_147]] : index
// CHECK: %[[VAL_150:.*]] = and %[[VAL_148]], %[[VAL_149]] : i1
// CHECK: scf.if %[[VAL_150]] {
-// CHECK: %[[VAL_151:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_152:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_142]]] : memref<?xf64>
+// CHECK: %[[VAL_151:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_152:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_142]]] : memref<?xf64>
// CHECK: %[[VAL_153:.*]] = addf %[[VAL_151]], %[[VAL_152]] : f64
-// CHECK: %[[VAL_154:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_143]]] : memref<?xf64>
+// CHECK: %[[VAL_154:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_143]]] : memref<?xf64>
// CHECK: %[[VAL_155:.*]] = addf %[[VAL_153]], %[[VAL_154]] : f64
-// CHECK: store %[[VAL_155]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_155]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_156:.*]] = cmpi eq, %[[VAL_145]], %[[VAL_147]] : index
// CHECK: scf.if %[[VAL_156]] {
-// CHECK: %[[VAL_157:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_158:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_143]]] : memref<?xf64>
+// CHECK: %[[VAL_157:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_158:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_143]]] : memref<?xf64>
// CHECK: %[[VAL_159:.*]] = addf %[[VAL_157]], %[[VAL_158]] : f64
-// CHECK: store %[[VAL_159]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_159]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_160:.*]] = cmpi eq, %[[VAL_144]], %[[VAL_147]] : index
// CHECK: scf.if %[[VAL_160]] {
-// CHECK: %[[VAL_161:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_162:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_142]]] : memref<?xf64>
+// CHECK: %[[VAL_161:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_162:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_142]]] : memref<?xf64>
// CHECK: %[[VAL_163:.*]] = addf %[[VAL_161]], %[[VAL_162]] : f64
-// CHECK: store %[[VAL_163]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_163]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1586,13 +1586,13 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: %[[VAL_169:.*]] = select %[[VAL_167]], %[[VAL_168]], %[[VAL_143]] : index
// CHECK: scf.yield %[[VAL_166]], %[[VAL_169]] : index, index
// CHECK: }
-// CHECK: %[[VAL_170:.*]] = load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_170:.*]] = memref.load %[[VAL_16]][] : memref<f64>
// CHECK: %[[VAL_171:.*]] = scf.for %[[VAL_172:.*]] = %[[VAL_173:.*]]#1 to %[[VAL_22]] step %[[VAL_5]] iter_args(%[[VAL_174:.*]] = %[[VAL_170]]) -> (f64) {
-// CHECK: %[[VAL_175:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_172]]] : memref<?xf64>
+// CHECK: %[[VAL_175:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_172]]] : memref<?xf64>
// CHECK: %[[VAL_176:.*]] = addf %[[VAL_174]], %[[VAL_175]] : f64
// CHECK: scf.yield %[[VAL_176]] : f64
// CHECK: }
-// CHECK: store %[[VAL_177:.*]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_177:.*]], %[[VAL_16]][] : memref<f64>
// CHECK: %[[VAL_178:.*]]:2 = scf.while (%[[VAL_179:.*]] = %[[VAL_180:.*]]#0, %[[VAL_181:.*]] = %[[VAL_182:.*]]#0) : (index, index) -> (index, index) {
// CHECK: %[[VAL_183:.*]] = cmpi ult, %[[VAL_179]], %[[VAL_18]] : index
// CHECK: %[[VAL_184:.*]] = cmpi ult, %[[VAL_181]], %[[VAL_20]] : index
@@ -1600,34 +1600,34 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: scf.condition(%[[VAL_185]]) %[[VAL_179]], %[[VAL_181]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_186:.*]]: index, %[[VAL_187:.*]]: index):
-// CHECK: %[[VAL_188:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_186]]] : memref<?xindex>
-// CHECK: %[[VAL_189:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_187]]] : memref<?xindex>
+// CHECK: %[[VAL_188:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_186]]] : memref<?xindex>
+// CHECK: %[[VAL_189:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_187]]] : memref<?xindex>
// CHECK: %[[VAL_190:.*]] = cmpi ult, %[[VAL_189]], %[[VAL_188]] : index
// CHECK: %[[VAL_191:.*]] = select %[[VAL_190]], %[[VAL_189]], %[[VAL_188]] : index
// CHECK: %[[VAL_192:.*]] = cmpi eq, %[[VAL_188]], %[[VAL_191]] : index
// CHECK: %[[VAL_193:.*]] = cmpi eq, %[[VAL_189]], %[[VAL_191]] : index
// CHECK: %[[VAL_194:.*]] = and %[[VAL_192]], %[[VAL_193]] : i1
// CHECK: scf.if %[[VAL_194]] {
-// CHECK: %[[VAL_195:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_196:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_186]]] : memref<?xf64>
+// CHECK: %[[VAL_195:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_196:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_186]]] : memref<?xf64>
// CHECK: %[[VAL_197:.*]] = addf %[[VAL_195]], %[[VAL_196]] : f64
-// CHECK: %[[VAL_198:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_187]]] : memref<?xf64>
+// CHECK: %[[VAL_198:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_187]]] : memref<?xf64>
// CHECK: %[[VAL_199:.*]] = addf %[[VAL_197]], %[[VAL_198]] : f64
-// CHECK: store %[[VAL_199]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_199]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_200:.*]] = cmpi eq, %[[VAL_189]], %[[VAL_191]] : index
// CHECK: scf.if %[[VAL_200]] {
-// CHECK: %[[VAL_201:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_202:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_187]]] : memref<?xf64>
+// CHECK: %[[VAL_201:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_202:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_187]]] : memref<?xf64>
// CHECK: %[[VAL_203:.*]] = addf %[[VAL_201]], %[[VAL_202]] : f64
-// CHECK: store %[[VAL_203]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_203]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: %[[VAL_204:.*]] = cmpi eq, %[[VAL_188]], %[[VAL_191]] : index
// CHECK: scf.if %[[VAL_204]] {
-// CHECK: %[[VAL_205:.*]] = load %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_206:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_186]]] : memref<?xf64>
+// CHECK: %[[VAL_205:.*]] = memref.load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_206:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_186]]] : memref<?xf64>
// CHECK: %[[VAL_207:.*]] = addf %[[VAL_205]], %[[VAL_206]] : f64
-// CHECK: store %[[VAL_207]], %[[VAL_16]][] : memref<f64>
+// CHECK: memref.store %[[VAL_207]], %[[VAL_16]][] : memref<f64>
// CHECK: } else {
// CHECK: }
// CHECK: }
@@ -1640,19 +1640,19 @@ func @four_tensors_op(%arga: tensor<?xf64>,
// CHECK: %[[VAL_213:.*]] = select %[[VAL_211]], %[[VAL_212]], %[[VAL_187]] : index
// CHECK: scf.yield %[[VAL_210]], %[[VAL_213]] : index, index
// CHECK: }
-// CHECK: %[[VAL_214:.*]] = load %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_214:.*]] = memref.load %[[VAL_16]][] : memref<f64>
// CHECK: %[[VAL_215:.*]] = scf.for %[[VAL_216:.*]] = %[[VAL_217:.*]]#1 to %[[VAL_20]] step %[[VAL_5]] iter_args(%[[VAL_218:.*]] = %[[VAL_214]]) -> (f64) {
-// CHECK: %[[VAL_219:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_216]]] : memref<?xf64>
+// CHECK: %[[VAL_219:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_216]]] : memref<?xf64>
// CHECK: %[[VAL_220:.*]] = addf %[[VAL_218]], %[[VAL_219]] : f64
// CHECK: scf.yield %[[VAL_220]] : f64
// CHECK: }
// CHECK: %[[VAL_221:.*]] = scf.for %[[VAL_222:.*]] = %[[VAL_223:.*]]#0 to %[[VAL_18]] step %[[VAL_5]] iter_args(%[[VAL_224:.*]] = %[[VAL_225:.*]]) -> (f64) {
-// CHECK: %[[VAL_226:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_222]]] : memref<?xf64>
+// CHECK: %[[VAL_226:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_222]]] : memref<?xf64>
// CHECK: %[[VAL_227:.*]] = addf %[[VAL_224]], %[[VAL_226]] : f64
// CHECK: scf.yield %[[VAL_227]] : f64
// CHECK: }
-// CHECK: store %[[VAL_228:.*]], %[[VAL_16]][] : memref<f64>
-// CHECK: %[[VAL_229:.*]] = tensor_load %[[VAL_16]] : memref<f64>
+// CHECK: memref.store %[[VAL_228:.*]], %[[VAL_16]][] : memref<f64>
+// CHECK: %[[VAL_229:.*]] = memref.tensor_load %[[VAL_16]] : memref<f64>
// CHECK: return %[[VAL_229]] : tensor<f64>
// CHECK: }
func @red3s(%arga: tensor<?xf64>,
diff --git a/mlir/test/Dialect/Linalg/sparse_2d.mlir b/mlir/test/Dialect/Linalg/sparse_2d.mlir
index d560e042865b..24ccdfc20b14 100644
--- a/mlir/test/Dialect/Linalg/sparse_2d.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_2d.mlir
@@ -24,20 +24,20 @@
// CHECK: %[[VAL_4:.*]] = constant 16 : index
// CHECK: %[[VAL_5:.*]] = constant 0 : index
// CHECK: %[[VAL_6:.*]] = constant 1 : index
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_0]] : memref<32x16xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_10:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32x16xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_10:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_9]], %[[VAL_10]]) : memref<32x16xf32>, memref<32x16xf32>
// CHECK: scf.for %[[VAL_11:.*]] = %[[VAL_5]] to %[[VAL_3]] step %[[VAL_6]] {
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
// CHECK: %[[VAL_15:.*]] = addf %[[VAL_13]], %[[VAL_14]] : f32
// CHECK: store %[[VAL_15]], %[[VAL_10]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_16:.*]] = tensor_load %[[VAL_10]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.tensor_load %[[VAL_10]] : memref<32x16xf32>
// CHECK: return %[[VAL_16]] : tensor<32x16xf32>
// CHECK: }
func @add_dd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -59,20 +59,20 @@ func @add_dd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_4:.*]] = constant 16 : index
// CHECK: %[[VAL_5:.*]] = constant 0 : index
// CHECK: %[[VAL_6:.*]] = constant 1 : index
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_0]] : memref<32x16xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_10:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32x16xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_10:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_9]], %[[VAL_10]]) : memref<32x16xf32>, memref<32x16xf32>
// CHECK: scf.for %[[VAL_11:.*]] = %[[VAL_5]] to %[[VAL_3]] step %[[VAL_6]] {
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
// CHECK: %[[VAL_15:.*]] = mulf %[[VAL_13]], %[[VAL_14]] : f32
// CHECK: store %[[VAL_15]], %[[VAL_10]]{{\[}}%[[VAL_11]], %[[VAL_12]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_16:.*]] = tensor_load %[[VAL_10]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.tensor_load %[[VAL_10]] : memref<32x16xf32>
// CHECK: return %[[VAL_16]] : tensor<32x16xf32>
// CHECK: }
func @mul_dd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -113,29 +113,29 @@ func @mul_dd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_7]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_7]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<32x16xf32>, memref<32x16xf32>
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_5]] to %[[VAL_3]] step %[[VAL_7]] {
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_14]]] : memref<?xindex>
// CHECK: %[[VAL_16:.*]] = addi %[[VAL_14]], %[[VAL_7]] : index
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xindex>
// CHECK: %[[VAL_18:.*]]:2 = scf.while (%[[VAL_19:.*]] = %[[VAL_15]], %[[VAL_20:.*]] = %[[VAL_5]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_17]] : index
// CHECK: scf.condition(%[[VAL_21]]) %[[VAL_19]], %[[VAL_20]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_22:.*]]: index, %[[VAL_23:.*]]: index):
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
// CHECK: %[[VAL_25:.*]] = cmpi eq, %[[VAL_24]], %[[VAL_23]] : index
// CHECK: scf.if %[[VAL_25]] {
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_22]]] : memref<?xf32>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_22]]] : memref<?xf32>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
// CHECK: %[[VAL_28:.*]] = addf %[[VAL_26]], %[[VAL_27]] : f32
// CHECK: store %[[VAL_28]], %[[VAL_13]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
// CHECK: store %[[VAL_29]], %[[VAL_13]]{{\[}}%[[VAL_14]], %[[VAL_23]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: }
@@ -147,11 +147,11 @@ func @mul_dd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.yield %[[VAL_32]], %[[VAL_33]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_34:.*]] = %[[VAL_35:.*]]#1 to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_34]]] : memref<32x16xf32>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: store %[[VAL_36]], %[[VAL_13]]{{\[}}%[[VAL_14]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_37:.*]] = tensor_load %[[VAL_13]] : memref<32x16xf32>
+// CHECK: %[[VAL_37:.*]] = memref.tensor_load %[[VAL_13]] : memref<32x16xf32>
// CHECK: return %[[VAL_37]] : tensor<32x16xf32>
// CHECK: }
func @add_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -175,23 +175,23 @@ func @add_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_5]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_5]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<32x16xf32>, memref<32x16xf32>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
// CHECK: %[[VAL_14:.*]] = addi %[[VAL_12]], %[[VAL_5]] : index
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_13]] to %[[VAL_15]] step %[[VAL_5]] {
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xf32>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_17]]] : memref<32x16xf32>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xf32>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_17]]] : memref<32x16xf32>
// CHECK: %[[VAL_20:.*]] = mulf %[[VAL_18]], %[[VAL_19]] : f32
// CHECK: store %[[VAL_20]], %[[VAL_11]]{{\[}}%[[VAL_12]], %[[VAL_17]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_21:.*]] = tensor_load %[[VAL_11]] : memref<32x16xf32>
+// CHECK: %[[VAL_21:.*]] = memref.tensor_load %[[VAL_11]] : memref<32x16xf32>
// CHECK: return %[[VAL_21]] : tensor<32x16xf32>
// CHECK: }
func @mul_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -232,32 +232,32 @@ func @mul_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_6]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_6]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_7]]] : memref<?xindex>
// CHECK: %[[VAL_16:.*]]:2 = scf.while (%[[VAL_17:.*]] = %[[VAL_14]], %[[VAL_18:.*]] = %[[VAL_6]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_19:.*]] = cmpi ult, %[[VAL_17]], %[[VAL_15]] : index
// CHECK: scf.condition(%[[VAL_19]]) %[[VAL_17]], %[[VAL_18]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_20:.*]]: index, %[[VAL_21:.*]]: index):
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<?xindex>
// CHECK: %[[VAL_23:.*]] = cmpi eq, %[[VAL_22]], %[[VAL_21]] : index
// CHECK: scf.if %[[VAL_23]] {
// CHECK: scf.for %[[VAL_24:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
// CHECK: %[[VAL_25:.*]] = muli %[[VAL_20]], %[[VAL_4]] : index
// CHECK: %[[VAL_26:.*]] = addi %[[VAL_25]], %[[VAL_24]] : index
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_26]]] : memref<?xf32>
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_24]]] : memref<32x16xf32>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_26]]] : memref<?xf32>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_24]]] : memref<32x16xf32>
// CHECK: %[[VAL_29:.*]] = addf %[[VAL_27]], %[[VAL_28]] : f32
// CHECK: store %[[VAL_29]], %[[VAL_13]]{{\[}}%[[VAL_21]], %[[VAL_24]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
// CHECK: scf.for %[[VAL_30:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_30]]] : memref<32x16xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_30]]] : memref<32x16xf32>
// CHECK: store %[[VAL_31]], %[[VAL_13]]{{\[}}%[[VAL_21]], %[[VAL_30]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
@@ -271,11 +271,11 @@ func @mul_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: }
// CHECK: scf.for %[[VAL_36:.*]] = %[[VAL_37:.*]]#1 to %[[VAL_3]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_38:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_36]], %[[VAL_38]]] : memref<32x16xf32>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_36]], %[[VAL_38]]] : memref<32x16xf32>
// CHECK: store %[[VAL_39]], %[[VAL_13]]{{\[}}%[[VAL_36]], %[[VAL_38]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_40:.*]] = tensor_load %[[VAL_13]] : memref<32x16xf32>
+// CHECK: %[[VAL_40:.*]] = memref.tensor_load %[[VAL_13]] : memref<32x16xf32>
// CHECK: return %[[VAL_40]] : tensor<32x16xf32>
// CHECK: }
func @add_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -299,24 +299,24 @@ func @add_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_12]] to %[[VAL_13]] step %[[VAL_5]] {
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_14]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
// CHECK: %[[VAL_17:.*]] = muli %[[VAL_14]], %[[VAL_3]] : index
// CHECK: %[[VAL_18:.*]] = addi %[[VAL_17]], %[[VAL_16]] : index
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf32>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_15]], %[[VAL_16]]] : memref<32x16xf32>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf32>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_15]], %[[VAL_16]]] : memref<32x16xf32>
// CHECK: %[[VAL_21:.*]] = mulf %[[VAL_19]], %[[VAL_20]] : f32
// CHECK: store %[[VAL_21]], %[[VAL_11]]{{\[}}%[[VAL_15]], %[[VAL_16]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_22:.*]] = tensor_load %[[VAL_11]] : memref<32x16xf32>
+// CHECK: %[[VAL_22:.*]] = memref.tensor_load %[[VAL_11]] : memref<32x16xf32>
// CHECK: return %[[VAL_22]] : tensor<32x16xf32>
// CHECK: }
func @mul_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -359,38 +359,38 @@ func @mul_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_10:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_7]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_7]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_12:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_14:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_15:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_14:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_14]], %[[VAL_15]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_7]]] : memref<?xindex>
// CHECK: %[[VAL_18:.*]]:2 = scf.while (%[[VAL_19:.*]] = %[[VAL_16]], %[[VAL_20:.*]] = %[[VAL_6]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_21:.*]] = cmpi ult, %[[VAL_19]], %[[VAL_17]] : index
// CHECK: scf.condition(%[[VAL_21]]) %[[VAL_19]], %[[VAL_20]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_22:.*]]: index, %[[VAL_23:.*]]: index):
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
// CHECK: %[[VAL_25:.*]] = cmpi eq, %[[VAL_24]], %[[VAL_23]] : index
// CHECK: scf.if %[[VAL_25]] {
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_22]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_22]]] : memref<?xindex>
// CHECK: %[[VAL_27:.*]] = addi %[[VAL_22]], %[[VAL_7]] : index
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_29:.*]]:2 = scf.while (%[[VAL_30:.*]] = %[[VAL_26]], %[[VAL_31:.*]] = %[[VAL_6]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_32:.*]] = cmpi ult, %[[VAL_30]], %[[VAL_28]] : index
// CHECK: scf.condition(%[[VAL_32]]) %[[VAL_30]], %[[VAL_31]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_33:.*]]: index, %[[VAL_34:.*]]: index):
-// CHECK: %[[VAL_35:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xindex>
+// CHECK: %[[VAL_35:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_33]]] : memref<?xindex>
// CHECK: %[[VAL_36:.*]] = cmpi eq, %[[VAL_35]], %[[VAL_34]] : index
// CHECK: scf.if %[[VAL_36]] {
-// CHECK: %[[VAL_37:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
+// CHECK: %[[VAL_37:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: %[[VAL_39:.*]] = addf %[[VAL_37]], %[[VAL_38]] : f32
// CHECK: store %[[VAL_39]], %[[VAL_15]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
-// CHECK: %[[VAL_40:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
+// CHECK: %[[VAL_40:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: store %[[VAL_40]], %[[VAL_15]]{{\[}}%[[VAL_23]], %[[VAL_34]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: }
@@ -402,13 +402,13 @@ func @mul_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.yield %[[VAL_43]], %[[VAL_44]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_45:.*]] = %[[VAL_46:.*]]#1 to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_45]]] : memref<32x16xf32>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_45]]] : memref<32x16xf32>
// CHECK: store %[[VAL_47]], %[[VAL_15]]{{\[}}%[[VAL_23]], %[[VAL_45]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_5]] {
// CHECK: scf.for %[[VAL_48:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_49:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_48]]] : memref<32x16xf32>
+// CHECK: %[[VAL_49:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_23]], %[[VAL_48]]] : memref<32x16xf32>
// CHECK: store %[[VAL_49]], %[[VAL_15]]{{\[}}%[[VAL_23]], %[[VAL_48]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
@@ -422,11 +422,11 @@ func @mul_sd(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: }
// CHECK: scf.for %[[VAL_54:.*]] = %[[VAL_55:.*]]#1 to %[[VAL_3]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_56:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_54]], %[[VAL_56]]] : memref<32x16xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_54]], %[[VAL_56]]] : memref<32x16xf32>
// CHECK: store %[[VAL_57]], %[[VAL_15]]{{\[}}%[[VAL_54]], %[[VAL_56]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_58:.*]] = tensor_load %[[VAL_15]] : memref<32x16xf32>
+// CHECK: %[[VAL_58:.*]] = memref.tensor_load %[[VAL_15]] : memref<32x16xf32>
// CHECK: return %[[VAL_58]] : tensor<32x16xf32>
// CHECK: }
func @add_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -451,26 +451,26 @@ func @add_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_13]] to %[[VAL_14]] step %[[VAL_4]] {
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: %[[VAL_18:.*]] = addi %[[VAL_15]], %[[VAL_4]] : index
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_20:.*]] = %[[VAL_17]] to %[[VAL_19]] step %[[VAL_4]] {
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xindex>
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<?xf32>
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_16]], %[[VAL_21]]] : memref<32x16xf32>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<?xf32>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_16]], %[[VAL_21]]] : memref<32x16xf32>
// CHECK: %[[VAL_24:.*]] = mulf %[[VAL_22]], %[[VAL_23]] : f32
// CHECK: store %[[VAL_24]], %[[VAL_12]]{{\[}}%[[VAL_16]], %[[VAL_21]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_25:.*]] = tensor_load %[[VAL_12]] : memref<32x16xf32>
+// CHECK: %[[VAL_25:.*]] = memref.tensor_load %[[VAL_12]] : memref<32x16xf32>
// CHECK: return %[[VAL_25]] : tensor<32x16xf32>
// CHECK: }
func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -515,13 +515,13 @@ func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]]:2 = scf.while (%[[VAL_22:.*]] = %[[VAL_17]], %[[VAL_23:.*]] = %[[VAL_19]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_24:.*]] = cmpi ult, %[[VAL_22]], %[[VAL_18]] : index
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_20]] : index
@@ -529,20 +529,20 @@ func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.condition(%[[VAL_26]]) %[[VAL_22]], %[[VAL_23]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_27:.*]]: index, %[[VAL_28:.*]]: index):
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = cmpi ult, %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = select %[[VAL_31]], %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_33:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: %[[VAL_34:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: %[[VAL_35:.*]] = and %[[VAL_33]], %[[VAL_34]] : i1
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
// CHECK: %[[VAL_42:.*]]:2 = scf.while (%[[VAL_43:.*]] = %[[VAL_36]], %[[VAL_44:.*]] = %[[VAL_39]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_45:.*]] = cmpi ult, %[[VAL_43]], %[[VAL_38]] : index
// CHECK: %[[VAL_46:.*]] = cmpi ult, %[[VAL_44]], %[[VAL_41]] : index
@@ -550,27 +550,27 @@ func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.condition(%[[VAL_47]]) %[[VAL_43]], %[[VAL_44]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_48:.*]]: index, %[[VAL_49:.*]]: index):
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
// CHECK: %[[VAL_52:.*]] = cmpi ult, %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_53:.*]] = select %[[VAL_52]], %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_54:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: %[[VAL_55:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: %[[VAL_56:.*]] = and %[[VAL_54]], %[[VAL_55]] : i1
// CHECK: scf.if %[[VAL_56]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: %[[VAL_59:.*]] = addf %[[VAL_57]], %[[VAL_58]] : f32
// CHECK: store %[[VAL_59]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: %[[VAL_60:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: scf.if %[[VAL_60]] {
-// CHECK: %[[VAL_61:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_61:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
// CHECK: store %[[VAL_61]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: %[[VAL_62:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: scf.if %[[VAL_62]] {
-// CHECK: %[[VAL_63:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_63:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: store %[[VAL_63]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: }
@@ -585,35 +585,35 @@ func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.yield %[[VAL_66]], %[[VAL_69]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_70:.*]] = %[[VAL_71:.*]]#0 to %[[VAL_38]] step %[[VAL_4]] {
-// CHECK: %[[VAL_72:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_70]]] : memref<?xindex>
-// CHECK: %[[VAL_73:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_70]]] : memref<?xf32>
+// CHECK: %[[VAL_72:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_70]]] : memref<?xindex>
+// CHECK: %[[VAL_73:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_70]]] : memref<?xf32>
// CHECK: store %[[VAL_73]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_72]]] : memref<32x16xf32>
// CHECK: }
// CHECK: scf.for %[[VAL_74:.*]] = %[[VAL_75:.*]]#1 to %[[VAL_41]] step %[[VAL_4]] {
-// CHECK: %[[VAL_76:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_74]]] : memref<?xindex>
-// CHECK: %[[VAL_77:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_74]]] : memref<?xf32>
+// CHECK: %[[VAL_76:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_74]]] : memref<?xindex>
+// CHECK: %[[VAL_77:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_74]]] : memref<?xf32>
// CHECK: store %[[VAL_77]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_76]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: %[[VAL_78:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_78]] {
-// CHECK: %[[VAL_79:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_79:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_80:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_81:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_80]]] : memref<?xindex>
+// CHECK: %[[VAL_81:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_80]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_82:.*]] = %[[VAL_79]] to %[[VAL_81]] step %[[VAL_4]] {
-// CHECK: %[[VAL_83:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_82]]] : memref<?xindex>
-// CHECK: %[[VAL_84:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_82]]] : memref<?xf32>
+// CHECK: %[[VAL_83:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_82]]] : memref<?xindex>
+// CHECK: %[[VAL_84:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_82]]] : memref<?xf32>
// CHECK: store %[[VAL_84]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_83]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: %[[VAL_85:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_85]] {
-// CHECK: %[[VAL_86:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_86:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_87:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_88:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_87]]] : memref<?xindex>
+// CHECK: %[[VAL_88:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_87]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_89:.*]] = %[[VAL_86]] to %[[VAL_88]] step %[[VAL_4]] {
-// CHECK: %[[VAL_90:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_89]]] : memref<?xindex>
-// CHECK: %[[VAL_91:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_89]]] : memref<?xf32>
+// CHECK: %[[VAL_90:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_89]]] : memref<?xindex>
+// CHECK: %[[VAL_91:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_89]]] : memref<?xf32>
// CHECK: store %[[VAL_91]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_90]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
@@ -629,28 +629,28 @@ func @mul_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<3
// CHECK: scf.yield %[[VAL_94]], %[[VAL_97]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_98:.*]] = %[[VAL_99:.*]]#0 to %[[VAL_18]] step %[[VAL_4]] {
-// CHECK: %[[VAL_100:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_98]]] : memref<?xindex>
-// CHECK: %[[VAL_101:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_98]]] : memref<?xindex>
+// CHECK: %[[VAL_100:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_98]]] : memref<?xindex>
+// CHECK: %[[VAL_101:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_98]]] : memref<?xindex>
// CHECK: %[[VAL_102:.*]] = addi %[[VAL_98]], %[[VAL_4]] : index
-// CHECK: %[[VAL_103:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_102]]] : memref<?xindex>
+// CHECK: %[[VAL_103:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_102]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_104:.*]] = %[[VAL_101]] to %[[VAL_103]] step %[[VAL_4]] {
-// CHECK: %[[VAL_105:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_104]]] : memref<?xindex>
-// CHECK: %[[VAL_106:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_104]]] : memref<?xf32>
+// CHECK: %[[VAL_105:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_104]]] : memref<?xindex>
+// CHECK: %[[VAL_106:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_104]]] : memref<?xf32>
// CHECK: store %[[VAL_106]], %[[VAL_16]]{{\[}}%[[VAL_100]], %[[VAL_105]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
// CHECK: scf.for %[[VAL_107:.*]] = %[[VAL_108:.*]]#1 to %[[VAL_20]] step %[[VAL_4]] {
-// CHECK: %[[VAL_109:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_107]]] : memref<?xindex>
-// CHECK: %[[VAL_110:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_107]]] : memref<?xindex>
+// CHECK: %[[VAL_109:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_107]]] : memref<?xindex>
+// CHECK: %[[VAL_110:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_107]]] : memref<?xindex>
// CHECK: %[[VAL_111:.*]] = addi %[[VAL_107]], %[[VAL_4]] : index
-// CHECK: %[[VAL_112:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_111]]] : memref<?xindex>
+// CHECK: %[[VAL_112:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_111]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_113:.*]] = %[[VAL_110]] to %[[VAL_112]] step %[[VAL_4]] {
-// CHECK: %[[VAL_114:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_113]]] : memref<?xindex>
-// CHECK: %[[VAL_115:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_113]]] : memref<?xf32>
+// CHECK: %[[VAL_114:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_113]]] : memref<?xindex>
+// CHECK: %[[VAL_115:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_113]]] : memref<?xf32>
// CHECK: store %[[VAL_115]], %[[VAL_16]]{{\[}}%[[VAL_109]], %[[VAL_114]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_116:.*]] = tensor_load %[[VAL_16]] : memref<32x16xf32>
+// CHECK: %[[VAL_116:.*]] = memref.tensor_load %[[VAL_16]] : memref<32x16xf32>
// CHECK: return %[[VAL_116]] : tensor<32x16xf32>
// CHECK: }
func @add_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -680,13 +680,13 @@ func @add_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]]:2 = scf.while (%[[VAL_22:.*]] = %[[VAL_17]], %[[VAL_23:.*]] = %[[VAL_19]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_24:.*]] = cmpi ult, %[[VAL_22]], %[[VAL_18]] : index
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_20]] : index
@@ -694,20 +694,20 @@ func @add_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_26]]) %[[VAL_22]], %[[VAL_23]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_27:.*]]: index, %[[VAL_28:.*]]: index):
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = cmpi ult, %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = select %[[VAL_31]], %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_33:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: %[[VAL_34:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: %[[VAL_35:.*]] = and %[[VAL_33]], %[[VAL_34]] : i1
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
// CHECK: %[[VAL_42:.*]]:2 = scf.while (%[[VAL_43:.*]] = %[[VAL_36]], %[[VAL_44:.*]] = %[[VAL_39]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_45:.*]] = cmpi ult, %[[VAL_43]], %[[VAL_38]] : index
// CHECK: %[[VAL_46:.*]] = cmpi ult, %[[VAL_44]], %[[VAL_41]] : index
@@ -715,16 +715,16 @@ func @add_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_47]]) %[[VAL_43]], %[[VAL_44]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_48:.*]]: index, %[[VAL_49:.*]]: index):
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
// CHECK: %[[VAL_52:.*]] = cmpi ult, %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_53:.*]] = select %[[VAL_52]], %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_54:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: %[[VAL_55:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: %[[VAL_56:.*]] = and %[[VAL_54]], %[[VAL_55]] : i1
// CHECK: scf.if %[[VAL_56]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: %[[VAL_59:.*]] = mulf %[[VAL_57]], %[[VAL_58]] : f32
// CHECK: store %[[VAL_59]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
@@ -747,7 +747,7 @@ func @add_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_71:.*]] = select %[[VAL_69]], %[[VAL_70]], %[[VAL_28]] : index
// CHECK: scf.yield %[[VAL_68]], %[[VAL_71]] : index, index
// CHECK: }
-// CHECK: %[[VAL_72:.*]] = tensor_load %[[VAL_16]] : memref<32x16xf32>
+// CHECK: %[[VAL_72:.*]] = memref.tensor_load %[[VAL_16]] : memref<32x16xf32>
// CHECK: return %[[VAL_72]] : tensor<32x16xf32>
// CHECK: }
func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -792,13 +792,13 @@ func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]]:2 = scf.while (%[[VAL_22:.*]] = %[[VAL_17]], %[[VAL_23:.*]] = %[[VAL_19]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_24:.*]] = cmpi ult, %[[VAL_22]], %[[VAL_18]] : index
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_20]] : index
@@ -806,20 +806,20 @@ func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_26]]) %[[VAL_22]], %[[VAL_23]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_27:.*]]: index, %[[VAL_28:.*]]: index):
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = cmpi ult, %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = select %[[VAL_31]], %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_33:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: %[[VAL_34:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: %[[VAL_35:.*]] = and %[[VAL_33]], %[[VAL_34]] : i1
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
// CHECK: %[[VAL_42:.*]]:2 = scf.while (%[[VAL_43:.*]] = %[[VAL_36]], %[[VAL_44:.*]] = %[[VAL_39]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_45:.*]] = cmpi ult, %[[VAL_43]], %[[VAL_38]] : index
// CHECK: %[[VAL_46:.*]] = cmpi ult, %[[VAL_44]], %[[VAL_41]] : index
@@ -827,27 +827,27 @@ func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_47]]) %[[VAL_43]], %[[VAL_44]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_48:.*]]: index, %[[VAL_49:.*]]: index):
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
// CHECK: %[[VAL_52:.*]] = cmpi ult, %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_53:.*]] = select %[[VAL_52]], %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_54:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: %[[VAL_55:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: %[[VAL_56:.*]] = and %[[VAL_54]], %[[VAL_55]] : i1
// CHECK: scf.if %[[VAL_56]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: %[[VAL_59:.*]] = addf %[[VAL_57]], %[[VAL_58]] : f32
// CHECK: store %[[VAL_59]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: %[[VAL_60:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: scf.if %[[VAL_60]] {
-// CHECK: %[[VAL_61:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_61:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
// CHECK: store %[[VAL_61]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: %[[VAL_62:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: scf.if %[[VAL_62]] {
-// CHECK: %[[VAL_63:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_63:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: store %[[VAL_63]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
// CHECK: }
@@ -862,35 +862,35 @@ func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.yield %[[VAL_66]], %[[VAL_69]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_70:.*]] = %[[VAL_71:.*]]#0 to %[[VAL_38]] step %[[VAL_4]] {
-// CHECK: %[[VAL_72:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_70]]] : memref<?xindex>
-// CHECK: %[[VAL_73:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_70]]] : memref<?xf32>
+// CHECK: %[[VAL_72:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_70]]] : memref<?xindex>
+// CHECK: %[[VAL_73:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_70]]] : memref<?xf32>
// CHECK: store %[[VAL_73]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_72]]] : memref<32x16xf32>
// CHECK: }
// CHECK: scf.for %[[VAL_74:.*]] = %[[VAL_75:.*]]#1 to %[[VAL_41]] step %[[VAL_4]] {
-// CHECK: %[[VAL_76:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_74]]] : memref<?xindex>
-// CHECK: %[[VAL_77:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_74]]] : memref<?xf32>
+// CHECK: %[[VAL_76:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_74]]] : memref<?xindex>
+// CHECK: %[[VAL_77:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_74]]] : memref<?xf32>
// CHECK: store %[[VAL_77]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_76]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: %[[VAL_78:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_78]] {
-// CHECK: %[[VAL_79:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_79:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_80:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_81:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_80]]] : memref<?xindex>
+// CHECK: %[[VAL_81:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_80]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_82:.*]] = %[[VAL_79]] to %[[VAL_81]] step %[[VAL_4]] {
-// CHECK: %[[VAL_83:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_82]]] : memref<?xindex>
-// CHECK: %[[VAL_84:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_82]]] : memref<?xf32>
+// CHECK: %[[VAL_83:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_82]]] : memref<?xindex>
+// CHECK: %[[VAL_84:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_82]]] : memref<?xf32>
// CHECK: store %[[VAL_84]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_83]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
// CHECK: %[[VAL_85:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: scf.if %[[VAL_85]] {
-// CHECK: %[[VAL_86:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_86:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_87:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_88:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_87]]] : memref<?xindex>
+// CHECK: %[[VAL_88:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_87]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_89:.*]] = %[[VAL_86]] to %[[VAL_88]] step %[[VAL_4]] {
-// CHECK: %[[VAL_90:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_89]]] : memref<?xindex>
-// CHECK: %[[VAL_91:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_89]]] : memref<?xf32>
+// CHECK: %[[VAL_90:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_89]]] : memref<?xindex>
+// CHECK: %[[VAL_91:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_89]]] : memref<?xf32>
// CHECK: store %[[VAL_91]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_90]]] : memref<32x16xf32>
// CHECK: }
// CHECK: } else {
@@ -906,28 +906,28 @@ func @mul_ss_ss(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.yield %[[VAL_94]], %[[VAL_97]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_98:.*]] = %[[VAL_99:.*]]#0 to %[[VAL_18]] step %[[VAL_4]] {
-// CHECK: %[[VAL_100:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_98]]] : memref<?xindex>
-// CHECK: %[[VAL_101:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_98]]] : memref<?xindex>
+// CHECK: %[[VAL_100:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_98]]] : memref<?xindex>
+// CHECK: %[[VAL_101:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_98]]] : memref<?xindex>
// CHECK: %[[VAL_102:.*]] = addi %[[VAL_98]], %[[VAL_4]] : index
-// CHECK: %[[VAL_103:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_102]]] : memref<?xindex>
+// CHECK: %[[VAL_103:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_102]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_104:.*]] = %[[VAL_101]] to %[[VAL_103]] step %[[VAL_4]] {
-// CHECK: %[[VAL_105:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_104]]] : memref<?xindex>
-// CHECK: %[[VAL_106:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_104]]] : memref<?xf32>
+// CHECK: %[[VAL_105:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_104]]] : memref<?xindex>
+// CHECK: %[[VAL_106:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_104]]] : memref<?xf32>
// CHECK: store %[[VAL_106]], %[[VAL_16]]{{\[}}%[[VAL_100]], %[[VAL_105]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
// CHECK: scf.for %[[VAL_107:.*]] = %[[VAL_108:.*]]#1 to %[[VAL_20]] step %[[VAL_4]] {
-// CHECK: %[[VAL_109:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_107]]] : memref<?xindex>
-// CHECK: %[[VAL_110:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_107]]] : memref<?xindex>
+// CHECK: %[[VAL_109:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_107]]] : memref<?xindex>
+// CHECK: %[[VAL_110:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_107]]] : memref<?xindex>
// CHECK: %[[VAL_111:.*]] = addi %[[VAL_107]], %[[VAL_4]] : index
-// CHECK: %[[VAL_112:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_111]]] : memref<?xindex>
+// CHECK: %[[VAL_112:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_111]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_113:.*]] = %[[VAL_110]] to %[[VAL_112]] step %[[VAL_4]] {
-// CHECK: %[[VAL_114:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_113]]] : memref<?xindex>
-// CHECK: %[[VAL_115:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_113]]] : memref<?xf32>
+// CHECK: %[[VAL_114:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_113]]] : memref<?xindex>
+// CHECK: %[[VAL_115:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_113]]] : memref<?xf32>
// CHECK: store %[[VAL_115]], %[[VAL_16]]{{\[}}%[[VAL_109]], %[[VAL_114]]] : memref<32x16xf32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_116:.*]] = tensor_load %[[VAL_16]] : memref<32x16xf32>
+// CHECK: %[[VAL_116:.*]] = memref.tensor_load %[[VAL_16]] : memref<32x16xf32>
// CHECK: return %[[VAL_116]] : tensor<32x16xf32>
// CHECK: }
func @add_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -957,13 +957,13 @@ func @add_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<32x16xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<32x16xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16xf32>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<32x16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<32x16xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<32x16xf32>, memref<32x16xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]]:2 = scf.while (%[[VAL_22:.*]] = %[[VAL_17]], %[[VAL_23:.*]] = %[[VAL_19]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_24:.*]] = cmpi ult, %[[VAL_22]], %[[VAL_18]] : index
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_20]] : index
@@ -971,20 +971,20 @@ func @add_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_26]]) %[[VAL_22]], %[[VAL_23]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_27:.*]]: index, %[[VAL_28:.*]]: index):
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = cmpi ult, %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_32:.*]] = select %[[VAL_31]], %[[VAL_30]], %[[VAL_29]] : index
// CHECK: %[[VAL_33:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_32]] : index
// CHECK: %[[VAL_34:.*]] = cmpi eq, %[[VAL_30]], %[[VAL_32]] : index
// CHECK: %[[VAL_35:.*]] = and %[[VAL_33]], %[[VAL_34]] : i1
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = addi %[[VAL_27]], %[[VAL_4]] : index
-// CHECK: %[[VAL_38:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_38:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_28]], %[[VAL_4]] : index
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_40]]] : memref<?xindex>
// CHECK: %[[VAL_42:.*]]:2 = scf.while (%[[VAL_43:.*]] = %[[VAL_36]], %[[VAL_44:.*]] = %[[VAL_39]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_45:.*]] = cmpi ult, %[[VAL_43]], %[[VAL_38]] : index
// CHECK: %[[VAL_46:.*]] = cmpi ult, %[[VAL_44]], %[[VAL_41]] : index
@@ -992,16 +992,16 @@ func @add_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: scf.condition(%[[VAL_47]]) %[[VAL_43]], %[[VAL_44]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_48:.*]]: index, %[[VAL_49:.*]]: index):
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_48]]] : memref<?xindex>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_49]]] : memref<?xindex>
// CHECK: %[[VAL_52:.*]] = cmpi ult, %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_53:.*]] = select %[[VAL_52]], %[[VAL_51]], %[[VAL_50]] : index
// CHECK: %[[VAL_54:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_53]] : index
// CHECK: %[[VAL_55:.*]] = cmpi eq, %[[VAL_51]], %[[VAL_53]] : index
// CHECK: %[[VAL_56:.*]] = and %[[VAL_54]], %[[VAL_55]] : i1
// CHECK: scf.if %[[VAL_56]] {
-// CHECK: %[[VAL_57:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
+// CHECK: %[[VAL_57:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_49]]] : memref<?xf32>
// CHECK: %[[VAL_59:.*]] = mulf %[[VAL_57]], %[[VAL_58]] : f32
// CHECK: store %[[VAL_59]], %[[VAL_16]]{{\[}}%[[VAL_32]], %[[VAL_53]]] : memref<32x16xf32>
// CHECK: } else {
@@ -1024,7 +1024,7 @@ func @add_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_71:.*]] = select %[[VAL_69]], %[[VAL_70]], %[[VAL_28]] : index
// CHECK: scf.yield %[[VAL_68]], %[[VAL_71]] : index, index
// CHECK: }
-// CHECK: %[[VAL_72:.*]] = tensor_load %[[VAL_16]] : memref<32x16xf32>
+// CHECK: %[[VAL_72:.*]] = memref.tensor_load %[[VAL_16]] : memref<32x16xf32>
// CHECK: return %[[VAL_72]] : tensor<32x16xf32>
// CHECK: }
func @mul_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tensor<32x16xf32>) -> tensor<32x16xf32> {
@@ -1063,26 +1063,26 @@ func @mul_sd_ds(%arga: tensor<32x16xf32>, %argb: tensor<32x16xf32>, %argx: tenso
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_5]] : tensor<16x32xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_5]] : tensor<16x32xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<16x32xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<32xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<16xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<16xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<16xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<16xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<16xf32>, memref<16xf32>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
// CHECK: %[[VAL_14:.*]] = addi %[[VAL_12]], %[[VAL_5]] : index
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_12]]] : memref<16xf32>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_12]]] : memref<16xf32>
// CHECK: %[[VAL_17:.*]] = scf.for %[[VAL_18:.*]] = %[[VAL_13]] to %[[VAL_15]] step %[[VAL_5]] iter_args(%[[VAL_19:.*]] = %[[VAL_16]]) -> (f32) {
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf32>
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<32xf32>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf32>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_20]]] : memref<32xf32>
// CHECK: %[[VAL_23:.*]] = mulf %[[VAL_21]], %[[VAL_22]] : f32
// CHECK: %[[VAL_24:.*]] = addf %[[VAL_23]], %[[VAL_19]] : f32
// CHECK: scf.yield %[[VAL_24]] : f32
// CHECK: }
// CHECK: store %[[VAL_25:.*]], %[[VAL_11]]{{\[}}%[[VAL_12]]] : memref<16xf32>
// CHECK: }
-// CHECK: %[[VAL_26:.*]] = tensor_load %[[VAL_11]] : memref<16xf32>
+// CHECK: %[[VAL_26:.*]] = memref.tensor_load %[[VAL_11]] : memref<16xf32>
// CHECK: return %[[VAL_26]] : tensor<16xf32>
// CHECK: }
func @matvec(%argA: tensor<16x32xf32>, %argb: tensor<32xf32>, %argx: tensor<16xf32>) -> tensor<16xf32> {
@@ -1118,22 +1118,22 @@ func @matvec(%argA: tensor<16x32xf32>, %argb: tensor<32xf32>, %argx: tensor<16xf
// CHECK: %[[VAL_4:.*]] = constant 1 : index
// CHECK: %[[VAL_5:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<10x20xf32> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<10x20xf32> to memref<?xf32>
-// CHECK: %[[VAL_7:.*]] = tensor_to_memref %[[VAL_1]] : memref<f32>
-// CHECK: %[[VAL_8:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_7:.*]] = memref.buffer_cast %[[VAL_1]] : memref<f32>
+// CHECK: %[[VAL_8:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_7]], %[[VAL_8]]) : memref<f32>, memref<f32>
// CHECK: scf.for %[[VAL_9:.*]] = %[[VAL_3]] to %[[VAL_2]] step %[[VAL_4]] {
-// CHECK: %[[VAL_10:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_9]]] : memref<?xindex>
+// CHECK: %[[VAL_10:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_9]]] : memref<?xindex>
// CHECK: %[[VAL_11:.*]] = addi %[[VAL_9]], %[[VAL_4]] : index
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_11]]] : memref<?xindex>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_8]][] : memref<f32>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_11]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_8]][] : memref<f32>
// CHECK: %[[VAL_14:.*]] = scf.for %[[VAL_15:.*]] = %[[VAL_10]] to %[[VAL_12]] step %[[VAL_4]] iter_args(%[[VAL_16:.*]] = %[[VAL_13]]) -> (f32) {
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xf32>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xf32>
// CHECK: %[[VAL_18:.*]] = addf %[[VAL_16]], %[[VAL_17]] : f32
// CHECK: scf.yield %[[VAL_18]] : f32
// CHECK: }
// CHECK: store %[[VAL_19:.*]], %[[VAL_8]][] : memref<f32>
// CHECK: }
-// CHECK: %[[VAL_20:.*]] = tensor_load %[[VAL_8]] : memref<f32>
+// CHECK: %[[VAL_20:.*]] = memref.tensor_load %[[VAL_8]] : memref<f32>
// CHECK: return %[[VAL_20]] : tensor<f32>
// CHECK: }
func @sum_reduction(%arga: tensor<10x20xf32>, %argx: tensor<f32>) -> tensor<f32> {
@@ -1169,23 +1169,23 @@ func @sum_reduction(%arga: tensor<10x20xf32>, %argx: tensor<f32>) -> tensor<f32>
// CHECK: %[[VAL_5:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<?x?xf64> to memref<?xindex>
// CHECK: %[[VAL_6:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_4]] : tensor<?x?xf64> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<?x?xf64> to memref<?xf64>
-// CHECK: %[[VAL_8:.*]] = dim %[[VAL_1]], %[[VAL_3]] : tensor<?x?xf64>
-// CHECK: %[[VAL_9:.*]] = dim %[[VAL_1]], %[[VAL_4]] : tensor<?x?xf64>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<?x?xf64>
-// CHECK: %[[VAL_11:.*]] = alloc(%[[VAL_8]], %[[VAL_9]]) : memref<?x?xf64>
+// CHECK: %[[VAL_8:.*]] = memref.dim %[[VAL_1]], %[[VAL_3]] : tensor<?x?xf64>
+// CHECK: %[[VAL_9:.*]] = memref.dim %[[VAL_1]], %[[VAL_4]] : tensor<?x?xf64>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<?x?xf64>
+// CHECK: %[[VAL_11:.*]] = memref.alloc(%[[VAL_8]], %[[VAL_9]]) : memref<?x?xf64>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<?x?xf64>, memref<?x?xf64>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_3]] to %[[VAL_8]] step %[[VAL_4]] {
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_12]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_12]]] : memref<?xindex>
// CHECK: %[[VAL_14:.*]] = addi %[[VAL_12]], %[[VAL_4]] : index
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_14]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_13]] to %[[VAL_15]] step %[[VAL_4]] {
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_16]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xf64>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xf64>
// CHECK: %[[VAL_19:.*]] = mulf %[[VAL_18]], %[[VAL_2]] : f64
// CHECK: store %[[VAL_19]], %[[VAL_11]]{{\[}}%[[VAL_12]], %[[VAL_17]]] : memref<?x?xf64>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_20:.*]] = tensor_load %[[VAL_11]] : memref<?x?xf64>
+// CHECK: %[[VAL_20:.*]] = memref.tensor_load %[[VAL_11]] : memref<?x?xf64>
// CHECK: return %[[VAL_20]] : tensor<?x?xf64>
// CHECK: }
func @scale(%arga: tensor<?x?xf64>, %argx: tensor<?x?xf64>) -> tensor<?x?xf64> {
@@ -1229,28 +1229,28 @@ func @scale(%arga: tensor<?x?xf64>, %argx: tensor<?x?xf64>) -> tensor<?x?xf64> {
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_5]] : tensor<?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_5]] : tensor<?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<?x?xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_1]] : memref<?x?xf32>
-// CHECK: %[[VAL_12:.*]] = dim %[[VAL_2]], %[[VAL_4]] : tensor<?x?xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_2]] : memref<?x?xf32>
-// CHECK: %[[VAL_14:.*]] = dim %[[VAL_3]], %[[VAL_4]] : tensor<?x?xf32>
-// CHECK: %[[VAL_15:.*]] = dim %[[VAL_3]], %[[VAL_5]] : tensor<?x?xf32>
-// CHECK: %[[VAL_16:.*]] = tensor_to_memref %[[VAL_3]] : memref<?x?xf32>
-// CHECK: %[[VAL_17:.*]] = alloc(%[[VAL_14]], %[[VAL_15]]) : memref<?x?xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_1]] : memref<?x?xf32>
+// CHECK: %[[VAL_12:.*]] = memref.dim %[[VAL_2]], %[[VAL_4]] : tensor<?x?xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_2]] : memref<?x?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.dim %[[VAL_3]], %[[VAL_4]] : tensor<?x?xf32>
+// CHECK: %[[VAL_15:.*]] = memref.dim %[[VAL_3]], %[[VAL_5]] : tensor<?x?xf32>
+// CHECK: %[[VAL_16:.*]] = memref.buffer_cast %[[VAL_3]] : memref<?x?xf32>
+// CHECK: %[[VAL_17:.*]] = memref.alloc(%[[VAL_14]], %[[VAL_15]]) : memref<?x?xf32>
// CHECK: linalg.copy(%[[VAL_16]], %[[VAL_17]]) : memref<?x?xf32>, memref<?x?xf32>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_20:.*]] = %[[VAL_18]] to %[[VAL_19]] step %[[VAL_5]] {
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_22:.*]] = %[[VAL_4]] to %[[VAL_12]] step %[[VAL_5]] {
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_22]]] : memref<?x?xf32>
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_21]], %[[VAL_22]]] : memref<?x?xf32>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xindex>
// CHECK: %[[VAL_25:.*]] = addi %[[VAL_20]], %[[VAL_5]] : index
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_25]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_25]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_27:.*]] = %[[VAL_24]] to %[[VAL_26]] step %[[VAL_5]] {
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_27]]] : memref<?xindex>
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_21]], %[[VAL_28]]] : memref<?x?xf32>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_27]]] : memref<?xf32>
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_22]], %[[VAL_28]]] : memref<?x?xf32>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_21]], %[[VAL_28]]] : memref<?x?xf32>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_27]]] : memref<?xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_22]], %[[VAL_28]]] : memref<?x?xf32>
// CHECK: %[[VAL_32:.*]] = mulf %[[VAL_23]], %[[VAL_31]] : f32
// CHECK: %[[VAL_33:.*]] = mulf %[[VAL_30]], %[[VAL_32]] : f32
// CHECK: %[[VAL_34:.*]] = addf %[[VAL_29]], %[[VAL_33]] : f32
@@ -1258,7 +1258,7 @@ func @scale(%arga: tensor<?x?xf64>, %argx: tensor<?x?xf64>) -> tensor<?x?xf64> {
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_35:.*]] = tensor_load %[[VAL_17]] : memref<?x?xf32>
+// CHECK: %[[VAL_35:.*]] = memref.tensor_load %[[VAL_17]] : memref<?x?xf32>
// CHECK: return %[[VAL_35]] : tensor<?x?xf32>
// CHECK: }
func @sampled_dense_dense(%args: tensor<?x?xf32>,
@@ -1319,33 +1319,33 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: %[[VAL_17:.*]] = linalg.sparse_pointers %[[VAL_2]], %[[VAL_8]] : tensor<?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_18:.*]] = linalg.sparse_indices %[[VAL_2]], %[[VAL_8]] : tensor<?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_19:.*]] = linalg.sparse_values %[[VAL_2]] : tensor<?x?xf32> to memref<?xf32>
-// CHECK: %[[VAL_20:.*]] = tensor_to_memref %[[VAL_3]] : memref<?xf32>
-// CHECK: %[[VAL_21:.*]] = tensor_to_memref %[[VAL_4]] : memref<f32>
-// CHECK: %[[VAL_22:.*]] = dim %[[VAL_5]], %[[VAL_6]] : tensor<?xf32>
-// CHECK: %[[VAL_23:.*]] = tensor_to_memref %[[VAL_5]] : memref<?xf32>
-// CHECK: %[[VAL_24:.*]] = alloc(%[[VAL_22]]) : memref<?xf32>
+// CHECK: %[[VAL_20:.*]] = memref.buffer_cast %[[VAL_3]] : memref<?xf32>
+// CHECK: %[[VAL_21:.*]] = memref.buffer_cast %[[VAL_4]] : memref<f32>
+// CHECK: %[[VAL_22:.*]] = memref.dim %[[VAL_5]], %[[VAL_6]] : tensor<?xf32>
+// CHECK: %[[VAL_23:.*]] = memref.buffer_cast %[[VAL_5]] : memref<?xf32>
+// CHECK: %[[VAL_24:.*]] = memref.alloc(%[[VAL_22]]) : memref<?xf32>
// CHECK: linalg.copy(%[[VAL_23]], %[[VAL_24]]) : memref<?xf32>, memref<?xf32>
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_21]][] : memref<f32>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_6]]] : memref<?xindex>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_21]][] : memref<f32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
// CHECK: %[[VAL_28:.*]]:2 = scf.while (%[[VAL_29:.*]] = %[[VAL_26]], %[[VAL_30:.*]] = %[[VAL_6]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_31:.*]] = cmpi ult, %[[VAL_29]], %[[VAL_27]] : index
// CHECK: scf.condition(%[[VAL_31]]) %[[VAL_29]], %[[VAL_30]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_32:.*]]: index, %[[VAL_33:.*]]: index):
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_32]]] : memref<?xindex>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_32]]] : memref<?xindex>
// CHECK: %[[VAL_35:.*]] = cmpi eq, %[[VAL_34]], %[[VAL_33]] : index
// CHECK: scf.if %[[VAL_35]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_20]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_37:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_32]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_20]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_37:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_32]]] : memref<?xindex>
// CHECK: %[[VAL_38:.*]] = addi %[[VAL_32]], %[[VAL_8]] : index
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_38]]] : memref<?xindex>
-// CHECK: %[[VAL_40:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_33]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_38]]] : memref<?xindex>
+// CHECK: %[[VAL_40:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_33]]] : memref<?xindex>
// CHECK: %[[VAL_41:.*]] = addi %[[VAL_33]], %[[VAL_8]] : index
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_41]]] : memref<?xindex>
-// CHECK: %[[VAL_43:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_33]]] : memref<?xindex>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_41]]] : memref<?xindex>
+// CHECK: %[[VAL_43:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_33]]] : memref<?xindex>
// CHECK: %[[VAL_44:.*]] = addi %[[VAL_33]], %[[VAL_8]] : index
-// CHECK: %[[VAL_45:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_44]]] : memref<?xindex>
+// CHECK: %[[VAL_45:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_44]]] : memref<?xindex>
// CHECK: %[[VAL_46:.*]]:3 = scf.while (%[[VAL_47:.*]] = %[[VAL_37]], %[[VAL_48:.*]] = %[[VAL_40]], %[[VAL_49:.*]] = %[[VAL_43]]) : (index, index, index) -> (index, index, index) {
// CHECK: %[[VAL_50:.*]] = cmpi ult, %[[VAL_47]], %[[VAL_39]] : index
// CHECK: %[[VAL_51:.*]] = cmpi ult, %[[VAL_48]], %[[VAL_42]] : index
@@ -1355,11 +1355,11 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: scf.condition(%[[VAL_54]]) %[[VAL_47]], %[[VAL_48]], %[[VAL_49]] : index, index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_55:.*]]: index, %[[VAL_56:.*]]: index, %[[VAL_57:.*]]: index):
-// CHECK: %[[VAL_58:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_55]]] : memref<?xindex>
-// CHECK: %[[VAL_59:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_56]]] : memref<?xindex>
+// CHECK: %[[VAL_58:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_55]]] : memref<?xindex>
+// CHECK: %[[VAL_59:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_56]]] : memref<?xindex>
// CHECK: %[[VAL_60:.*]] = cmpi ult, %[[VAL_59]], %[[VAL_58]] : index
// CHECK: %[[VAL_61:.*]] = select %[[VAL_60]], %[[VAL_59]], %[[VAL_58]] : index
-// CHECK: %[[VAL_62:.*]] = load %[[VAL_18]]{{\[}}%[[VAL_57]]] : memref<?xindex>
+// CHECK: %[[VAL_62:.*]] = memref.load %[[VAL_18]]{{\[}}%[[VAL_57]]] : memref<?xindex>
// CHECK: %[[VAL_63:.*]] = cmpi ult, %[[VAL_62]], %[[VAL_61]] : index
// CHECK: %[[VAL_64:.*]] = select %[[VAL_63]], %[[VAL_62]], %[[VAL_61]] : index
// CHECK: %[[VAL_65:.*]] = cmpi eq, %[[VAL_58]], %[[VAL_64]] : index
@@ -1368,13 +1368,13 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: %[[VAL_68:.*]] = cmpi eq, %[[VAL_62]], %[[VAL_64]] : index
// CHECK: %[[VAL_69:.*]] = and %[[VAL_67]], %[[VAL_68]] : i1
// CHECK: scf.if %[[VAL_69]] {
-// CHECK: %[[VAL_70:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_71:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_55]]] : memref<?xf32>
-// CHECK: %[[VAL_72:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_56]]] : memref<?xf32>
+// CHECK: %[[VAL_70:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_71:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_55]]] : memref<?xf32>
+// CHECK: %[[VAL_72:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_56]]] : memref<?xf32>
// CHECK: %[[VAL_73:.*]] = mulf %[[VAL_71]], %[[VAL_72]] : f32
// CHECK: %[[VAL_74:.*]] = mulf %[[VAL_73]], %[[VAL_36]] : f32
// CHECK: %[[VAL_75:.*]] = mulf %[[VAL_74]], %[[VAL_25]] : f32
-// CHECK: %[[VAL_76:.*]] = load %[[VAL_19]]{{\[}}%[[VAL_57]]] : memref<?xf32>
+// CHECK: %[[VAL_76:.*]] = memref.load %[[VAL_19]]{{\[}}%[[VAL_57]]] : memref<?xf32>
// CHECK: %[[VAL_77:.*]] = addf %[[VAL_75]], %[[VAL_76]] : f32
// CHECK: %[[VAL_78:.*]] = addf %[[VAL_70]], %[[VAL_77]] : f32
// CHECK: store %[[VAL_78]], %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
@@ -1383,9 +1383,9 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: %[[VAL_80:.*]] = cmpi eq, %[[VAL_59]], %[[VAL_64]] : index
// CHECK: %[[VAL_81:.*]] = and %[[VAL_79]], %[[VAL_80]] : i1
// CHECK: scf.if %[[VAL_81]] {
-// CHECK: %[[VAL_82:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_83:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_55]]] : memref<?xf32>
-// CHECK: %[[VAL_84:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_56]]] : memref<?xf32>
+// CHECK: %[[VAL_82:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_83:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_55]]] : memref<?xf32>
+// CHECK: %[[VAL_84:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_56]]] : memref<?xf32>
// CHECK: %[[VAL_85:.*]] = mulf %[[VAL_83]], %[[VAL_84]] : f32
// CHECK: %[[VAL_86:.*]] = mulf %[[VAL_85]], %[[VAL_36]] : f32
// CHECK: %[[VAL_87:.*]] = mulf %[[VAL_86]], %[[VAL_25]] : f32
@@ -1394,8 +1394,8 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: } else {
// CHECK: %[[VAL_89:.*]] = cmpi eq, %[[VAL_62]], %[[VAL_64]] : index
// CHECK: scf.if %[[VAL_89]] {
-// CHECK: %[[VAL_90:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_91:.*]] = load %[[VAL_19]]{{\[}}%[[VAL_57]]] : memref<?xf32>
+// CHECK: %[[VAL_90:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_91:.*]] = memref.load %[[VAL_19]]{{\[}}%[[VAL_57]]] : memref<?xf32>
// CHECK: %[[VAL_92:.*]] = addf %[[VAL_90]], %[[VAL_91]] : f32
// CHECK: store %[[VAL_92]], %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
// CHECK: } else {
@@ -1420,17 +1420,17 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: scf.condition(%[[VAL_108]]) %[[VAL_103]], %[[VAL_105]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_109:.*]]: index, %[[VAL_110:.*]]: index):
-// CHECK: %[[VAL_111:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_109]]] : memref<?xindex>
-// CHECK: %[[VAL_112:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_110]]] : memref<?xindex>
+// CHECK: %[[VAL_111:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_109]]] : memref<?xindex>
+// CHECK: %[[VAL_112:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_110]]] : memref<?xindex>
// CHECK: %[[VAL_113:.*]] = cmpi ult, %[[VAL_112]], %[[VAL_111]] : index
// CHECK: %[[VAL_114:.*]] = select %[[VAL_113]], %[[VAL_112]], %[[VAL_111]] : index
// CHECK: %[[VAL_115:.*]] = cmpi eq, %[[VAL_111]], %[[VAL_114]] : index
// CHECK: %[[VAL_116:.*]] = cmpi eq, %[[VAL_112]], %[[VAL_114]] : index
// CHECK: %[[VAL_117:.*]] = and %[[VAL_115]], %[[VAL_116]] : i1
// CHECK: scf.if %[[VAL_117]] {
-// CHECK: %[[VAL_118:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
-// CHECK: %[[VAL_119:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_109]]] : memref<?xf32>
-// CHECK: %[[VAL_120:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_110]]] : memref<?xf32>
+// CHECK: %[[VAL_118:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_119:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_109]]] : memref<?xf32>
+// CHECK: %[[VAL_120:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_110]]] : memref<?xf32>
// CHECK: %[[VAL_121:.*]] = mulf %[[VAL_119]], %[[VAL_120]] : f32
// CHECK: %[[VAL_122:.*]] = mulf %[[VAL_121]], %[[VAL_36]] : f32
// CHECK: %[[VAL_123:.*]] = mulf %[[VAL_122]], %[[VAL_25]] : f32
@@ -1446,21 +1446,21 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: %[[VAL_130:.*]] = select %[[VAL_128]], %[[VAL_129]], %[[VAL_110]] : index
// CHECK: scf.yield %[[VAL_127]], %[[VAL_130]] : index, index
// CHECK: }
-// CHECK: %[[VAL_131:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_131:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
// CHECK: %[[VAL_132:.*]] = scf.for %[[VAL_133:.*]] = %[[VAL_134:.*]]#2 to %[[VAL_45]] step %[[VAL_8]] iter_args(%[[VAL_135:.*]] = %[[VAL_131]]) -> (f32) {
-// CHECK: %[[VAL_136:.*]] = load %[[VAL_19]]{{\[}}%[[VAL_133]]] : memref<?xf32>
+// CHECK: %[[VAL_136:.*]] = memref.load %[[VAL_19]]{{\[}}%[[VAL_133]]] : memref<?xf32>
// CHECK: %[[VAL_137:.*]] = addf %[[VAL_135]], %[[VAL_136]] : f32
// CHECK: scf.yield %[[VAL_137]] : f32
// CHECK: }
// CHECK: store %[[VAL_138:.*]], %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
-// CHECK: %[[VAL_139:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_33]]] : memref<?xindex>
+// CHECK: %[[VAL_139:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_33]]] : memref<?xindex>
// CHECK: %[[VAL_140:.*]] = addi %[[VAL_33]], %[[VAL_8]] : index
-// CHECK: %[[VAL_141:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_140]]] : memref<?xindex>
-// CHECK: %[[VAL_142:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
+// CHECK: %[[VAL_141:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_140]]] : memref<?xindex>
+// CHECK: %[[VAL_142:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_33]]] : memref<?xf32>
// CHECK: %[[VAL_143:.*]] = scf.for %[[VAL_144:.*]] = %[[VAL_139]] to %[[VAL_141]] step %[[VAL_8]] iter_args(%[[VAL_145:.*]] = %[[VAL_142]]) -> (f32) {
-// CHECK: %[[VAL_146:.*]] = load %[[VAL_19]]{{\[}}%[[VAL_144]]] : memref<?xf32>
+// CHECK: %[[VAL_146:.*]] = memref.load %[[VAL_19]]{{\[}}%[[VAL_144]]] : memref<?xf32>
// CHECK: %[[VAL_147:.*]] = addf %[[VAL_145]], %[[VAL_146]] : f32
// CHECK: scf.yield %[[VAL_147]] : f32
// CHECK: }
@@ -1475,18 +1475,18 @@ func @sampled_dense_dense(%args: tensor<?x?xf32>,
// CHECK: scf.yield %[[VAL_151]], %[[VAL_152]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_153:.*]] = %[[VAL_154:.*]]#1 to %[[VAL_22]] step %[[VAL_8]] {
-// CHECK: %[[VAL_155:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_153]]] : memref<?xindex>
+// CHECK: %[[VAL_155:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_153]]] : memref<?xindex>
// CHECK: %[[VAL_156:.*]] = addi %[[VAL_153]], %[[VAL_8]] : index
-// CHECK: %[[VAL_157:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_156]]] : memref<?xindex>
-// CHECK: %[[VAL_158:.*]] = load %[[VAL_24]]{{\[}}%[[VAL_153]]] : memref<?xf32>
+// CHECK: %[[VAL_157:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_156]]] : memref<?xindex>
+// CHECK: %[[VAL_158:.*]] = memref.load %[[VAL_24]]{{\[}}%[[VAL_153]]] : memref<?xf32>
// CHECK: %[[VAL_159:.*]] = scf.for %[[VAL_160:.*]] = %[[VAL_155]] to %[[VAL_157]] step %[[VAL_8]] iter_args(%[[VAL_161:.*]] = %[[VAL_158]]) -> (f32) {
-// CHECK: %[[VAL_162:.*]] = load %[[VAL_19]]{{\[}}%[[VAL_160]]] : memref<?xf32>
+// CHECK: %[[VAL_162:.*]] = memref.load %[[VAL_19]]{{\[}}%[[VAL_160]]] : memref<?xf32>
// CHECK: %[[VAL_163:.*]] = addf %[[VAL_161]], %[[VAL_162]] : f32
// CHECK: scf.yield %[[VAL_163]] : f32
// CHECK: }
// CHECK: store %[[VAL_164:.*]], %[[VAL_24]]{{\[}}%[[VAL_153]]] : memref<?xf32>
// CHECK: }
-// CHECK: %[[VAL_165:.*]] = tensor_load %[[VAL_24]] : memref<?xf32>
+// CHECK: %[[VAL_165:.*]] = memref.tensor_load %[[VAL_24]] : memref<?xf32>
// CHECK: return %[[VAL_165]] : tensor<?xf32>
// CHECK: }
func @sum_kernel_with_inv(%arga: tensor<?x?xf32>,
diff --git a/mlir/test/Dialect/Linalg/sparse_3d.mlir b/mlir/test/Dialect/Linalg/sparse_3d.mlir
index d0148199a606..cecc09ee7e7d 100644
--- a/mlir/test/Dialect/Linalg/sparse_3d.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_3d.mlir
@@ -25,22 +25,22 @@
// CHECK: %[[VAL_5:.*]] = constant 8 : index
// CHECK: %[[VAL_6:.*]] = constant 0 : index
// CHECK: %[[VAL_7:.*]] = constant 1 : index
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_0]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_6]] to %[[VAL_3]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_6]] to %[[VAL_5]] step %[[VAL_7]] {
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_17:.*]] = addf %[[VAL_15]], %[[VAL_16]] : f32
// CHECK: store %[[VAL_17]], %[[VAL_11]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_18:.*]] = tensor_load %[[VAL_11]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_18:.*]] = memref.tensor_load %[[VAL_11]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_18]] : tensor<32x16x8xf32>
// CHECK: }
func @add_ddd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -63,22 +63,22 @@ func @add_ddd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_5:.*]] = constant 8 : index
// CHECK: %[[VAL_6:.*]] = constant 0 : index
// CHECK: %[[VAL_7:.*]] = constant 1 : index
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_0]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_11:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_0]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_10]], %[[VAL_11]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_12:.*]] = %[[VAL_6]] to %[[VAL_3]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_6]] to %[[VAL_5]] step %[[VAL_7]] {
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_17:.*]] = mulf %[[VAL_15]], %[[VAL_16]] : f32
// CHECK: store %[[VAL_17]], %[[VAL_11]]{{\[}}%[[VAL_12]], %[[VAL_13]], %[[VAL_14]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_18:.*]] = tensor_load %[[VAL_11]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_18:.*]] = memref.tensor_load %[[VAL_11]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_18]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_ddd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -121,32 +121,32 @@ func @mul_ddd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_10:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_12:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_15:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_14]], %[[VAL_15]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: %[[VAL_18:.*]] = muli %[[VAL_16]], %[[VAL_5]] : index
// CHECK: %[[VAL_19:.*]] = addi %[[VAL_18]], %[[VAL_17]] : index
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]] = addi %[[VAL_19]], %[[VAL_9]] : index
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xindex>
// CHECK: %[[VAL_23:.*]]:2 = scf.while (%[[VAL_24:.*]] = %[[VAL_20]], %[[VAL_25:.*]] = %[[VAL_7]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_26:.*]] = cmpi ult, %[[VAL_24]], %[[VAL_22]] : index
// CHECK: scf.condition(%[[VAL_26]]) %[[VAL_24]], %[[VAL_25]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_27:.*]]: index, %[[VAL_28:.*]]: index):
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_30:.*]] = cmpi eq, %[[VAL_29]], %[[VAL_28]] : index
// CHECK: scf.if %[[VAL_30]] {
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_27]]] : memref<?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_33:.*]] = addf %[[VAL_31]], %[[VAL_32]] : f32
// CHECK: store %[[VAL_33]], %[[VAL_15]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_8]] {
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_34]], %[[VAL_15]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: }
@@ -158,12 +158,12 @@ func @mul_ddd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.yield %[[VAL_37]], %[[VAL_38]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_39:.*]] = %[[VAL_40:.*]]#1 to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_39]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_41]], %[[VAL_15]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_42:.*]] = tensor_load %[[VAL_15]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_42:.*]] = memref.tensor_load %[[VAL_15]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_42]] : tensor<32x16x8xf32>
// CHECK: }
func @add_dds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -189,27 +189,27 @@ func @add_dds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_6]] to %[[VAL_4]] step %[[VAL_7]] {
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_6]] to %[[VAL_5]] step %[[VAL_7]] {
// CHECK: %[[VAL_16:.*]] = muli %[[VAL_14]], %[[VAL_5]] : index
// CHECK: %[[VAL_17:.*]] = addi %[[VAL_16]], %[[VAL_15]] : index
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: %[[VAL_19:.*]] = addi %[[VAL_17]], %[[VAL_7]] : index
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_21:.*]] = %[[VAL_18]] to %[[VAL_20]] step %[[VAL_7]] {
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xf32>
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_15]], %[[VAL_22]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xf32>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_14]], %[[VAL_15]], %[[VAL_22]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_25:.*]] = mulf %[[VAL_23]], %[[VAL_24]] : f32
// CHECK: store %[[VAL_25]], %[[VAL_13]]{{\[}}%[[VAL_14]], %[[VAL_15]], %[[VAL_22]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_26:.*]] = tensor_load %[[VAL_13]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_26:.*]] = memref.tensor_load %[[VAL_13]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_26]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_dds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -251,34 +251,34 @@ func @mul_dds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_8]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_8]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_13]], %[[VAL_14]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_7]] to %[[VAL_3]] step %[[VAL_8]] {
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]] = addi %[[VAL_15]], %[[VAL_8]] : index
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: %[[VAL_19:.*]]:2 = scf.while (%[[VAL_20:.*]] = %[[VAL_16]], %[[VAL_21:.*]] = %[[VAL_7]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_22:.*]] = cmpi ult, %[[VAL_20]], %[[VAL_18]] : index
// CHECK: scf.condition(%[[VAL_22]]) %[[VAL_20]], %[[VAL_21]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_23:.*]]: index, %[[VAL_24:.*]]: index):
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
// CHECK: %[[VAL_26:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_24]] : index
// CHECK: scf.if %[[VAL_26]] {
// CHECK: scf.for %[[VAL_27:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
// CHECK: %[[VAL_28:.*]] = muli %[[VAL_23]], %[[VAL_5]] : index
// CHECK: %[[VAL_29:.*]] = addi %[[VAL_28]], %[[VAL_27]] : index
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_29]]] : memref<?xf32>
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_27]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_29]]] : memref<?xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_27]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_32:.*]] = addf %[[VAL_30]], %[[VAL_31]] : f32
// CHECK: store %[[VAL_32]], %[[VAL_14]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_27]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
// CHECK: scf.for %[[VAL_33:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_34:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_33]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_34:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_33]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_34]], %[[VAL_14]]{{\[}}%[[VAL_15]], %[[VAL_24]], %[[VAL_33]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
@@ -292,12 +292,12 @@ func @mul_dds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: }
// CHECK: scf.for %[[VAL_39:.*]] = %[[VAL_40:.*]]#1 to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_41:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_39]], %[[VAL_41]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_39]], %[[VAL_41]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_42]], %[[VAL_14]]{{\[}}%[[VAL_15]], %[[VAL_39]], %[[VAL_41]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_43:.*]] = tensor_load %[[VAL_14]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_43:.*]] = memref.tensor_load %[[VAL_14]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_43]] : tensor<32x16x8xf32>
// CHECK: }
func @add_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -322,27 +322,27 @@ func @add_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_6]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_6]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_5]] to %[[VAL_3]] step %[[VAL_6]] {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
// CHECK: %[[VAL_15:.*]] = addi %[[VAL_13]], %[[VAL_6]] : index
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_14]] to %[[VAL_16]] step %[[VAL_6]] {
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_19:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
// CHECK: %[[VAL_20:.*]] = muli %[[VAL_17]], %[[VAL_4]] : index
// CHECK: %[[VAL_21:.*]] = addi %[[VAL_20]], %[[VAL_19]] : index
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xf32>
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_13]], %[[VAL_18]], %[[VAL_19]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xf32>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_13]], %[[VAL_18]], %[[VAL_19]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_24:.*]] = mulf %[[VAL_22]], %[[VAL_23]] : f32
// CHECK: store %[[VAL_24]], %[[VAL_12]]{{\[}}%[[VAL_13]], %[[VAL_18]], %[[VAL_19]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_25:.*]] = tensor_load %[[VAL_12]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_25:.*]] = memref.tensor_load %[[VAL_12]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_25]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -387,40 +387,40 @@ func @mul_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_17:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_16:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_17:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_16]], %[[VAL_17]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_18:.*]] = %[[VAL_8]] to %[[VAL_4]] step %[[VAL_9]] {
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_18]]] : memref<?xindex>
// CHECK: %[[VAL_20:.*]] = addi %[[VAL_18]], %[[VAL_9]] : index
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_20]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]]:2 = scf.while (%[[VAL_23:.*]] = %[[VAL_19]], %[[VAL_24:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_21]] : index
// CHECK: scf.condition(%[[VAL_25]]) %[[VAL_23]], %[[VAL_24]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_26:.*]]: index, %[[VAL_27:.*]]: index):
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_26]]] : memref<?xindex>
// CHECK: %[[VAL_29:.*]] = cmpi eq, %[[VAL_28]], %[[VAL_27]] : index
// CHECK: scf.if %[[VAL_29]] {
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_26]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = addi %[[VAL_26]], %[[VAL_9]] : index
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<?xindex>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<?xindex>
// CHECK: %[[VAL_33:.*]]:2 = scf.while (%[[VAL_34:.*]] = %[[VAL_30]], %[[VAL_35:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_36:.*]] = cmpi ult, %[[VAL_34]], %[[VAL_32]] : index
// CHECK: scf.condition(%[[VAL_36]]) %[[VAL_34]], %[[VAL_35]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_37:.*]]: index, %[[VAL_38:.*]]: index):
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_37]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_38]] : index
// CHECK: scf.if %[[VAL_40]] {
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_37]]] : memref<?xf32>
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_37]]] : memref<?xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_43:.*]] = addf %[[VAL_41]], %[[VAL_42]] : f32
// CHECK: store %[[VAL_43]], %[[VAL_17]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
-// CHECK: %[[VAL_44:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_44:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_44]], %[[VAL_17]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: }
@@ -432,13 +432,13 @@ func @mul_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.yield %[[VAL_47]], %[[VAL_48]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_49:.*]] = %[[VAL_50:.*]]#1 to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_51:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_49]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_51:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_51]], %[[VAL_17]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
// CHECK: scf.for %[[VAL_52:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_53:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_52]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_53:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_52]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_53]], %[[VAL_17]]{{\[}}%[[VAL_18]], %[[VAL_27]], %[[VAL_52]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
@@ -452,12 +452,12 @@ func @mul_dsd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: }
// CHECK: scf.for %[[VAL_58:.*]] = %[[VAL_59:.*]]#1 to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_60:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_61:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_58]], %[[VAL_60]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_61:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_18]], %[[VAL_58]], %[[VAL_60]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_61]], %[[VAL_17]]{{\[}}%[[VAL_18]], %[[VAL_58]], %[[VAL_60]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_62:.*]] = tensor_load %[[VAL_17]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_62:.*]] = memref.tensor_load %[[VAL_17]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_62]] : tensor<32x16x8xf32>
// CHECK: }
func @add_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -484,29 +484,29 @@ func @add_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_13]], %[[VAL_14]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]] = addi %[[VAL_15]], %[[VAL_6]] : index
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_19:.*]] = %[[VAL_16]] to %[[VAL_18]] step %[[VAL_6]] {
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]] = addi %[[VAL_19]], %[[VAL_6]] : index
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_24:.*]] = %[[VAL_21]] to %[[VAL_23]] step %[[VAL_6]] {
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xindex>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_24]]] : memref<?xf32>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_20]], %[[VAL_25]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_15]], %[[VAL_20]], %[[VAL_25]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_28:.*]] = mulf %[[VAL_26]], %[[VAL_27]] : f32
// CHECK: store %[[VAL_28]], %[[VAL_14]]{{\[}}%[[VAL_15]], %[[VAL_20]], %[[VAL_25]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_29:.*]] = tensor_load %[[VAL_14]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_29:.*]] = memref.tensor_load %[[VAL_14]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_29]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -548,18 +548,18 @@ func @mul_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_7]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_7]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_13]], %[[VAL_14]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
// CHECK: %[[VAL_17:.*]]:2 = scf.while (%[[VAL_18:.*]] = %[[VAL_15]], %[[VAL_19:.*]] = %[[VAL_7]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_20:.*]] = cmpi ult, %[[VAL_18]], %[[VAL_16]] : index
// CHECK: scf.condition(%[[VAL_20]]) %[[VAL_18]], %[[VAL_19]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_21:.*]]: index, %[[VAL_22:.*]]: index):
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<?xindex>
// CHECK: %[[VAL_24:.*]] = cmpi eq, %[[VAL_23]], %[[VAL_22]] : index
// CHECK: scf.if %[[VAL_24]] {
// CHECK: scf.for %[[VAL_25:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
@@ -568,8 +568,8 @@ func @mul_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.for %[[VAL_28:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
// CHECK: %[[VAL_29:.*]] = muli %[[VAL_27]], %[[VAL_5]] : index
// CHECK: %[[VAL_30:.*]] = addi %[[VAL_29]], %[[VAL_28]] : index
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf32>
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_22]], %[[VAL_25]], %[[VAL_28]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_30]]] : memref<?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_22]], %[[VAL_25]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_33:.*]] = addf %[[VAL_31]], %[[VAL_32]] : f32
// CHECK: store %[[VAL_33]], %[[VAL_14]]{{\[}}%[[VAL_22]], %[[VAL_25]], %[[VAL_28]]] : memref<32x16x8xf32>
// CHECK: }
@@ -578,7 +578,7 @@ func @mul_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.if %[[VAL_6]] {
// CHECK: scf.for %[[VAL_34:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_35:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_22]], %[[VAL_34]], %[[VAL_35]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_22]], %[[VAL_34]], %[[VAL_35]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_36]], %[[VAL_14]]{{\[}}%[[VAL_22]], %[[VAL_34]], %[[VAL_35]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -594,12 +594,12 @@ func @mul_dss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.for %[[VAL_41:.*]] = %[[VAL_42:.*]]#1 to %[[VAL_3]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_43:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_44:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_45:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_41]], %[[VAL_43]], %[[VAL_44]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_45:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_41]], %[[VAL_43]], %[[VAL_44]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_45]], %[[VAL_14]]{{\[}}%[[VAL_41]], %[[VAL_43]], %[[VAL_44]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_46:.*]] = tensor_load %[[VAL_14]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_46:.*]] = memref.tensor_load %[[VAL_14]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_46]] : tensor<32x16x8xf32>
// CHECK: }
func @add_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -624,28 +624,28 @@ func @add_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_5]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_5]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_13:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_13]] to %[[VAL_14]] step %[[VAL_6]] {
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_5]] to %[[VAL_3]] step %[[VAL_6]] {
// CHECK: %[[VAL_18:.*]] = muli %[[VAL_15]], %[[VAL_3]] : index
// CHECK: %[[VAL_19:.*]] = addi %[[VAL_18]], %[[VAL_17]] : index
// CHECK: scf.for %[[VAL_20:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
// CHECK: %[[VAL_21:.*]] = muli %[[VAL_19]], %[[VAL_4]] : index
// CHECK: %[[VAL_22:.*]] = addi %[[VAL_21]], %[[VAL_20]] : index
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xf32>
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_20]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<?xf32>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_20]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_25:.*]] = mulf %[[VAL_23]], %[[VAL_24]] : f32
// CHECK: store %[[VAL_25]], %[[VAL_12]]{{\[}}%[[VAL_16]], %[[VAL_17]], %[[VAL_20]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_26:.*]] = tensor_load %[[VAL_12]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_26:.*]] = memref.tensor_load %[[VAL_12]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_26]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -690,41 +690,41 @@ func @mul_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_12:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_17:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_16:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_17:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_16]], %[[VAL_17]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_8]]] : memref<?xindex>
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_9]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_8]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_9]]] : memref<?xindex>
// CHECK: %[[VAL_20:.*]]:2 = scf.while (%[[VAL_21:.*]] = %[[VAL_18]], %[[VAL_22:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_23:.*]] = cmpi ult, %[[VAL_21]], %[[VAL_19]] : index
// CHECK: scf.condition(%[[VAL_23]]) %[[VAL_21]], %[[VAL_22]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_24:.*]]: index, %[[VAL_25:.*]]: index):
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_24]]] : memref<?xindex>
// CHECK: %[[VAL_27:.*]] = cmpi eq, %[[VAL_26]], %[[VAL_25]] : index
// CHECK: scf.if %[[VAL_27]] {
// CHECK: scf.for %[[VAL_28:.*]] = %[[VAL_8]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: %[[VAL_29:.*]] = muli %[[VAL_24]], %[[VAL_5]] : index
// CHECK: %[[VAL_30:.*]] = addi %[[VAL_29]], %[[VAL_28]] : index
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_30]]] : memref<?xindex>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_30]]] : memref<?xindex>
// CHECK: %[[VAL_32:.*]] = addi %[[VAL_30]], %[[VAL_9]] : index
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xindex>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_32]]] : memref<?xindex>
// CHECK: %[[VAL_34:.*]]:2 = scf.while (%[[VAL_35:.*]] = %[[VAL_31]], %[[VAL_36:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_37:.*]] = cmpi ult, %[[VAL_35]], %[[VAL_33]] : index
// CHECK: scf.condition(%[[VAL_37]]) %[[VAL_35]], %[[VAL_36]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_38:.*]]: index, %[[VAL_39:.*]]: index):
-// CHECK: %[[VAL_40:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_38]]] : memref<?xindex>
+// CHECK: %[[VAL_40:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_38]]] : memref<?xindex>
// CHECK: %[[VAL_41:.*]] = cmpi eq, %[[VAL_40]], %[[VAL_39]] : index
// CHECK: scf.if %[[VAL_41]] {
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_38]]] : memref<?xf32>
-// CHECK: %[[VAL_43:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_38]]] : memref<?xf32>
+// CHECK: %[[VAL_43:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_44:.*]] = addf %[[VAL_42]], %[[VAL_43]] : f32
// CHECK: store %[[VAL_44]], %[[VAL_17]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
-// CHECK: %[[VAL_45:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_45:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_45]], %[[VAL_17]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_39]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: }
@@ -736,7 +736,7 @@ func @mul_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.yield %[[VAL_48]], %[[VAL_49]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_50:.*]] = %[[VAL_51:.*]]#1 to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_50]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_50]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_52]], %[[VAL_17]]{{\[}}%[[VAL_25]], %[[VAL_28]], %[[VAL_50]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -744,7 +744,7 @@ func @mul_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.if %[[VAL_7]] {
// CHECK: scf.for %[[VAL_53:.*]] = %[[VAL_8]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_54:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_55:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_53]], %[[VAL_54]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_55:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_25]], %[[VAL_53]], %[[VAL_54]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_55]], %[[VAL_17]]{{\[}}%[[VAL_25]], %[[VAL_53]], %[[VAL_54]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -760,12 +760,12 @@ func @mul_sdd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.for %[[VAL_60:.*]] = %[[VAL_61:.*]]#1 to %[[VAL_4]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_62:.*]] = %[[VAL_8]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_63:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_64:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_60]], %[[VAL_62]], %[[VAL_63]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_64:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_60]], %[[VAL_62]], %[[VAL_63]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_64]], %[[VAL_17]]{{\[}}%[[VAL_60]], %[[VAL_62]], %[[VAL_63]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_65:.*]] = tensor_load %[[VAL_17]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_65:.*]] = memref.tensor_load %[[VAL_17]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_65]] : tensor<32x16x8xf32>
// CHECK: }
func @add_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -792,30 +792,30 @@ func @add_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_9:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_13]], %[[VAL_14]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_5]]] : memref<?xindex>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_6]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_15]] to %[[VAL_16]] step %[[VAL_6]] {
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_19:.*]] = %[[VAL_5]] to %[[VAL_4]] step %[[VAL_6]] {
// CHECK: %[[VAL_20:.*]] = muli %[[VAL_17]], %[[VAL_4]] : index
// CHECK: %[[VAL_21:.*]] = addi %[[VAL_20]], %[[VAL_19]] : index
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
// CHECK: %[[VAL_23:.*]] = addi %[[VAL_21]], %[[VAL_6]] : index
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_23]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_25:.*]] = %[[VAL_22]] to %[[VAL_24]] step %[[VAL_6]] {
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_18]], %[[VAL_19]], %[[VAL_26]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_18]], %[[VAL_19]], %[[VAL_26]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_29:.*]] = mulf %[[VAL_27]], %[[VAL_28]] : f32
// CHECK: store %[[VAL_29]], %[[VAL_14]]{{\[}}%[[VAL_18]], %[[VAL_19]], %[[VAL_26]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_30:.*]] = tensor_load %[[VAL_14]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_30:.*]] = memref.tensor_load %[[VAL_14]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_30]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -859,43 +859,43 @@ func @mul_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_11:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_8]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_12:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_8]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_13:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_14:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_7]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_8]]] : memref<?xindex>
// CHECK: %[[VAL_19:.*]]:2 = scf.while (%[[VAL_20:.*]] = %[[VAL_17]], %[[VAL_21:.*]] = %[[VAL_7]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_22:.*]] = cmpi ult, %[[VAL_20]], %[[VAL_18]] : index
// CHECK: scf.condition(%[[VAL_22]]) %[[VAL_20]], %[[VAL_21]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_23:.*]]: index, %[[VAL_24:.*]]: index):
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
// CHECK: %[[VAL_26:.*]] = cmpi eq, %[[VAL_25]], %[[VAL_24]] : index
// CHECK: scf.if %[[VAL_26]] {
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_23]]] : memref<?xindex>
// CHECK: %[[VAL_28:.*]] = addi %[[VAL_23]], %[[VAL_8]] : index
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
// CHECK: %[[VAL_30:.*]]:2 = scf.while (%[[VAL_31:.*]] = %[[VAL_27]], %[[VAL_32:.*]] = %[[VAL_7]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_33:.*]] = cmpi ult, %[[VAL_31]], %[[VAL_29]] : index
// CHECK: scf.condition(%[[VAL_33]]) %[[VAL_31]], %[[VAL_32]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_34:.*]]: index, %[[VAL_35:.*]]: index):
-// CHECK: %[[VAL_36:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_34]]] : memref<?xindex>
+// CHECK: %[[VAL_36:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_34]]] : memref<?xindex>
// CHECK: %[[VAL_37:.*]] = cmpi eq, %[[VAL_36]], %[[VAL_35]] : index
// CHECK: scf.if %[[VAL_37]] {
// CHECK: scf.for %[[VAL_38:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
// CHECK: %[[VAL_39:.*]] = muli %[[VAL_34]], %[[VAL_5]] : index
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_39]], %[[VAL_38]] : index
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_40]]] : memref<?xf32>
-// CHECK: %[[VAL_42:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_38]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_40]]] : memref<?xf32>
+// CHECK: %[[VAL_42:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_43:.*]] = addf %[[VAL_41]], %[[VAL_42]] : f32
// CHECK: store %[[VAL_43]], %[[VAL_16]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_38]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_6]] {
// CHECK: scf.for %[[VAL_44:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_45:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_44]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_45:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_44]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_45]], %[[VAL_16]]{{\[}}%[[VAL_24]], %[[VAL_35]], %[[VAL_44]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
@@ -909,7 +909,7 @@ func @mul_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: }
// CHECK: scf.for %[[VAL_50:.*]] = %[[VAL_51:.*]]#1 to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_52:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_53:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_50]], %[[VAL_52]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_53:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_50]], %[[VAL_52]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_53]], %[[VAL_16]]{{\[}}%[[VAL_24]], %[[VAL_50]], %[[VAL_52]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -917,7 +917,7 @@ func @mul_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.if %[[VAL_6]] {
// CHECK: scf.for %[[VAL_54:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_55:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_56:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_54]], %[[VAL_55]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_56:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_24]], %[[VAL_54]], %[[VAL_55]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_56]], %[[VAL_16]]{{\[}}%[[VAL_24]], %[[VAL_54]], %[[VAL_55]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -933,12 +933,12 @@ func @mul_sds(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.for %[[VAL_61:.*]] = %[[VAL_62:.*]]#1 to %[[VAL_3]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_63:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
// CHECK: scf.for %[[VAL_64:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_65:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_61]], %[[VAL_63]], %[[VAL_64]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_65:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_61]], %[[VAL_63]], %[[VAL_64]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_65]], %[[VAL_16]]{{\[}}%[[VAL_61]], %[[VAL_63]], %[[VAL_64]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_66:.*]] = tensor_load %[[VAL_16]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_66:.*]] = memref.tensor_load %[[VAL_16]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_66]] : tensor<32x16x8xf32>
// CHECK: }
func @add_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -964,30 +964,30 @@ func @add_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_8:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_5]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_5]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_10:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_14]] to %[[VAL_15]] step %[[VAL_5]] {
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xindex>
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_16]]] : memref<?xindex>
// CHECK: %[[VAL_19:.*]] = addi %[[VAL_16]], %[[VAL_5]] : index
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_21:.*]] = %[[VAL_18]] to %[[VAL_20]] step %[[VAL_5]] {
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_21]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_23:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
// CHECK: %[[VAL_24:.*]] = muli %[[VAL_21]], %[[VAL_3]] : index
// CHECK: %[[VAL_25:.*]] = addi %[[VAL_24]], %[[VAL_23]] : index
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xf32>
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_17]], %[[VAL_22]], %[[VAL_23]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_25]]] : memref<?xf32>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_17]], %[[VAL_22]], %[[VAL_23]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_28:.*]] = mulf %[[VAL_26]], %[[VAL_27]] : f32
// CHECK: store %[[VAL_28]], %[[VAL_13]]{{\[}}%[[VAL_17]], %[[VAL_22]], %[[VAL_23]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_29:.*]] = tensor_load %[[VAL_13]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_29:.*]] = memref.tensor_load %[[VAL_13]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_29]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -1034,49 +1034,49 @@ func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_14:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_15:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_16:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_17:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_18:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_19:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_17:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_18:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_19:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_18]], %[[VAL_19]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_8]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_9]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_8]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_9]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]]:2 = scf.while (%[[VAL_23:.*]] = %[[VAL_20]], %[[VAL_24:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_25:.*]] = cmpi ult, %[[VAL_23]], %[[VAL_21]] : index
// CHECK: scf.condition(%[[VAL_25]]) %[[VAL_23]], %[[VAL_24]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_26:.*]]: index, %[[VAL_27:.*]]: index):
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_26]]] : memref<?xindex>
// CHECK: %[[VAL_29:.*]] = cmpi eq, %[[VAL_28]], %[[VAL_27]] : index
// CHECK: scf.if %[[VAL_29]] {
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_26]]] : memref<?xindex>
// CHECK: %[[VAL_31:.*]] = addi %[[VAL_26]], %[[VAL_9]] : index
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<?xindex>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_31]]] : memref<?xindex>
// CHECK: %[[VAL_33:.*]]:2 = scf.while (%[[VAL_34:.*]] = %[[VAL_30]], %[[VAL_35:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_36:.*]] = cmpi ult, %[[VAL_34]], %[[VAL_32]] : index
// CHECK: scf.condition(%[[VAL_36]]) %[[VAL_34]], %[[VAL_35]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_37:.*]]: index, %[[VAL_38:.*]]: index):
-// CHECK: %[[VAL_39:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_39:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_37]]] : memref<?xindex>
// CHECK: %[[VAL_40:.*]] = cmpi eq, %[[VAL_39]], %[[VAL_38]] : index
// CHECK: scf.if %[[VAL_40]] {
-// CHECK: %[[VAL_41:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_37]]] : memref<?xindex>
+// CHECK: %[[VAL_41:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_37]]] : memref<?xindex>
// CHECK: %[[VAL_42:.*]] = addi %[[VAL_37]], %[[VAL_9]] : index
-// CHECK: %[[VAL_43:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_42]]] : memref<?xindex>
+// CHECK: %[[VAL_43:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_42]]] : memref<?xindex>
// CHECK: %[[VAL_44:.*]]:2 = scf.while (%[[VAL_45:.*]] = %[[VAL_41]], %[[VAL_46:.*]] = %[[VAL_8]]) : (index, index) -> (index, index) {
// CHECK: %[[VAL_47:.*]] = cmpi ult, %[[VAL_45]], %[[VAL_43]] : index
// CHECK: scf.condition(%[[VAL_47]]) %[[VAL_45]], %[[VAL_46]] : index, index
// CHECK: } do {
// CHECK: ^bb0(%[[VAL_48:.*]]: index, %[[VAL_49:.*]]: index):
-// CHECK: %[[VAL_50:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_48]]] : memref<?xindex>
+// CHECK: %[[VAL_50:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_48]]] : memref<?xindex>
// CHECK: %[[VAL_51:.*]] = cmpi eq, %[[VAL_50]], %[[VAL_49]] : index
// CHECK: scf.if %[[VAL_51]] {
-// CHECK: %[[VAL_52:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_48]]] : memref<?xf32>
-// CHECK: %[[VAL_53:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_52:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_48]]] : memref<?xf32>
+// CHECK: %[[VAL_53:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_54:.*]] = addf %[[VAL_52]], %[[VAL_53]] : f32
// CHECK: store %[[VAL_54]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
-// CHECK: %[[VAL_55:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_55:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_55]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_49]]] : memref<32x16x8xf32>
// CHECK: } else {
// CHECK: }
@@ -1088,13 +1088,13 @@ func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.yield %[[VAL_58]], %[[VAL_59]] : index, index
// CHECK: }
// CHECK: scf.for %[[VAL_60:.*]] = %[[VAL_61:.*]]#1 to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_62:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_60]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_62:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_60]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_62]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_60]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
// CHECK: scf.if %[[VAL_7]] {
// CHECK: scf.for %[[VAL_63:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_64:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_63]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_64:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_63]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_64]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_38]], %[[VAL_63]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: } else {
@@ -1108,7 +1108,7 @@ func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: }
// CHECK: scf.for %[[VAL_69:.*]] = %[[VAL_70:.*]]#1 to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_71:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_72:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_69]], %[[VAL_71]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_72:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_69]], %[[VAL_71]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_72]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_69]], %[[VAL_71]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -1116,7 +1116,7 @@ func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.if %[[VAL_7]] {
// CHECK: scf.for %[[VAL_73:.*]] = %[[VAL_8]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_74:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_75:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_73]], %[[VAL_74]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_75:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_27]], %[[VAL_73]], %[[VAL_74]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_75]], %[[VAL_19]]{{\[}}%[[VAL_27]], %[[VAL_73]], %[[VAL_74]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
@@ -1132,12 +1132,12 @@ func @mul_ssd(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: scf.for %[[VAL_80:.*]] = %[[VAL_81:.*]]#1 to %[[VAL_4]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_82:.*]] = %[[VAL_8]] to %[[VAL_5]] step %[[VAL_9]] {
// CHECK: scf.for %[[VAL_83:.*]] = %[[VAL_8]] to %[[VAL_6]] step %[[VAL_9]] {
-// CHECK: %[[VAL_84:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_80]], %[[VAL_82]], %[[VAL_83]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_84:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_80]], %[[VAL_82]], %[[VAL_83]]] : memref<32x16x8xf32>
// CHECK: store %[[VAL_84]], %[[VAL_19]]{{\[}}%[[VAL_80]], %[[VAL_82]], %[[VAL_83]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_85:.*]] = tensor_load %[[VAL_19]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_85:.*]] = memref.tensor_load %[[VAL_19]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_85]] : tensor<32x16x8xf32>
// CHECK: }
func @add_sss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -1165,32 +1165,32 @@ func @add_sss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_10:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_11:.*]] = linalg.sparse_indices %[[VAL_0]], %[[VAL_3]] : tensor<32x16x8xf32> to memref<?xindex>
// CHECK: %[[VAL_12:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<32x16x8xf32> to memref<?xf32>
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_1]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_14:.*]] = tensor_to_memref %[[VAL_2]] : memref<32x16x8xf32>
-// CHECK: %[[VAL_15:.*]] = alloc() : memref<32x16x8xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_1]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_14:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_15:.*]] = memref.alloc() : memref<32x16x8xf32>
// CHECK: linalg.copy(%[[VAL_14]], %[[VAL_15]]) : memref<32x16x8xf32>, memref<32x16x8xf32>
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_5]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_18:.*]] = %[[VAL_16]] to %[[VAL_17]] step %[[VAL_5]] {
-// CHECK: %[[VAL_19:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_19:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xindex>
// CHECK: %[[VAL_21:.*]] = addi %[[VAL_18]], %[[VAL_5]] : index
-// CHECK: %[[VAL_22:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_21]]] : memref<?xindex>
+// CHECK: %[[VAL_22:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_21]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_23:.*]] = %[[VAL_20]] to %[[VAL_22]] step %[[VAL_5]] {
-// CHECK: %[[VAL_24:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_23]]] : memref<?xindex>
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_24:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_23]]] : memref<?xindex>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_23]]] : memref<?xindex>
// CHECK: %[[VAL_26:.*]] = addi %[[VAL_23]], %[[VAL_5]] : index
-// CHECK: %[[VAL_27:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_26]]] : memref<?xindex>
+// CHECK: %[[VAL_27:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_26]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_28:.*]] = %[[VAL_25]] to %[[VAL_27]] step %[[VAL_5]] {
-// CHECK: %[[VAL_29:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xf32>
-// CHECK: %[[VAL_31:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_19]], %[[VAL_24]], %[[VAL_29]]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_29:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_28]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_28]]] : memref<?xf32>
+// CHECK: %[[VAL_31:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_19]], %[[VAL_24]], %[[VAL_29]]] : memref<32x16x8xf32>
// CHECK: %[[VAL_32:.*]] = mulf %[[VAL_30]], %[[VAL_31]] : f32
// CHECK: store %[[VAL_32]], %[[VAL_15]]{{\[}}%[[VAL_19]], %[[VAL_24]], %[[VAL_29]]] : memref<32x16x8xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_33:.*]] = tensor_load %[[VAL_15]] : memref<32x16x8xf32>
+// CHECK: %[[VAL_33:.*]] = memref.tensor_load %[[VAL_15]] : memref<32x16x8xf32>
// CHECK: return %[[VAL_33]] : tensor<32x16x8xf32>
// CHECK: }
func @mul_sss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: tensor<32x16x8xf32>) -> tensor<32x16x8xf32> {
@@ -1232,37 +1232,37 @@ func @mul_sss(%arga: tensor<32x16x8xf32>, %argb: tensor<32x16x8xf32>, %argx: ten
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<?x?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<?x?x?xf32> to memref<?xindex>
// CHECK: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<?x?x?xf32> to memref<?xf32>
-// CHECK: %[[VAL_10:.*]] = dim %[[VAL_2]], %[[VAL_5]] : tensor<?x?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<?x?xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_3]] : memref<?x?xf32>
-// CHECK: %[[VAL_13:.*]] = dim %[[VAL_0]], %[[VAL_5]] : tensor<?x?xf32>
-// CHECK: %[[VAL_14:.*]] = dim %[[VAL_0]], %[[VAL_6]] : tensor<?x?xf32>
-// CHECK: %[[VAL_15:.*]] = tensor_to_memref %[[VAL_0]] : memref<?x?xf32>
-// CHECK: %[[VAL_16:.*]] = alloc(%[[VAL_13]], %[[VAL_14]]) : memref<?x?xf32>
+// CHECK: %[[VAL_10:.*]] = memref.dim %[[VAL_2]], %[[VAL_5]] : tensor<?x?xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<?x?xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_3]] : memref<?x?xf32>
+// CHECK: %[[VAL_13:.*]] = memref.dim %[[VAL_0]], %[[VAL_5]] : tensor<?x?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.dim %[[VAL_0]], %[[VAL_6]] : tensor<?x?xf32>
+// CHECK: %[[VAL_15:.*]] = memref.buffer_cast %[[VAL_0]] : memref<?x?xf32>
+// CHECK: %[[VAL_16:.*]] = memref.alloc(%[[VAL_13]], %[[VAL_14]]) : memref<?x?xf32>
// CHECK: linalg.copy(%[[VAL_15]], %[[VAL_16]]) : memref<?x?xf32>, memref<?x?xf32>
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_5]] to %[[VAL_13]] step %[[VAL_6]] {
// CHECK: scf.for %[[VAL_18:.*]] = %[[VAL_5]] to %[[VAL_10]] step %[[VAL_6]] {
// CHECK: %[[VAL_19:.*]] = muli %[[VAL_10]], %[[VAL_17]] : index
// CHECK: %[[VAL_20:.*]] = addi %[[VAL_19]], %[[VAL_18]] : index
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
// CHECK: %[[VAL_22:.*]] = addi %[[VAL_20]], %[[VAL_6]] : index
-// CHECK: %[[VAL_23:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_22]]] : memref<?xindex>
+// CHECK: %[[VAL_23:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_22]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_24:.*]] = %[[VAL_21]] to %[[VAL_23]] step %[[VAL_6]] {
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xindex>
-// CHECK: %[[VAL_26:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xf32>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_24]]] : memref<?xindex>
+// CHECK: %[[VAL_26:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_24]]] : memref<?xf32>
// CHECK: scf.for %[[VAL_27:.*]] = %[[VAL_5]] to %[[VAL_14]] step %[[VAL_6]] {
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_18]], %[[VAL_27]]] : memref<?x?xf32>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_18]], %[[VAL_27]]] : memref<?x?xf32>
// CHECK: %[[VAL_29:.*]] = mulf %[[VAL_26]], %[[VAL_28]] : f32
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_25]], %[[VAL_27]]] : memref<?x?xf32>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_25]], %[[VAL_27]]] : memref<?x?xf32>
// CHECK: %[[VAL_31:.*]] = mulf %[[VAL_29]], %[[VAL_30]] : f32
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_17]], %[[VAL_27]]] : memref<?x?xf32>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_17]], %[[VAL_27]]] : memref<?x?xf32>
// CHECK: %[[VAL_33:.*]] = addf %[[VAL_31]], %[[VAL_32]] : f32
// CHECK: store %[[VAL_33]], %[[VAL_16]]{{\[}}%[[VAL_17]], %[[VAL_27]]] : memref<?x?xf32>
// CHECK: }
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_34:.*]] = tensor_load %[[VAL_16]] : memref<?x?xf32>
+// CHECK: %[[VAL_34:.*]] = memref.tensor_load %[[VAL_16]] : memref<?x?xf32>
// CHECK: return %[[VAL_34]] : tensor<?x?xf32>
// CHECK: }
func @kernel_3d(%arga: tensor<?x?xf32>,
@@ -1304,29 +1304,29 @@ func @kernel_3d(%arga: tensor<?x?xf32>,
// CHECK: %[[VAL_6:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_4]] : tensor<10x20x30xf32> to memref<?xindex>
// CHECK: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_0]], %[[VAL_2]] : tensor<10x20x30xf32> to memref<?xindex>
// CHECK: %[[VAL_8:.*]] = linalg.sparse_values %[[VAL_0]] : tensor<10x20x30xf32> to memref<?xf32>
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<f32>
-// CHECK: %[[VAL_10:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<f32>
+// CHECK: %[[VAL_10:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_9]], %[[VAL_10]]) : memref<f32>, memref<f32>
-// CHECK: %[[VAL_11:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
-// CHECK: %[[VAL_12:.*]] = load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
+// CHECK: %[[VAL_11:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_3]]] : memref<?xindex>
+// CHECK: %[[VAL_12:.*]] = memref.load %[[VAL_5]]{{\[}}%[[VAL_4]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_11]] to %[[VAL_12]] step %[[VAL_4]] {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
// CHECK: %[[VAL_15:.*]] = addi %[[VAL_13]], %[[VAL_4]] : index
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_15]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_17:.*]] = %[[VAL_14]] to %[[VAL_16]] step %[[VAL_4]] {
-// CHECK: %[[VAL_18:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
+// CHECK: %[[VAL_18:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
// CHECK: %[[VAL_19:.*]] = addi %[[VAL_17]], %[[VAL_4]] : index
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_19]]] : memref<?xindex>
-// CHECK: %[[VAL_21:.*]] = load %[[VAL_10]][] : memref<f32>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK: %[[VAL_21:.*]] = memref.load %[[VAL_10]][] : memref<f32>
// CHECK: %[[VAL_22:.*]] = scf.for %[[VAL_23:.*]] = %[[VAL_18]] to %[[VAL_20]] step %[[VAL_4]] iter_args(%[[VAL_24:.*]] = %[[VAL_21]]) -> (f32) {
-// CHECK: %[[VAL_25:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_23]]] : memref<?xf32>
+// CHECK: %[[VAL_25:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_23]]] : memref<?xf32>
// CHECK: %[[VAL_26:.*]] = addf %[[VAL_24]], %[[VAL_25]] : f32
// CHECK: scf.yield %[[VAL_26]] : f32
// CHECK: }
// CHECK: store %[[VAL_27:.*]], %[[VAL_10]][] : memref<f32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_28:.*]] = tensor_load %[[VAL_10]] : memref<f32>
+// CHECK: %[[VAL_28:.*]] = memref.tensor_load %[[VAL_10]] : memref<f32>
// CHECK: return %[[VAL_28]] : tensor<f32>
// CHECK: }
func @sum_reduction(%arga: tensor<10x20x30xf32>, %argx: tensor<f32>) -> tensor<f32> {
@@ -1362,20 +1362,20 @@ func @sum_reduction(%arga: tensor<10x20x30xf32>, %argx: tensor<f32>) -> tensor<f
// CHECK: %[[VAL_3:.*]] = constant 2 : index
// CHECK: %[[VAL_4:.*]] = constant 0 : index
// CHECK: %[[VAL_5:.*]] = constant 1 : index
-// CHECK: %[[VAL_6:.*]] = dim %[[VAL_0]], %[[VAL_5]] : tensor<?x?x?xf32>
-// CHECK: %[[VAL_7:.*]] = dim %[[VAL_0]], %[[VAL_3]] : tensor<?x?x?xf32>
-// CHECK: %[[VAL_8:.*]] = tensor_to_memref %[[VAL_0]] : memref<?x?x?xf32>
-// CHECK: %[[VAL_9:.*]] = dim %[[VAL_1]], %[[VAL_4]] : tensor<?xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<?xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<f32>
-// CHECK: %[[VAL_12:.*]] = alloc() : memref<f32>
+// CHECK: %[[VAL_6:.*]] = memref.dim %[[VAL_0]], %[[VAL_5]] : tensor<?x?x?xf32>
+// CHECK: %[[VAL_7:.*]] = memref.dim %[[VAL_0]], %[[VAL_3]] : tensor<?x?x?xf32>
+// CHECK: %[[VAL_8:.*]] = memref.buffer_cast %[[VAL_0]] : memref<?x?x?xf32>
+// CHECK: %[[VAL_9:.*]] = memref.dim %[[VAL_1]], %[[VAL_4]] : tensor<?xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<?xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<f32>
+// CHECK: %[[VAL_12:.*]] = memref.alloc() : memref<f32>
// CHECK: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<f32>, memref<f32>
// CHECK: scf.for %[[VAL_13:.*]] = %[[VAL_4]] to %[[VAL_9]] step %[[VAL_5]] {
-// CHECK: %[[VAL_14:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_13]]] : memref<?xf32>
+// CHECK: %[[VAL_14:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_13]]] : memref<?xf32>
// CHECK: scf.for %[[VAL_15:.*]] = %[[VAL_4]] to %[[VAL_6]] step %[[VAL_5]] {
-// CHECK: %[[VAL_16:.*]] = load %[[VAL_12]][] : memref<f32>
+// CHECK: %[[VAL_16:.*]] = memref.load %[[VAL_12]][] : memref<f32>
// CHECK: %[[VAL_17:.*]] = scf.for %[[VAL_18:.*]] = %[[VAL_4]] to %[[VAL_7]] step %[[VAL_5]] iter_args(%[[VAL_19:.*]] = %[[VAL_16]]) -> (f32) {
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_13]], %[[VAL_15]], %[[VAL_18]]] : memref<?x?x?xf32>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_13]], %[[VAL_15]], %[[VAL_18]]] : memref<?x?x?xf32>
// CHECK: %[[VAL_21:.*]] = mulf %[[VAL_20]], %[[VAL_14]] : f32
// CHECK: %[[VAL_22:.*]] = addf %[[VAL_19]], %[[VAL_21]] : f32
// CHECK: scf.yield %[[VAL_22]] : f32
@@ -1383,7 +1383,7 @@ func @sum_reduction(%arga: tensor<10x20x30xf32>, %argx: tensor<f32>) -> tensor<f
// CHECK: store %[[VAL_23:.*]], %[[VAL_12]][] : memref<f32>
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_24:.*]] = tensor_load %[[VAL_12]] : memref<f32>
+// CHECK: %[[VAL_24:.*]] = memref.tensor_load %[[VAL_12]] : memref<f32>
// CHECK: return %[[VAL_24]] : tensor<f32>
// CHECK: }
func @sum_reduction_inv(%arga: tensor<?x?x?xf32>,
@@ -1427,25 +1427,25 @@ func @sum_reduction_inv(%arga: tensor<?x?x?xf32>,
// CHECK: %[[VAL_6:.*]] = constant 30 : index
// CHECK: %[[VAL_7:.*]] = constant 0 : index
// CHECK: %[[VAL_8:.*]] = constant 1 : index
-// CHECK: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_0]] : memref<10xf32>
-// CHECK: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<20xf32>
-// CHECK: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<30xf32>
-// CHECK: %[[VAL_12:.*]] = tensor_to_memref %[[VAL_3]] : memref<10x20x30xf32>
-// CHECK: %[[VAL_13:.*]] = alloc() : memref<10x20x30xf32>
+// CHECK: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_0]] : memref<10xf32>
+// CHECK: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<20xf32>
+// CHECK: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<30xf32>
+// CHECK: %[[VAL_12:.*]] = memref.buffer_cast %[[VAL_3]] : memref<10x20x30xf32>
+// CHECK: %[[VAL_13:.*]] = memref.alloc() : memref<10x20x30xf32>
// CHECK: linalg.copy(%[[VAL_12]], %[[VAL_13]]) : memref<10x20x30xf32>, memref<10x20x30xf32>
// CHECK: scf.for %[[VAL_14:.*]] = %[[VAL_7]] to %[[VAL_4]] step %[[VAL_8]] {
-// CHECK: %[[VAL_15:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_14]]] : memref<10xf32>
+// CHECK: %[[VAL_15:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_14]]] : memref<10xf32>
// CHECK: scf.for %[[VAL_16:.*]] = %[[VAL_7]] to %[[VAL_5]] step %[[VAL_8]] {
-// CHECK: %[[VAL_17:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_16]]] : memref<20xf32>
+// CHECK: %[[VAL_17:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_16]]] : memref<20xf32>
// CHECK: scf.for %[[VAL_18:.*]] = %[[VAL_7]] to %[[VAL_6]] step %[[VAL_8]] {
// CHECK: %[[VAL_19:.*]] = mulf %[[VAL_15]], %[[VAL_17]] : f32
-// CHECK: %[[VAL_20:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_18]]] : memref<30xf32>
+// CHECK: %[[VAL_20:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_18]]] : memref<30xf32>
// CHECK: %[[VAL_21:.*]] = mulf %[[VAL_19]], %[[VAL_20]] : f32
// CHECK: store %[[VAL_21]], %[[VAL_13]]{{\[}}%[[VAL_14]], %[[VAL_16]], %[[VAL_18]]] : memref<10x20x30xf32>
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_22:.*]] = tensor_load %[[VAL_13]] : memref<10x20x30xf32>
+// CHECK: %[[VAL_22:.*]] = memref.tensor_load %[[VAL_13]] : memref<10x20x30xf32>
// CHECK: return %[[VAL_22]] : tensor<10x20x30xf32>
// CHECK: }
func @invariants(%arga: tensor<10xf32>,
diff --git a/mlir/test/Dialect/Linalg/sparse_lower.mlir b/mlir/test/Dialect/Linalg/sparse_lower.mlir
index 26888d979126..58ebd7a12041 100644
--- a/mlir/test/Dialect/Linalg/sparse_lower.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_lower.mlir
@@ -45,26 +45,26 @@
// CHECK-HIR: %[[VAL_7:.*]] = linalg.sparse_pointers %[[VAL_6]], %[[VAL_5]] : tensor<64x64xf64> to memref<?xindex>
// CHECK-HIR: %[[VAL_8:.*]] = linalg.sparse_indices %[[VAL_6]], %[[VAL_5]] : tensor<64x64xf64> to memref<?xindex>
// CHECK-HIR: %[[VAL_9:.*]] = linalg.sparse_values %[[VAL_6]] : tensor<64x64xf64> to memref<?xf64>
-// CHECK-HIR: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_1]] : memref<64xf64>
-// CHECK-HIR: %[[VAL_11:.*]] = tensor_to_memref %[[VAL_2]] : memref<64xf64>
-// CHECK-HIR: %[[VAL_12:.*]] = alloc() : memref<64xf64>
+// CHECK-HIR: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_1]] : memref<64xf64>
+// CHECK-HIR: %[[VAL_11:.*]] = memref.buffer_cast %[[VAL_2]] : memref<64xf64>
+// CHECK-HIR: %[[VAL_12:.*]] = memref.alloc() : memref<64xf64>
// CHECK-HIR: linalg.copy(%[[VAL_11]], %[[VAL_12]]) : memref<64xf64>, memref<64xf64>
// CHECK-HIR: scf.for %[[VAL_13:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-HIR: %[[VAL_14:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
+// CHECK-HIR: %[[VAL_14:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_13]]] : memref<?xindex>
// CHECK-HIR: %[[VAL_15:.*]] = addi %[[VAL_13]], %[[VAL_5]] : index
-// CHECK-HIR: %[[VAL_16:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
-// CHECK-HIR: %[[VAL_17:.*]] = load %[[VAL_12]]{{\[}}%[[VAL_13]]] : memref<64xf64>
+// CHECK-HIR: %[[VAL_16:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK-HIR: %[[VAL_17:.*]] = memref.load %[[VAL_12]]{{\[}}%[[VAL_13]]] : memref<64xf64>
// CHECK-HIR: %[[VAL_18:.*]] = scf.for %[[VAL_19:.*]] = %[[VAL_14]] to %[[VAL_16]] step %[[VAL_5]] iter_args(%[[VAL_20:.*]] = %[[VAL_17]]) -> (f64) {
-// CHECK-HIR: %[[VAL_21:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
-// CHECK-HIR: %[[VAL_22:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xf64>
-// CHECK-HIR: %[[VAL_23:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<64xf64>
+// CHECK-HIR: %[[VAL_21:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_19]]] : memref<?xindex>
+// CHECK-HIR: %[[VAL_22:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<?xf64>
+// CHECK-HIR: %[[VAL_23:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_21]]] : memref<64xf64>
// CHECK-HIR: %[[VAL_24:.*]] = mulf %[[VAL_22]], %[[VAL_23]] : f64
// CHECK-HIR: %[[VAL_25:.*]] = addf %[[VAL_20]], %[[VAL_24]] : f64
// CHECK-HIR: scf.yield %[[VAL_25]] : f64
// CHECK-HIR: }
// CHECK-HIR: store %[[VAL_26:.*]], %[[VAL_12]]{{\[}}%[[VAL_13]]] : memref<64xf64>
// CHECK-HIR: }
-// CHECK-HIR: %[[VAL_27:.*]] = tensor_load %[[VAL_12]] : memref<64xf64>
+// CHECK-HIR: %[[VAL_27:.*]] = memref.tensor_load %[[VAL_12]] : memref<64xf64>
// CHECK-HIR: return %[[VAL_27]] : tensor<64xf64>
// CHECK-HIR: }
@@ -78,29 +78,29 @@
// CHECK-MIR: %[[VAL_6:.*]] = call @sparsePointers64(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
// CHECK-MIR: %[[VAL_7:.*]] = call @sparseIndices64(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
// CHECK-MIR: %[[VAL_8:.*]] = call @sparseValuesF64(%[[VAL_0]]) : (!llvm.ptr<i8>) -> memref<?xf64>
-// CHECK-MIR: %[[VAL_9:.*]] = tensor_to_memref %[[VAL_1]] : memref<64xf64>
-// CHECK-MIR: %[[VAL_10:.*]] = tensor_to_memref %[[VAL_2]] : memref<64xf64>
-// CHECK-MIR: %[[VAL_11:.*]] = alloc() : memref<64xf64>
+// CHECK-MIR: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_11:.*]] = memref.alloc() : memref<64xf64>
// CHECK-MIR: scf.for %[[VAL_12:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-MIR: %[[VAL_13:.*]] = load %[[VAL_10]]{{\[}}%[[VAL_12]]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_13:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_12]]] : memref<64xf64>
// CHECK-MIR: store %[[VAL_13]], %[[VAL_11]]{{\[}}%[[VAL_12]]] : memref<64xf64>
// CHECK-MIR: }
// CHECK-MIR: scf.for %[[VAL_14:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-MIR: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK-MIR: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
// CHECK-MIR: %[[VAL_16:.*]] = addi %[[VAL_14]], %[[VAL_5]] : index
-// CHECK-MIR: %[[VAL_17:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_16]]] : memref<?xindex>
-// CHECK-MIR: %[[VAL_18:.*]] = load %[[VAL_11]]{{\[}}%[[VAL_14]]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_17:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_16]]] : memref<?xindex>
+// CHECK-MIR: %[[VAL_18:.*]] = memref.load %[[VAL_11]]{{\[}}%[[VAL_14]]] : memref<64xf64>
// CHECK-MIR: %[[VAL_19:.*]] = scf.for %[[VAL_20:.*]] = %[[VAL_15]] to %[[VAL_17]] step %[[VAL_5]] iter_args(%[[VAL_21:.*]] = %[[VAL_18]]) -> (f64) {
-// CHECK-MIR: %[[VAL_22:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
-// CHECK-MIR: %[[VAL_23:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xf64>
-// CHECK-MIR: %[[VAL_24:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_22:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_20]]] : memref<?xindex>
+// CHECK-MIR: %[[VAL_23:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_20]]] : memref<?xf64>
+// CHECK-MIR: %[[VAL_24:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_22]]] : memref<64xf64>
// CHECK-MIR: %[[VAL_25:.*]] = mulf %[[VAL_23]], %[[VAL_24]] : f64
// CHECK-MIR: %[[VAL_26:.*]] = addf %[[VAL_21]], %[[VAL_25]] : f64
// CHECK-MIR: scf.yield %[[VAL_26]] : f64
// CHECK-MIR: }
// CHECK-MIR: store %[[VAL_27:.*]], %[[VAL_11]]{{\[}}%[[VAL_14]]] : memref<64xf64>
// CHECK-MIR: }
-// CHECK-MIR: %[[VAL_28:.*]] = tensor_load %[[VAL_11]] : memref<64xf64>
+// CHECK-MIR: %[[VAL_28:.*]] = memref.tensor_load %[[VAL_11]] : memref<64xf64>
// CHECK-MIR: return %[[VAL_28]] : tensor<64xf64>
// CHECK-MIR: }
@@ -114,20 +114,20 @@
// CHECK-LIR: %[[VAL_6:.*]] = call @sparsePointers64(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
// CHECK-LIR: %[[VAL_7:.*]] = call @sparseIndices64(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
// CHECK-LIR: %[[VAL_8:.*]] = call @sparseValuesF64(%[[VAL_0]]) : (!llvm.ptr<i8>) -> memref<?xf64>
-// CHECK-LIR: %[[VAL_9:.*]] = alloc() : memref<64xf64>
+// CHECK-LIR: %[[VAL_9:.*]] = memref.alloc() : memref<64xf64>
// CHECK-LIR: scf.for %[[VAL_10:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-LIR: %[[VAL_11:.*]] = load %[[VAL_2]]{{\[}}%[[VAL_10]]] : memref<64xf64>
+// CHECK-LIR: %[[VAL_11:.*]] = memref.load %[[VAL_2]]{{\[}}%[[VAL_10]]] : memref<64xf64>
// CHECK-LIR: store %[[VAL_11]], %[[VAL_9]]{{\[}}%[[VAL_10]]] : memref<64xf64>
// CHECK-LIR: }
// CHECK-LIR: scf.for %[[VAL_12:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-LIR: %[[VAL_13:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
+// CHECK-LIR: %[[VAL_13:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_12]]] : memref<?xindex>
// CHECK-LIR: %[[VAL_14:.*]] = addi %[[VAL_12]], %[[VAL_5]] : index
-// CHECK-LIR: %[[VAL_15:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
-// CHECK-LIR: %[[VAL_16:.*]] = load %[[VAL_9]]{{\[}}%[[VAL_12]]] : memref<64xf64>
+// CHECK-LIR: %[[VAL_15:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_14]]] : memref<?xindex>
+// CHECK-LIR: %[[VAL_16:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_12]]] : memref<64xf64>
// CHECK-LIR: %[[VAL_17:.*]] = scf.for %[[VAL_18:.*]] = %[[VAL_13]] to %[[VAL_15]] step %[[VAL_5]] iter_args(%[[VAL_19:.*]] = %[[VAL_16]]) -> (f64) {
-// CHECK-LIR: %[[VAL_20:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
-// CHECK-LIR: %[[VAL_21:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf64>
-// CHECK-LIR: %[[VAL_22:.*]] = load %[[VAL_1]]{{\[}}%[[VAL_20]]] : memref<64xf64>
+// CHECK-LIR: %[[VAL_20:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_18]]] : memref<?xindex>
+// CHECK-LIR: %[[VAL_21:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_18]]] : memref<?xf64>
+// CHECK-LIR: %[[VAL_22:.*]] = memref.load %[[VAL_1]]{{\[}}%[[VAL_20]]] : memref<64xf64>
// CHECK-LIR: %[[VAL_23:.*]] = mulf %[[VAL_21]], %[[VAL_22]] : f64
// CHECK-LIR: %[[VAL_24:.*]] = addf %[[VAL_19]], %[[VAL_23]] : f64
// CHECK-LIR: scf.yield %[[VAL_24]] : f64
@@ -148,14 +148,14 @@
// CHECK-FAST: %[[VAL_7:.*]] = call @sparseIndices64(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
// CHECK-FAST: %[[VAL_8:.*]] = call @sparseValuesF64(%[[VAL_0]]) : (!llvm.ptr<i8>) -> memref<?xf64>
// CHECK-FAST: scf.for %[[VAL_9:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
-// CHECK-FAST: %[[VAL_10:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<?xindex>
+// CHECK-FAST: %[[VAL_10:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<?xindex>
// CHECK-FAST: %[[VAL_11:.*]] = addi %[[VAL_9]], %[[VAL_5]] : index
-// CHECK-FAST: %[[VAL_12:.*]] = load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
-// CHECK-FAST: %[[VAL_13:.*]] = load %[[VAL_2]]{{\[}}%[[VAL_9]]] : memref<64xf64>
+// CHECK-FAST: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
+// CHECK-FAST: %[[VAL_13:.*]] = memref.load %[[VAL_2]]{{\[}}%[[VAL_9]]] : memref<64xf64>
// CHECK-FAST: %[[VAL_14:.*]] = scf.for %[[VAL_15:.*]] = %[[VAL_10]] to %[[VAL_12]] step %[[VAL_5]] iter_args(%[[VAL_16:.*]] = %[[VAL_13]]) -> (f64) {
-// CHECK-FAST: %[[VAL_17:.*]] = load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
-// CHECK-FAST: %[[VAL_18:.*]] = load %[[VAL_8]]{{\[}}%[[VAL_15]]] : memref<?xf64>
-// CHECK-FAST: %[[VAL_19:.*]] = load %[[VAL_1]]{{\[}}%[[VAL_17]]] : memref<64xf64>
+// CHECK-FAST: %[[VAL_17:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
+// CHECK-FAST: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_15]]] : memref<?xf64>
+// CHECK-FAST: %[[VAL_19:.*]] = memref.load %[[VAL_1]]{{\[}}%[[VAL_17]]] : memref<64xf64>
// CHECK-FAST: %[[VAL_20:.*]] = mulf %[[VAL_18]], %[[VAL_19]] : f64
// CHECK-FAST: %[[VAL_21:.*]] = addf %[[VAL_16]], %[[VAL_20]] : f64
// CHECK-FAST: scf.yield %[[VAL_21]] : f64
diff --git a/mlir/test/Dialect/Linalg/sparse_nd.mlir b/mlir/test/Dialect/Linalg/sparse_nd.mlir
index 62f3c5e4dded..56ab7de4f0f0 100644
--- a/mlir/test/Dialect/Linalg/sparse_nd.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_nd.mlir
@@ -33,14 +33,14 @@
// CHECK: %[[VAL_10:.*]] = constant 80 : index
// CHECK: %[[VAL_11:.*]] = constant 0 : index
// CHECK: %[[VAL_12:.*]] = constant 1 : index
-// CHECK: %[[VAL_13:.*]] = tensor_to_memref %[[VAL_0]] : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: %[[VAL_13:.*]] = memref.buffer_cast %[[VAL_0]] : memref<10x20x30x40x50x60x70x80xf32>
// CHECK: %[[VAL_14:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_3]] : tensor<10x20x30x40x50x60x70x80xf32> to memref<?xindex>
// CHECK: %[[VAL_15:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_3]] : tensor<10x20x30x40x50x60x70x80xf32> to memref<?xindex>
// CHECK: %[[VAL_16:.*]] = linalg.sparse_pointers %[[VAL_1]], %[[VAL_4]] : tensor<10x20x30x40x50x60x70x80xf32> to memref<?xindex>
// CHECK: %[[VAL_17:.*]] = linalg.sparse_indices %[[VAL_1]], %[[VAL_4]] : tensor<10x20x30x40x50x60x70x80xf32> to memref<?xindex>
// CHECK: %[[VAL_18:.*]] = linalg.sparse_values %[[VAL_1]] : tensor<10x20x30x40x50x60x70x80xf32> to memref<?xf32>
-// CHECK: %[[VAL_19:.*]] = tensor_to_memref %[[VAL_2]] : memref<10x20x30x40x50x60x70x80xf32>
-// CHECK: %[[VAL_20:.*]] = alloc() : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: %[[VAL_19:.*]] = memref.buffer_cast %[[VAL_2]] : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: %[[VAL_20:.*]] = memref.alloc() : memref<10x20x30x40x50x60x70x80xf32>
// CHECK: linalg.copy(%[[VAL_19]], %[[VAL_20]]) : memref<10x20x30x40x50x60x70x80xf32>, memref<10x20x30x40x50x60x70x80xf32>
// CHECK: scf.for %[[VAL_21:.*]] = %[[VAL_11]] to %[[VAL_10]] step %[[VAL_12]] {
// CHECK: scf.for %[[VAL_22:.*]] = %[[VAL_11]] to %[[VAL_9]] step %[[VAL_12]] {
@@ -49,16 +49,16 @@
// CHECK: scf.for %[[VAL_25:.*]] = %[[VAL_11]] to %[[VAL_8]] step %[[VAL_12]] {
// CHECK: %[[VAL_26:.*]] = muli %[[VAL_24]], %[[VAL_8]] : index
// CHECK: %[[VAL_27:.*]] = addi %[[VAL_26]], %[[VAL_25]] : index
-// CHECK: %[[VAL_28:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_27]]] : memref<?xindex>
+// CHECK: %[[VAL_28:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_27]]] : memref<?xindex>
// CHECK: %[[VAL_29:.*]] = addi %[[VAL_27]], %[[VAL_12]] : index
-// CHECK: %[[VAL_30:.*]] = load %[[VAL_14]]{{\[}}%[[VAL_29]]] : memref<?xindex>
+// CHECK: %[[VAL_30:.*]] = memref.load %[[VAL_14]]{{\[}}%[[VAL_29]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_31:.*]] = %[[VAL_28]] to %[[VAL_30]] step %[[VAL_12]] {
-// CHECK: %[[VAL_32:.*]] = load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xindex>
-// CHECK: %[[VAL_33:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_31]]] : memref<?xindex>
+// CHECK: %[[VAL_32:.*]] = memref.load %[[VAL_15]]{{\[}}%[[VAL_31]]] : memref<?xindex>
+// CHECK: %[[VAL_33:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_31]]] : memref<?xindex>
// CHECK: %[[VAL_34:.*]] = addi %[[VAL_31]], %[[VAL_12]] : index
-// CHECK: %[[VAL_35:.*]] = load %[[VAL_16]]{{\[}}%[[VAL_34]]] : memref<?xindex>
+// CHECK: %[[VAL_35:.*]] = memref.load %[[VAL_16]]{{\[}}%[[VAL_34]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_36:.*]] = %[[VAL_33]] to %[[VAL_35]] step %[[VAL_12]] {
-// CHECK: %[[VAL_37:.*]] = load %[[VAL_17]]{{\[}}%[[VAL_36]]] : memref<?xindex>
+// CHECK: %[[VAL_37:.*]] = memref.load %[[VAL_17]]{{\[}}%[[VAL_36]]] : memref<?xindex>
// CHECK: scf.for %[[VAL_38:.*]] = %[[VAL_11]] to %[[VAL_7]] step %[[VAL_12]] {
// CHECK: %[[VAL_39:.*]] = muli %[[VAL_36]], %[[VAL_7]] : index
// CHECK: %[[VAL_40:.*]] = addi %[[VAL_39]], %[[VAL_38]] : index
@@ -68,10 +68,10 @@
// CHECK: scf.for %[[VAL_44:.*]] = %[[VAL_11]] to %[[VAL_5]] step %[[VAL_12]] {
// CHECK: %[[VAL_45:.*]] = muli %[[VAL_43]], %[[VAL_5]] : index
// CHECK: %[[VAL_46:.*]] = addi %[[VAL_45]], %[[VAL_44]] : index
-// CHECK: %[[VAL_47:.*]] = load %[[VAL_13]]{{\[}}%[[VAL_44]], %[[VAL_41]], %[[VAL_38]], %[[VAL_37]], %[[VAL_32]], %[[VAL_25]], %[[VAL_22]], %[[VAL_21]]] : memref<10x20x30x40x50x60x70x80xf32>
-// CHECK: %[[VAL_48:.*]] = load %[[VAL_18]]{{\[}}%[[VAL_46]]] : memref<?xf32>
+// CHECK: %[[VAL_47:.*]] = memref.load %[[VAL_13]]{{\[}}%[[VAL_44]], %[[VAL_41]], %[[VAL_38]], %[[VAL_37]], %[[VAL_32]], %[[VAL_25]], %[[VAL_22]], %[[VAL_21]]] : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: %[[VAL_48:.*]] = memref.load %[[VAL_18]]{{\[}}%[[VAL_46]]] : memref<?xf32>
// CHECK: %[[VAL_49:.*]] = mulf %[[VAL_47]], %[[VAL_48]] : f32
-// CHECK: store %[[VAL_49]], %[[VAL_20]]{{\[}}%[[VAL_44]], %[[VAL_41]], %[[VAL_38]], %[[VAL_37]], %[[VAL_32]], %[[VAL_25]], %[[VAL_22]], %[[VAL_21]]] : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: memref.store %[[VAL_49]], %[[VAL_20]]{{\[}}%[[VAL_44]], %[[VAL_41]], %[[VAL_38]], %[[VAL_37]], %[[VAL_32]], %[[VAL_25]], %[[VAL_22]], %[[VAL_21]]] : memref<10x20x30x40x50x60x70x80xf32>
// CHECK: }
// CHECK: }
// CHECK: }
@@ -80,7 +80,7 @@
// CHECK: }
// CHECK: }
// CHECK: }
-// CHECK: %[[VAL_50:.*]] = tensor_load %[[VAL_20]] : memref<10x20x30x40x50x60x70x80xf32>
+// CHECK: %[[VAL_50:.*]] = memref.tensor_load %[[VAL_20]] : memref<10x20x30x40x50x60x70x80xf32>
// CHECK: return %[[VAL_50]] : tensor<10x20x30x40x50x60x70x80xf32>
// CHECK: }
func @mul(%arga: tensor<10x20x30x40x50x60x70x80xf32>,
diff --git a/mlir/test/Dialect/Linalg/sparse_storage.mlir b/mlir/test/Dialect/Linalg/sparse_storage.mlir
index cc1207ec2204..81127513ea35 100644
--- a/mlir/test/Dialect/Linalg/sparse_storage.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_storage.mlir
@@ -29,15 +29,15 @@
// CHECK-TYPE0-LABEL: func @mul_dd(
// CHECK-TYPE0: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE0: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE0: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi64>
+// CHECK-TYPE0: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi64>
// CHECK-TYPE0: %[[B0:.*]] = index_cast %[[P0]] : i64 to index
-// CHECK-TYPE0: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi64>
+// CHECK-TYPE0: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi64>
// CHECK-TYPE0: %[[B1:.*]] = index_cast %[[P1]] : i64 to index
// CHECK-TYPE0: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE0: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi64>
+// CHECK-TYPE0: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi64>
// CHECK-TYPE0: %[[INDC:.*]] = index_cast %[[IND0]] : i64 to index
-// CHECK-TYPE0: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE0: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE0: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE0: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE0: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE0: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE0: }
@@ -45,15 +45,15 @@
// CHECK-TYPE1-LABEL: func @mul_dd(
// CHECK-TYPE1: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE1: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE1: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi64>
+// CHECK-TYPE1: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi64>
// CHECK-TYPE1: %[[B0:.*]] = index_cast %[[P0]] : i64 to index
-// CHECK-TYPE1: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi64>
+// CHECK-TYPE1: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi64>
// CHECK-TYPE1: %[[B1:.*]] = index_cast %[[P1]] : i64 to index
// CHECK-TYPE1: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE1: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi32>
+// CHECK-TYPE1: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi32>
// CHECK-TYPE1: %[[INDC:.*]] = index_cast %[[IND0]] : i32 to index
-// CHECK-TYPE1: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE1: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE1: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE1: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE1: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE1: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE1: }
@@ -61,15 +61,15 @@
// CHECK-TYPE2-LABEL: func @mul_dd(
// CHECK-TYPE2: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE2: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE2: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi32>
+// CHECK-TYPE2: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi32>
// CHECK-TYPE2: %[[B0:.*]] = index_cast %[[P0]] : i32 to index
-// CHECK-TYPE2: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi32>
+// CHECK-TYPE2: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi32>
// CHECK-TYPE2: %[[B1:.*]] = index_cast %[[P1]] : i32 to index
// CHECK-TYPE2: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE2: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi64>
+// CHECK-TYPE2: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi64>
// CHECK-TYPE2: %[[INDC:.*]] = index_cast %[[IND0]] : i64 to index
-// CHECK-TYPE2: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE2: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE2: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE2: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE2: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE2: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE2: }
@@ -77,15 +77,15 @@
// CHECK-TYPE3-LABEL: func @mul_dd(
// CHECK-TYPE3: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE3: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE3: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi32>
+// CHECK-TYPE3: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi32>
// CHECK-TYPE3: %[[B0:.*]] = index_cast %[[P0]] : i32 to index
-// CHECK-TYPE3: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi32>
+// CHECK-TYPE3: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi32>
// CHECK-TYPE3: %[[B1:.*]] = index_cast %[[P1]] : i32 to index
// CHECK-TYPE3: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE3: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi32>
+// CHECK-TYPE3: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi32>
// CHECK-TYPE3: %[[INDC:.*]] = index_cast %[[IND0]] : i32 to index
-// CHECK-TYPE3: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE3: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE3: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE3: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE3: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE3: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE3: }
@@ -93,15 +93,15 @@
// CHECK-TYPE4-LABEL: func @mul_dd(
// CHECK-TYPE4: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE4: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE4: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi16>
+// CHECK-TYPE4: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi16>
// CHECK-TYPE4: %[[B0:.*]] = index_cast %[[P0]] : i16 to index
-// CHECK-TYPE4: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi16>
+// CHECK-TYPE4: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi16>
// CHECK-TYPE4: %[[B1:.*]] = index_cast %[[P1]] : i16 to index
// CHECK-TYPE4: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE4: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi16>
+// CHECK-TYPE4: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi16>
// CHECK-TYPE4: %[[INDC:.*]] = index_cast %[[IND0]] : i16 to index
-// CHECK-TYPE4: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE4: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE4: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE4: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE4: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE4: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE4: }
@@ -109,15 +109,15 @@
// CHECK-TYPE5-LABEL: func @mul_dd(
// CHECK-TYPE5: %[[C0:.*]] = constant 0 : index
// CHECK-TYPE5: %[[C1:.*]] = constant 1 : index
-// CHECK-TYPE5: %[[P0:.*]] = load %{{.*}}[%[[C0]]] : memref<?xi8>
+// CHECK-TYPE5: %[[P0:.*]] = memref.load %{{.*}}[%[[C0]]] : memref<?xi8>
// CHECK-TYPE5: %[[B0:.*]] = index_cast %[[P0]] : i8 to index
-// CHECK-TYPE5: %[[P1:.*]] = load %{{.*}}[%[[C1]]] : memref<?xi8>
+// CHECK-TYPE5: %[[P1:.*]] = memref.load %{{.*}}[%[[C1]]] : memref<?xi8>
// CHECK-TYPE5: %[[B1:.*]] = index_cast %[[P1]] : i8 to index
// CHECK-TYPE5: scf.for %[[I:.*]] = %[[B0]] to %[[B1]] step %[[C1]] {
-// CHECK-TYPE5: %[[IND0:.*]] = load %{{.*}}[%[[I]]] : memref<?xi8>
+// CHECK-TYPE5: %[[IND0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xi8>
// CHECK-TYPE5: %[[INDC:.*]] = index_cast %[[IND0]] : i8 to index
-// CHECK-TYPE5: %[[VAL0:.*]] = load %{{.*}}[%[[I]]] : memref<?xf64>
-// CHECK-TYPE5: %[[VAL1:.*]] = load %{{.*}}[%[[INDC]]] : memref<32xf64>
+// CHECK-TYPE5: %[[VAL0:.*]] = memref.load %{{.*}}[%[[I]]] : memref<?xf64>
+// CHECK-TYPE5: %[[VAL1:.*]] = memref.load %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE5: %[[MUL:.*]] = mulf %[[VAL0]], %[[VAL1]] : f64
// CHECK-TYPE5: store %[[MUL]], %{{.*}}[%[[INDC]]] : memref<32xf64>
// CHECK-TYPE5: }
diff --git a/mlir/test/Dialect/Linalg/sparse_vector.mlir b/mlir/test/Dialect/Linalg/sparse_vector.mlir
index 5409e4b80681..882ef077cb01 100644
--- a/mlir/test/Dialect/Linalg/sparse_vector.mlir
+++ b/mlir/test/Dialect/Linalg/sparse_vector.mlir
@@ -24,7 +24,7 @@
// CHECK-VEC0-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC0-DAG: %[[c1024:.*]] = constant 1024 : index
// CHECK-VEC0: scf.for %[[i:.*]] = %[[c0]] to %[[c1024]] step %[[c1]] {
-// CHECK-VEC0: %[[l:.*]] = load %{{.*}}[%[[i]]] : memref<1024xf32>
+// CHECK-VEC0: %[[l:.*]] = memref.load %{{.*}}[%[[i]]] : memref<1024xf32>
// CHECK-VEC0: %[[m:.*]] = mulf %[[l]], %{{.*}} : f32
// CHECK-VEC0: store %[[m]], %{{.*}}[%[[i]]] : memref<1024xf32>
// CHECK-VEC0: }
@@ -84,15 +84,15 @@ func @scale_d(%arga: tensor<1024xf32>, %scale: f32, %argx: tensor<1024xf32>) ->
// CHECK-VEC0-LABEL: func @mul_s
// CHECK-VEC0-DAG: %[[c0:.*]] = constant 0 : index
// CHECK-VEC0-DAG: %[[c1:.*]] = constant 1 : index
-// CHECK-VEC0: %[[p:.*]] = load %{{.*}}[%[[c0]]] : memref<?xi32>
+// CHECK-VEC0: %[[p:.*]] = memref.load %{{.*}}[%[[c0]]] : memref<?xi32>
// CHECK-VEC0: %[[q:.*]] = index_cast %[[p]] : i32 to index
-// CHECK-VEC0: %[[r:.*]] = load %{{.*}}[%[[c1]]] : memref<?xi32>
+// CHECK-VEC0: %[[r:.*]] = memref.load %{{.*}}[%[[c1]]] : memref<?xi32>
// CHECK-VEC0: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC0: scf.for %[[i:.*]] = %[[q]] to %[[s]] step %[[c1]] {
-// CHECK-VEC0: %[[li:.*]] = load %{{.*}}[%[[i]]] : memref<?xi32>
+// CHECK-VEC0: %[[li:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xi32>
// CHECK-VEC0: %[[ci:.*]] = index_cast %[[li]] : i32 to index
-// CHECK-VEC0: %[[la:.*]] = load %{{.*}}[%[[i]]] : memref<?xf32>
-// CHECK-VEC0: %[[lb:.*]] = load %{{.*}}[%[[ci]]] : memref<1024xf32>
+// CHECK-VEC0: %[[la:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xf32>
+// CHECK-VEC0: %[[lb:.*]] = memref.load %{{.*}}[%[[ci]]] : memref<1024xf32>
// CHECK-VEC0: %[[m:.*]] = mulf %[[la]], %[[lb]] : f32
// CHECK-VEC0: store %[[m]], %{{.*}}[%[[ci]]] : memref<1024xf32>
// CHECK-VEC0: }
@@ -101,15 +101,15 @@ func @scale_d(%arga: tensor<1024xf32>, %scale: f32, %argx: tensor<1024xf32>) ->
// CHECK-VEC1-LABEL: func @mul_s
// CHECK-VEC1-DAG: %[[c0:.*]] = constant 0 : index
// CHECK-VEC1-DAG: %[[c1:.*]] = constant 1 : index
-// CHECK-VEC1: %[[p:.*]] = load %{{.*}}[%[[c0]]] : memref<?xi32>
+// CHECK-VEC1: %[[p:.*]] = memref.load %{{.*}}[%[[c0]]] : memref<?xi32>
// CHECK-VEC1: %[[q:.*]] = index_cast %[[p]] : i32 to index
-// CHECK-VEC1: %[[r:.*]] = load %{{.*}}[%[[c1]]] : memref<?xi32>
+// CHECK-VEC1: %[[r:.*]] = memref.load %{{.*}}[%[[c1]]] : memref<?xi32>
// CHECK-VEC1: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC1: scf.for %[[i:.*]] = %[[q]] to %[[s]] step %[[c1]] {
-// CHECK-VEC1: %[[li:.*]] = load %{{.*}}[%[[i]]] : memref<?xi32>
+// CHECK-VEC1: %[[li:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xi32>
// CHECK-VEC1: %[[ci:.*]] = index_cast %[[li]] : i32 to index
-// CHECK-VEC1: %[[la:.*]] = load %{{.*}}[%[[i]]] : memref<?xf32>
-// CHECK-VEC1: %[[lb:.*]] = load %{{.*}}[%[[ci]]] : memref<1024xf32>
+// CHECK-VEC1: %[[la:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xf32>
+// CHECK-VEC1: %[[lb:.*]] = memref.load %{{.*}}[%[[ci]]] : memref<1024xf32>
// CHECK-VEC1: %[[m:.*]] = mulf %[[la]], %[[lb]] : f32
// CHECK-VEC1: store %[[m]], %{{.*}}[%[[ci]]] : memref<1024xf32>
// CHECK-VEC1: }
@@ -119,9 +119,9 @@ func @scale_d(%arga: tensor<1024xf32>, %scale: f32, %argx: tensor<1024xf32>) ->
// CHECK-VEC2-DAG: %[[c0:.*]] = constant 0 : index
// CHECK-VEC2-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC2-DAG: %[[c16:.*]] = constant 16 : index
-// CHECK-VEC2: %[[p:.*]] = load %{{.*}}[%[[c0]]] : memref<?xi32>
+// CHECK-VEC2: %[[p:.*]] = memref.load %{{.*}}[%[[c0]]] : memref<?xi32>
// CHECK-VEC2: %[[q:.*]] = index_cast %[[p]] : i32 to index
-// CHECK-VEC2: %[[r:.*]] = load %{{.*}}[%[[c1]]] : memref<?xi32>
+// CHECK-VEC2: %[[r:.*]] = memref.load %{{.*}}[%[[c1]]] : memref<?xi32>
// CHECK-VEC2: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC2: scf.for %[[i:.*]] = %[[q]] to %[[s]] step %[[c16]] {
// CHECK-VEC2: %[[sub:.*]] = subi %[[s]], %[[i]] : index
@@ -150,9 +150,9 @@ func @mul_s(%arga: tensor<1024xf32>, %argb: tensor<1024xf32>, %argx: tensor<1024
// CHECK-VEC2-DAG: %[[c0:.*]] = constant 0 : index
// CHECK-VEC2-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC2-DAG: %[[c16:.*]] = constant 16 : index
-// CHECK-VEC2: %[[p:.*]] = load %{{.*}}[%[[c0]]] : memref<?xi32>
+// CHECK-VEC2: %[[p:.*]] = memref.load %{{.*}}[%[[c0]]] : memref<?xi32>
// CHECK-VEC2: %[[q:.*]] = index_cast %[[p]] : i32 to index
-// CHECK-VEC2: %[[r:.*]] = load %{{.*}}[%[[c1]]] : memref<?xi32>
+// CHECK-VEC2: %[[r:.*]] = memref.load %{{.*}}[%[[c1]]] : memref<?xi32>
// CHECK-VEC2: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC2: scf.for %[[i:.*]] = %[[q]] to %[[s]] step %[[c16]] {
// CHECK-VEC2: %[[sub:.*]] = subi %[[s]], %[[i]] : index
@@ -200,8 +200,8 @@ func @mul_s_alt(%argA: !SparseTensor, %argB: !SparseTensor, %argx: tensor<1024xf
// CHECK-VEC0-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC0-DAG: %[[c1024:.*]] = constant 1024 : index
// CHECK-VEC0: %[[red:.*]] = scf.for %[[i:.*]] = %[[c0]] to %[[c1024]] step %[[c1]] iter_args(%[[red_in:.*]] = %{{.*}}) -> (f32) {
-// CHECK-VEC0: %[[la:.*]] = load %{{.*}}[%[[i]]] : memref<1024xf32>
-// CHECK-VEC0: %[[lb:.*]] = load %{{.*}}[%[[i]]] : memref<1024xf32>
+// CHECK-VEC0: %[[la:.*]] = memref.load %{{.*}}[%[[i]]] : memref<1024xf32>
+// CHECK-VEC0: %[[lb:.*]] = memref.load %{{.*}}[%[[i]]] : memref<1024xf32>
// CHECK-VEC0: %[[m:.*]] = mulf %[[la]], %[[lb]] : f32
// CHECK-VEC0: %[[a:.*]] = addf %[[red_in]], %[[m]] : f32
// CHECK-VEC0: scf.yield %[[a]] : f32
@@ -302,16 +302,16 @@ func @reduction_17(%arga: tensor<17xf32>, %argb: tensor<17xf32>, %argx: tensor<f
// CHECK-VEC0-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC0-DAG: %[[c512:.*]] = constant 512 : index
// CHECK-VEC0: scf.for %[[i:.*]] = %[[c0]] to %[[c512]] step %[[c1]] {
-// CHECK-VEC0: %[[p:.*]] = load %{{.*}}[%[[i]]] : memref<?xi32>
+// CHECK-VEC0: %[[p:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xi32>
// CHECK-VEC0: %[[q:.*]] = index_cast %[[p]] : i32 to index
// CHECK-VEC0: %[[a:.*]] = addi %[[i]], %[[c1]] : index
-// CHECK-VEC0: %[[r:.*]] = load %{{.*}}[%[[a]]] : memref<?xi32>
+// CHECK-VEC0: %[[r:.*]] = memref.load %{{.*}}[%[[a]]] : memref<?xi32>
// CHECK-VEC0: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC0: scf.for %[[j:.*]] = %[[q]] to %[[s]] step %[[c1]] {
-// CHECK-VEC0: %[[lj:.*]] = load %{{.*}}[%[[j]]] : memref<?xi32>
+// CHECK-VEC0: %[[lj:.*]] = memref.load %{{.*}}[%[[j]]] : memref<?xi32>
// CHECK-VEC0: %[[cj:.*]] = index_cast %[[lj]] : i32 to index
-// CHECK-VEC0: %[[la:.*]] = load %{{.*}}[%[[j]]] : memref<?xf32>
-// CHECK-VEC0: %[[lb:.*]] = load %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
+// CHECK-VEC0: %[[la:.*]] = memref.load %{{.*}}[%[[j]]] : memref<?xf32>
+// CHECK-VEC0: %[[lb:.*]] = memref.load %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
// CHECK-VEC0: %[[m:.*]] = mulf %[[la]], %[[lb]] : f32
// CHECK-VEC0: store %[[m]], %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
// CHECK-VEC0: }
@@ -323,16 +323,16 @@ func @reduction_17(%arga: tensor<17xf32>, %argb: tensor<17xf32>, %argx: tensor<f
// CHECK-VEC1-DAG: %[[c1:.*]] = constant 1 : index
// CHECK-VEC1-DAG: %[[c512:.*]] = constant 512 : index
// CHECK-VEC1: scf.for %[[i:.*]] = %[[c0]] to %[[c512]] step %[[c1]] {
-// CHECK-VEC1: %[[p:.*]] = load %{{.*}}[%[[i]]] : memref<?xi32>
+// CHECK-VEC1: %[[p:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xi32>
// CHECK-VEC1: %[[q:.*]] = index_cast %[[p]] : i32 to index
// CHECK-VEC1: %[[a:.*]] = addi %[[i]], %[[c1]] : index
-// CHECK-VEC1: %[[r:.*]] = load %{{.*}}[%[[a]]] : memref<?xi32>
+// CHECK-VEC1: %[[r:.*]] = memref.load %{{.*}}[%[[a]]] : memref<?xi32>
// CHECK-VEC1: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC1: scf.for %[[j:.*]] = %[[q]] to %[[s]] step %[[c1]] {
-// CHECK-VEC1: %[[lj:.*]] = load %{{.*}}[%[[j]]] : memref<?xi32>
+// CHECK-VEC1: %[[lj:.*]] = memref.load %{{.*}}[%[[j]]] : memref<?xi32>
// CHECK-VEC1: %[[cj:.*]] = index_cast %[[lj]] : i32 to index
-// CHECK-VEC1: %[[la:.*]] = load %{{.*}}[%[[j]]] : memref<?xf32>
-// CHECK-VEC1: %[[lb:.*]] = load %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
+// CHECK-VEC1: %[[la:.*]] = memref.load %{{.*}}[%[[j]]] : memref<?xf32>
+// CHECK-VEC1: %[[lb:.*]] = memref.load %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
// CHECK-VEC1: %[[m:.*]] = mulf %[[la]], %[[lb]] : f32
// CHECK-VEC1: store %[[m]], %{{.*}}[%[[i]], %[[cj]]] : memref<512x1024xf32>
// CHECK-VEC1: }
@@ -345,10 +345,10 @@ func @reduction_17(%arga: tensor<17xf32>, %argb: tensor<17xf32>, %argx: tensor<f
// CHECK-VEC2-DAG: %[[c16:.*]] = constant 16 : index
// CHECK-VEC2-DAG: %[[c512:.*]] = constant 512 : index
// CHECK-VEC2: scf.for %[[i:.*]] = %[[c0]] to %[[c512]] step %[[c1]] {
-// CHECK-VEC2: %[[p:.*]] = load %{{.*}}[%[[i]]] : memref<?xi32>
+// CHECK-VEC2: %[[p:.*]] = memref.load %{{.*}}[%[[i]]] : memref<?xi32>
// CHECK-VEC2: %[[q:.*]] = index_cast %[[p]] : i32 to index
// CHECK-VEC2: %[[a:.*]] = addi %[[i]], %[[c1]] : index
-// CHECK-VEC2: %[[r:.*]] = load %{{.*}}[%[[a]]] : memref<?xi32>
+// CHECK-VEC2: %[[r:.*]] = memref.load %{{.*}}[%[[a]]] : memref<?xi32>
// CHECK-VEC2: %[[s:.*]] = index_cast %[[r]] : i32 to index
// CHECK-VEC2: scf.for %[[j:.*]] = %[[q]] to %[[s]] step %[[c16]] {
// CHECK-VEC2: %[[sub:.*]] = subi %[[s]], %[[j]] : index
diff --git a/mlir/test/Dialect/Linalg/standard.mlir b/mlir/test/Dialect/Linalg/standard.mlir
index 15cfca2ea679..27687aaf60b3 100644
--- a/mlir/test/Dialect/Linalg/standard.mlir
+++ b/mlir/test/Dialect/Linalg/standard.mlir
@@ -20,11 +20,11 @@ func @dot(%arg0: memref<?xf32, offset: ?, strides: [1]>,
// CHECK-SAME: %[[arg0:[a-zA-z0-9]*]]: memref<?xf32, #[[$map0]]>,
// CHECK-SAME: %[[arg1:[a-zA-z0-9]*]]: memref<?xf32, #[[$map0]]>,
// CHECK-SAME: %[[arg2:[a-zA-z0-9]*]]: memref<f32>) {
-// CHECK: %[[o0:.*]] = memref_cast %[[arg0]] :
+// CHECK: %[[o0:.*]] = memref.cast %[[arg0]] :
// CHECK-SAME: memref<?xf32, #[[$map0]]> to memref<?xf32, #[[$map6]]>
-// CHECK: %[[o1:.*]] = memref_cast %[[arg1]] :
+// CHECK: %[[o1:.*]] = memref.cast %[[arg1]] :
// CHECK-SAME: memref<?xf32, #[[$map0]]> to memref<?xf32, #[[$map6]]>
-// CHECK: %[[o2:.*]] = memref_cast %[[arg2]] :
+// CHECK: %[[o2:.*]] = memref.cast %[[arg2]] :
// CHECK-SAME: memref<f32> to memref<f32, #[[$map7]]>
// CHECK: call @linalg_dot_viewsxf32_viewsxf32_viewf32(
// CHECK-SAME: %[[o0]], %[[o1]], %[[o2]]) :
@@ -37,9 +37,9 @@ func @copy(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %arg1: memre
// CHECK-LABEL: func @copy(
// CHECK-SAME: %[[arg0:[a-zA-z0-9]*]]: memref<?x?x?xf32, #[[$map1]]>,
// CHECK-SAME: %[[arg1:[a-zA-z0-9]*]]: memref<?x?x?xf32, #[[$map1]]>) {
-// CHECK: %[[o0:.*]] = memref_cast %[[arg0]] :
+// CHECK: %[[o0:.*]] = memref.cast %[[arg0]] :
// CHECK-SAME: memref<?x?x?xf32, #[[$map1]]> to memref<?x?x?xf32, #[[$map8]]>
-// CHECK: %[[o1:.*]] = memref_cast %[[arg1]] :
+// CHECK: %[[o1:.*]] = memref.cast %[[arg1]] :
// CHECK-SAME: memref<?x?x?xf32, #[[$map1]]> to memref<?x?x?xf32, #[[$map8]]>
// CHECK: call @linalg_copy_viewsxsxsxf32_viewsxsxsxf32(%[[o0]], %[[o1]]) :
// CHECK-SAME: memref<?x?x?xf32, #[[$map8]]>, memref<?x?x?xf32, #[[$map8]]>
@@ -53,13 +53,13 @@ func @copy_transpose(%arg0: memref<?x?x?xf32, offset: ?, strides: [?, ?, 1]>, %a
// CHECK-LABEL: func @copy_transpose(
// CHECK-SAME: %[[arg0:[a-zA-z0-9]*]]: memref<?x?x?xf32, #[[$map1]]>,
// CHECK-SAME: %[[arg1:[a-zA-z0-9]*]]: memref<?x?x?xf32, #[[$map1]]>) {
-// CHECK: %[[t0:.*]] = transpose %[[arg0]]
+// CHECK: %[[t0:.*]] = memref.transpose %[[arg0]]
// CHECK-SAME: (d0, d1, d2) -> (d0, d2, d1) : memref<?x?x?xf32, #[[$map1]]>
-// CHECK: %[[t1:.*]] = transpose %[[arg1]]
+// CHECK: %[[t1:.*]] = memref.transpose %[[arg1]]
// CHECK-SAME: (d0, d1, d2) -> (d2, d1, d0) : memref<?x?x?xf32, #[[$map1]]>
-// CHECK: %[[o0:.*]] = memref_cast %[[t0]] :
+// CHECK: %[[o0:.*]] = memref.cast %[[t0]] :
// CHECK-SAME: memref<?x?x?xf32, #[[$map2]]> to memref<?x?x?xf32, #[[$map8]]>
-// CHECK: %[[o1:.*]] = memref_cast %[[t1]] :
+// CHECK: %[[o1:.*]] = memref.cast %[[t1]] :
// CHECK-SAME: memref<?x?x?xf32, #[[$map4]]> to memref<?x?x?xf32, #[[$map8]]>
// CHECK: call @linalg_copy_viewsxsxsxf32_viewsxsxsxf32(%[[o0]], %[[o1]]) :
// CHECK-SAME: memref<?x?x?xf32, #[[$map8]]>, memref<?x?x?xf32, #[[$map8]]>
diff --git a/mlir/test/Dialect/Linalg/tile-and-distribute.mlir b/mlir/test/Dialect/Linalg/tile-and-distribute.mlir
index d566701d7bb6..6f8c2b0ce76e 100644
--- a/mlir/test/Dialect/Linalg/tile-and-distribute.mlir
+++ b/mlir/test/Dialect/Linalg/tile-and-distribute.mlir
@@ -16,12 +16,12 @@ func @gemm1(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: %[[BIDX:.*]] = "gpu.block_id"() {dimension = "x"}
// CHECK: scf.for %[[ARG3:.*]] =
// CHECK: %[[OFFSETY:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
// CHECK: %[[OFFSETX:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
// CHECK: %[[OFFSETY_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
// CHECK: %[[OFFSETX:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -48,12 +48,12 @@ func @gemm2(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: scf.if %[[INBOUNDS]]
// CHECK: scf.for %[[ARG3:.*]] =
// CHECK: %[[OFFSETY:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
// CHECK: %[[OFFSETX:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
// CHECK: %[[OFFSETY_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
// CHECK: %[[OFFSETX_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX_2]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX_2]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -80,9 +80,9 @@ func @gemm3(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: %[[STEPX:.*]] = affine.apply #[[MAP0]]()[%[[NBLOCKSX]]]
// CHECK: scf.parallel (%[[ARG3:.*]], %[[ARG4:.*]]) = (%[[LBY]], %[[LBX]]) to (%{{.*}}, %{{.*}}) step (%[[STEPY]], %[[STEPX]])
// CHECK: scf.for %[[ARG5:.*]] =
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[ARG3]], %[[ARG5]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG5]], %[[ARG4]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[ARG3]], %[[ARG4]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[ARG3]], %[[ARG5]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG5]], %[[ARG4]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[ARG3]], %[[ARG4]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -106,12 +106,12 @@ func @gemm4(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: scf.if %[[INBOUNDS]]
// CHECK: scf.for %[[ARG3:.*]] =
// CHECK: %[[OFFSETY:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[OFFSETY]], %[[ARG3]]]
// CHECK: %[[OFFSETX:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG3]], %[[OFFSETX]]]
// CHECK: %[[OFFSETY_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
// CHECK: %[[OFFSETX_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX_2]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[OFFSETY_2]], %[[OFFSETX_2]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -139,10 +139,10 @@ func @gemm5(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: scf.parallel (%[[ARG3.*]]) = (%[[LBX]]) to (%{{.*}}) step (%[[STEPX]])
// CHECK: scf.for %[[ARG4:.*]] =
// CHECK: %[[OFFSETY:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[OFFSETY]], %[[ARG4]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG4]], %[[ARG3]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[OFFSETY]], %[[ARG4]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG4]], %[[ARG3]]]
// CHECK: %[[OFFSETY_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDY]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[OFFSETY_2]], %[[ARG3]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[OFFSETY_2]], %[[ARG3]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -166,11 +166,11 @@ func @gemm6(%a : memref<?x?xf32>, %b : memref<?x?xf32>, %c : memref<?x?xf32>)
// CHECK: %[[STEPY:.*]] = affine.apply #[[MAP0]]()[%[[NBLOCKSY]]]
// CHECK: scf.parallel (%[[ARG3.*]]) = (%[[LBY]]) to (%{{.*}}) step (%[[STEPY]])
// CHECK: scf.for %[[ARG4:.*]] =
-// CHECK: %[[SV1:.*]] = subview %[[ARG0]][%[[ARG3]], %[[ARG4]]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG0]][%[[ARG3]], %[[ARG4]]]
// CHECK: %[[OFFSETX:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG1]][%[[ARG4]], %[[OFFSETX]]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG1]][%[[ARG4]], %[[OFFSETX]]]
// CHECK: %[[OFFSETX_2:.*]] = affine.apply #[[MAP0]]()[%[[BIDX]]]
-// CHECK: %[[SV3:.*]] = subview %[[ARG2]][%[[ARG3]], %[[OFFSETX_2]]]
+// CHECK: %[[SV3:.*]] = memref.subview %[[ARG2]][%[[ARG3]], %[[OFFSETX_2]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
diff --git a/mlir/test/Dialect/Linalg/tile-and-fuse-tensors.mlir b/mlir/test/Dialect/Linalg/tile-and-fuse-tensors.mlir
index 9e9688088568..b742f8148dac 100644
--- a/mlir/test/Dialect/Linalg/tile-and-fuse-tensors.mlir
+++ b/mlir/test/Dialect/Linalg/tile-and-fuse-tensors.mlir
@@ -16,9 +16,9 @@ func @matmul_tensors(%arg0: tensor<?x?xf32>, %arg1: tensor<?x?xf32>, %arg2: tens
%c0 = constant 0 : index
%c3 = constant 3 : index
%c1 = constant 1 : index
- %0 = dim %t0, %c0 : tensor<?x?xf32>
- %1 = dim %t0, %c1 : tensor<?x?xf32>
- %2 = dim %arg1, %c1 : tensor<?x?xf32>
+ %0 = memref.dim %t0, %c0 : tensor<?x?xf32>
+ %1 = memref.dim %t0, %c1 : tensor<?x?xf32>
+ %2 = memref.dim %arg1, %c1 : tensor<?x?xf32>
%3 = scf.for %arg3 = %c0 to %0 step %c2 iter_args(%arg4 = %arg2) -> (tensor<?x?xf32>) {
%4 = scf.for %arg5 = %c0 to %2 step %c3 iter_args(%arg6 = %arg4) -> (tensor<?x?xf32>) {
%5 = scf.for %arg7 = %c0 to %1 step %c4 iter_args(%arg8 = %arg6) -> (tensor<?x?xf32>) {
@@ -42,7 +42,7 @@ func @matmul_tensors(%arg0: tensor<?x?xf32>, %arg1: tensor<?x?xf32>, %arg2: tens
// CHECK-SAME: %[[C:[0-9a-z]*]]: tensor<?x?xf32>
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
// CHECK-DAG: %[[C1:.*]] = constant 1 : index
-// CHECK-DAG: %[[dA1:.*]] = dim %[[A]], %[[C1]] : tensor<?x?xf32>
+// CHECK-DAG: %[[dA1:.*]] = memref.dim %[[A]], %[[C1]] : tensor<?x?xf32>
// CHECK: scf.for %[[I:[0-9a-z]*]]
// CHECK: %[[stA:.*]] = subtensor %[[A]][%[[I]], 0] [2, %[[dA1]]] [1, 1] : tensor<?x?xf32> to tensor<2x?xf32>
// CHECK-NEXT: scf.for %[[J:[0-9a-z]*]]
diff --git a/mlir/test/Dialect/Linalg/tile-conv-padding.mlir b/mlir/test/Dialect/Linalg/tile-conv-padding.mlir
index 262bfff17855..28b18e22d1de 100644
--- a/mlir/test/Dialect/Linalg/tile-conv-padding.mlir
+++ b/mlir/test/Dialect/Linalg/tile-conv-padding.mlir
@@ -21,18 +21,18 @@ func @conv_padding(%arg0: memref<?x?x?x?xf32, offset: ?, strides: [?, ?, ?, 1]>,
// TILE-20000-SAME: %[[ARG2:[a-zA-Z0-9_]*]]: memref<?x?x?x?xf32, #[[$strided4D]]>)
// TILE-20000-DAG: %[[C0:.*]] = constant 0 : index
// TILE-20000-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-20000: %[[B:.*]] = dim %[[ARG1]], %c0
+// TILE-20000: %[[B:.*]] = memref.dim %[[ARG1]], %c0
// TILE-20000: scf.for %[[ivI:.*]] = %[[C0]] to %[[B]] step %[[C2]] {
-// TILE-20000: %[[DIM10:.*]] = dim %[[ARG1]], %c0
+// TILE-20000: %[[DIM10:.*]] = memref.dim %[[ARG1]], %c0
// TILE-20000: %[[EXTENT:.*]] = affine.min #[[$minmap]](%[[ivI]])[%[[DIM10]]]
-// TILE-20000: %[[DIM11:.*]] = dim %[[ARG1]], %c1
-// TILE-20000: %[[DIM12:.*]] = dim %[[ARG1]], %c2
-// TILE-20000: %[[DIM13:.*]] = dim %[[ARG1]], %c3
-// TILE-20000: %[[SUBVIEW1:.*]] = subview %[[ARG1]][%[[ivI]], 0, 0, 0] [%[[EXTENT]], %[[DIM11]], %[[DIM12]], %[[DIM13]]]
-// TILE-20000: %[[DIM20:.*]] = dim %[[ARG2]], %c0
+// TILE-20000: %[[DIM11:.*]] = memref.dim %[[ARG1]], %c1
+// TILE-20000: %[[DIM12:.*]] = memref.dim %[[ARG1]], %c2
+// TILE-20000: %[[DIM13:.*]] = memref.dim %[[ARG1]], %c3
+// TILE-20000: %[[SUBVIEW1:.*]] = memref.subview %[[ARG1]][%[[ivI]], 0, 0, 0] [%[[EXTENT]], %[[DIM11]], %[[DIM12]], %[[DIM13]]]
+// TILE-20000: %[[DIM20:.*]] = memref.dim %[[ARG2]], %c0
// TILE-20000: %[[EXTENT:.*]] = affine.min #[[$minmap]](%[[ivI]])[%[[DIM20]]]
-// TILE-20000: %[[DIM21:.*]] = dim %[[ARG2]], %c1
-// TILE-20000: %[[DIM22:.*]] = dim %[[ARG2]], %c2
-// TILE-20000: %[[DIM23:.*]] = dim %[[ARG2]], %c3
-// TILE-20000: %[[SUBVIEW2:.*]] = subview %[[ARG2]][%[[ivI]], 0, 0, 0] [%[[EXTENT]], %[[DIM21]], %[[DIM22]], %[[DIM23]]]
+// TILE-20000: %[[DIM21:.*]] = memref.dim %[[ARG2]], %c1
+// TILE-20000: %[[DIM22:.*]] = memref.dim %[[ARG2]], %c2
+// TILE-20000: %[[DIM23:.*]] = memref.dim %[[ARG2]], %c3
+// TILE-20000: %[[SUBVIEW2:.*]] = memref.subview %[[ARG2]][%[[ivI]], 0, 0, 0] [%[[EXTENT]], %[[DIM21]], %[[DIM22]], %[[DIM23]]]
// TILE-20000: linalg.conv(%[[ARG0]], %[[SUBVIEW1]], %[[SUBVIEW2]])
diff --git a/mlir/test/Dialect/Linalg/tile-conv.mlir b/mlir/test/Dialect/Linalg/tile-conv.mlir
index 3b76f8a3139c..530214b8b8d1 100644
--- a/mlir/test/Dialect/Linalg/tile-conv.mlir
+++ b/mlir/test/Dialect/Linalg/tile-conv.mlir
@@ -17,30 +17,30 @@ func @conv(%arg0: memref<?x?x?x?xf32, offset: ?, strides: [?, ?, ?, 1]>, %arg1:
// TILE-23004-DAG: %[[C2:.*]] = constant 2 : index
// TILE-23004-DAG: %[[C3:.*]] = constant 3 : index
// TILE-23004-DAG: %[[C4:.*]] = constant 4 : index
-// TILE-23004: %[[Z0:.*]] = dim %[[ARG0]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[Q:.*]] = dim %[[ARG0]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[B:.*]] = dim %[[ARG1]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[X0:.*]] = dim %[[ARG2]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[Z0:.*]] = memref.dim %[[ARG0]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[Q:.*]] = memref.dim %[[ARG0]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[B:.*]] = memref.dim %[[ARG1]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[X0:.*]] = memref.dim %[[ARG2]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
// TILE-23004: scf.for %[[ivI:.*]] = %{{.*}} to %[[B]] step %{{.*}} {
// TILE-23004: scf.for %[[ivJ:.*]] = %{{.*}} to %[[X0]] step %{{.*}} {
// TILE-23004: scf.for %[[ivK:.*]] = %{{.*}} to %[[Q]] step %{{.*}} {
-// TILE-23004: %[[Z0_1:.*]] = dim %[[ARG0]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[Z1:.*]] = dim %[[ARG0]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[Z2:.*]] = dim %[[ARG0]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[Z0_1:.*]] = memref.dim %[[ARG0]], %c0 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[Z1:.*]] = memref.dim %[[ARG0]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[Z2:.*]] = memref.dim %[[ARG0]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
// TILE-23004: %[[szK:.*]] = affine.min #[[$bound_map_4]](%[[ivK]])[%[[Z2]]]
-// TILE-23004: %[[K:.*]] = dim %[[ARG0]], %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[FilterView:.*]] = subview %{{.*}}[0, 0, %[[ivK]], 0] [%[[Z0_1]], %[[Z1]], %[[szK]], %[[K]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[K:.*]] = memref.dim %[[ARG0]], %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[FilterView:.*]] = memref.subview %{{.*}}[0, 0, %[[ivK]], 0] [%[[Z0_1]], %[[Z1]], %[[szK]], %[[K]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
//
// TILE-23004: %[[J1:.*]] = affine.apply #[[$D0x30pS0x10]](%[[ivJ]])
-// TILE-23004: %[[PaddedInput0b:.*]] = dim %[[ARG1]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[PaddedInput0b:.*]] = memref.dim %[[ARG1]], %c1 : memref<?x?x?x?xf32, #[[$strided4D]]>
// TILE-23004: %[[I1pStep:.*]] = affine.min #[[$S0x10p90D0x30pS1]](%[[ivJ]])[%[[Z0]], %[[PaddedInput0b]]]
-// TILE-23004: %[[SZ2:.*]] = dim %[[ARG1]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[dim3:.*]] = dim %[[ARG1]], %c3
+// TILE-23004: %[[SZ2:.*]] = memref.dim %[[ARG1]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[dim3:.*]] = memref.dim %[[ARG1]], %c3
// TILE-23004: %[[sz3:.*]] = affine.min #[[$bound_map_4]](%[[ivK]])[%[[dim3]]]
-// TILE-23004: %[[InputView:.*]] = subview %{{.*}}[%[[ivI]], %[[J1]], 0, %[[ivK]]] [%{{.*}}, %{{.*}}, %[[SZ2]], %[[sz3]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[InputView:.*]] = memref.subview %{{.*}}[%[[ivI]], %[[J1]], 0, %[[ivK]]] [%{{.*}}, %{{.*}}, %[[SZ2]], %[[sz3]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
//
-// TILE-23004: %[[X0:.*]] = dim %[[ARG2]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[X1:.*]] = dim %[[ARG2]], %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
-// TILE-23004: %[[OutputView:.*]] = subview %{{.*}}[%[[ivI]], %[[ivJ]], 0, 0] [%{{.*}}, %{{.*}}, %[[X0]], %[[X1]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[X0:.*]] = memref.dim %[[ARG2]], %c2 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[X1:.*]] = memref.dim %[[ARG2]], %c3 : memref<?x?x?x?xf32, #[[$strided4D]]>
+// TILE-23004: %[[OutputView:.*]] = memref.subview %{{.*}}[%[[ivI]], %[[ivJ]], 0, 0] [%{{.*}}, %{{.*}}, %[[X0]], %[[X1]]] [1, 1, 1, 1] : memref<?x?x?x?xf32, #[[$strided4D]]> to memref<?x?x?x?xf32, #[[$strided4D]]>
//
// TILE-23004: linalg.conv(%[[FilterView]], %[[InputView]], %[[OutputView]]) {dilations = [10, 20], strides = [30, 40]} : memref<?x?x?x?xf32, #[[$strided4D]]>, memref<?x?x?x?xf32, #[[$strided4D]]>, memref<?x?x?x?xf32, #[[$strided4D]]>
diff --git a/mlir/test/Dialect/Linalg/tile-parallel-reduce.mlir b/mlir/test/Dialect/Linalg/tile-parallel-reduce.mlir
index 7b647426d3b7..8386bb0a3594 100644
--- a/mlir/test/Dialect/Linalg/tile-parallel-reduce.mlir
+++ b/mlir/test/Dialect/Linalg/tile-parallel-reduce.mlir
@@ -18,18 +18,18 @@ func @gemm(%arg0 : memref<?x?xf32>,
// CHECK-SAME: step (%[[C2]], %[[C4]])
// CHECK: scf.for %[[ARG5:.*]] =
// CHECK-SAME: step %[[C8]]
-// CHECK: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG5]]]
-// CHECK: %[[SV2:.*]] = subview %{{.*}}[%[[ARG5]], %[[ARG4]]]
-// CHECK: %[[SV3:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG4]]]
+// CHECK: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG5]]]
+// CHECK: %[[SV2:.*]] = memref.subview %{{.*}}[%[[ARG5]], %[[ARG4]]]
+// CHECK: %[[SV3:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG4]]]
// CHECK: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// TILE1-LABEL: func @gemm
// TILE1-DAG: %[[C2:.*]] = constant 2 : index
// TILE1: scf.parallel (%[[ARG3:.*]]) =
// TILE1-SAME: step (%[[C2]])
-// TILE1: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], 0]
-// TILE1: %[[SV3:.*]] = subview %{{.*}}[%[[ARG3]], 0]
-// TILE1-NOT: subview
+// TILE1: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0]
+// TILE1: %[[SV3:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0]
+// TILE1-NOT: memref.subview
// TILE1: linalg.matmul ins(%[[SV1]], %{{.*}} outs(%[[SV3]]
// TILE2-LABEL: func @gemm
@@ -37,9 +37,9 @@ func @gemm(%arg0 : memref<?x?xf32>,
// TILE2-DAG: %[[C4:.*]] = constant 4 : index
// TILE2: scf.parallel (%[[ARG3:.*]], %[[ARG4:.*]]) =
// TILE2-SAME: step (%[[C2]], %[[C4]])
-// TILE2: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], 0]
-// TILE2: %[[SV2:.*]] = subview %{{.*}}[0, %[[ARG4]]]
-// TILE2: %[[SV3:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG4]]]
+// TILE2: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0]
+// TILE2: %[[SV2:.*]] = memref.subview %{{.*}}[0, %[[ARG4]]]
+// TILE2: %[[SV3:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG4]]]
// TILE2: linalg.matmul ins(%[[SV1]], %[[SV2]]{{.*}} outs(%[[SV3]]
// -----
@@ -80,9 +80,9 @@ func @reduction(%arg0 : memref<?x?x?xf32>,
// CHECK-SAME: step (%[[C4]])
// CHECK: scf.for %[[ARG5:.*]] =
// CHECK-SAME: step %[[C8]]
-// CHECK: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG4]], %[[ARG5]]]
-// CHECK: %[[SV2:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG5]]]
-// CHECK: %[[SV3:.*]] = subview %{{.*}}[%[[ARG4]]]
+// CHECK: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG4]], %[[ARG5]]]
+// CHECK: %[[SV2:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG5]]]
+// CHECK: %[[SV3:.*]] = memref.subview %{{.*}}[%[[ARG4]]]
// CHECK: linalg.generic
// CHECK-SAME: ins(%[[SV1]], %[[SV2]]
// CHECK-SAME: outs(%[[SV3]]
@@ -91,9 +91,9 @@ func @reduction(%arg0 : memref<?x?x?xf32>,
// TILE1-DAG: %[[C2:.*]] = constant 2 : index
// TILE1: scf.for %[[ARG3:.*]] =
// TILE1-SAME: step %[[C2]]
-// TILE1: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], 0, 0]
-// TILE1: %[[SV2:.*]] = subview %{{.*}}[%[[ARG3]], 0]
-// TILE1-NOT: subview
+// TILE1: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0, 0]
+// TILE1: %[[SV2:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0]
+// TILE1-NOT: memref.subview
// TILE1: linalg.generic
// TILE1-SAME: ins(%[[SV1]], %[[SV2]]
// TILE1-SAME: outs(%{{.*}}
@@ -105,9 +105,9 @@ func @reduction(%arg0 : memref<?x?x?xf32>,
// TILE2-SAME: step %[[C2]]
// TILE2: scf.parallel (%[[ARG4:.*]]) =
// TILE2-SAME: step (%[[C4]])
-// TILE2: %[[SV1:.*]] = subview %{{.*}}[%[[ARG3]], %[[ARG4]], 0]
-// TILE2: %[[SV2:.*]] = subview %{{.*}}[%[[ARG3]], 0]
-// TILE2: %[[SV3:.*]] = subview %{{.*}}[%[[ARG4]]]
+// TILE2: %[[SV1:.*]] = memref.subview %{{.*}}[%[[ARG3]], %[[ARG4]], 0]
+// TILE2: %[[SV2:.*]] = memref.subview %{{.*}}[%[[ARG3]], 0]
+// TILE2: %[[SV3:.*]] = memref.subview %{{.*}}[%[[ARG4]]]
// TILE2: linalg.generic
// TILE2-SAME: ins(%[[SV1]], %[[SV2]]
// TILE2-SAME: outs(%[[SV3]]
diff --git a/mlir/test/Dialect/Linalg/tile-parallel.mlir b/mlir/test/Dialect/Linalg/tile-parallel.mlir
index 586823c8503a..55cc0c354cef 100644
--- a/mlir/test/Dialect/Linalg/tile-parallel.mlir
+++ b/mlir/test/Dialect/Linalg/tile-parallel.mlir
@@ -28,24 +28,24 @@ func @sum(%lhs: memref<?x?xf32, offset: ?, strides: [?, 1]>,
// TILE-2-SAME: [[LHS:%.*]]: {{.*}}, [[RHS:%.*]]: {{.*}}, [[SUM:%.*]]: {{.*}}) {
// TILE-2-DAG: [[C0:%.*]] = constant 0 : index
// TILE-2-DAG: [[C2:%.*]] = constant 2 : index
-// TILE-2: [[LHS_ROWS:%.*]] = dim [[LHS]], %c0
+// TILE-2: [[LHS_ROWS:%.*]] = memref.dim [[LHS]], %c0
// TILE-2: scf.parallel ([[I:%.*]]) = ([[C0]]) to ([[LHS_ROWS]]) step ([[C2]]) {
// TILE-2-NO: scf.parallel
-// TILE-2: [[LHS_SUBVIEW:%.*]] = subview [[LHS]]
-// TILE-2: [[RHS_SUBVIEW:%.*]] = subview [[RHS]]
-// TILE-2: [[SUM_SUBVIEW:%.*]] = subview [[SUM]]
+// TILE-2: [[LHS_SUBVIEW:%.*]] = memref.subview [[LHS]]
+// TILE-2: [[RHS_SUBVIEW:%.*]] = memref.subview [[RHS]]
+// TILE-2: [[SUM_SUBVIEW:%.*]] = memref.subview [[SUM]]
// TILE-2: linalg.generic {{.*}} ins([[LHS_SUBVIEW]], [[RHS_SUBVIEW]]{{.*}} outs([[SUM_SUBVIEW]]
// TILE-02-LABEL: func @sum(
// TILE-02-SAME: [[LHS:%.*]]: {{.*}}, [[RHS:%.*]]: {{.*}}, [[SUM:%.*]]: {{.*}}) {
// TILE-02-DAG: [[C0:%.*]] = constant 0 : index
// TILE-02-DAG: [[C2:%.*]] = constant 2 : index
-// TILE-02: [[LHS_COLS:%.*]] = dim [[LHS]], %c1
+// TILE-02: [[LHS_COLS:%.*]] = memref.dim [[LHS]], %c1
// TILE-02: scf.parallel ([[I:%.*]]) = ([[C0]]) to ([[LHS_COLS]]) step ([[C2]]) {
// TILE-02-NO: scf.parallel
-// TILE-02: [[LHS_SUBVIEW:%.*]] = subview [[LHS]]
-// TILE-02: [[RHS_SUBVIEW:%.*]] = subview [[RHS]]
-// TILE-02: [[SUM_SUBVIEW:%.*]] = subview [[SUM]]
+// TILE-02: [[LHS_SUBVIEW:%.*]] = memref.subview [[LHS]]
+// TILE-02: [[RHS_SUBVIEW:%.*]] = memref.subview [[RHS]]
+// TILE-02: [[SUM_SUBVIEW:%.*]] = memref.subview [[SUM]]
// TILE-02: linalg.generic {{.*}} ins([[LHS_SUBVIEW]], [[RHS_SUBVIEW]]{{.*}} outs([[SUM_SUBVIEW]]
// TILE-002-LABEL: func @sum(
@@ -58,11 +58,11 @@ func @sum(%lhs: memref<?x?xf32, offset: ?, strides: [?, 1]>,
// TILE-234-DAG: [[C0:%.*]] = constant 0 : index
// TILE-234-DAG: [[C2:%.*]] = constant 2 : index
// TILE-234-DAG: [[C3:%.*]] = constant 3 : index
-// TILE-234: [[LHS_ROWS:%.*]] = dim [[LHS]], %c0
-// TILE-234: [[LHS_COLS:%.*]] = dim [[LHS]], %c1
+// TILE-234: [[LHS_ROWS:%.*]] = memref.dim [[LHS]], %c0
+// TILE-234: [[LHS_COLS:%.*]] = memref.dim [[LHS]], %c1
// TILE-234: scf.parallel ([[I:%.*]], [[J:%.*]]) = ([[C0]], [[C0]]) to ([[LHS_ROWS]], [[LHS_COLS]]) step ([[C2]], [[C3]]) {
// TILE-234-NO: scf.parallel
-// TILE-234: [[LHS_SUBVIEW:%.*]] = subview [[LHS]]
-// TILE-234: [[RHS_SUBVIEW:%.*]] = subview [[RHS]]
-// TILE-234: [[SUM_SUBVIEW:%.*]] = subview [[SUM]]
+// TILE-234: [[LHS_SUBVIEW:%.*]] = memref.subview [[LHS]]
+// TILE-234: [[RHS_SUBVIEW:%.*]] = memref.subview [[RHS]]
+// TILE-234: [[SUM_SUBVIEW:%.*]] = memref.subview [[SUM]]
// TILE-234: linalg.generic {{.*}} ins([[LHS_SUBVIEW]], [[RHS_SUBVIEW]]{{.*}} outs([[SUM_SUBVIEW]]
diff --git a/mlir/test/Dialect/Linalg/tile-simple-conv.mlir b/mlir/test/Dialect/Linalg/tile-simple-conv.mlir
index b71f4bc0d3a8..4c49fec7536a 100644
--- a/mlir/test/Dialect/Linalg/tile-simple-conv.mlir
+++ b/mlir/test/Dialect/Linalg/tile-simple-conv.mlir
@@ -20,30 +20,30 @@ func @conv(%arg0 : memref<?x?x?x?xf32>, %arg1 : memref<?x?x?x?xf32>, %arg2 : mem
// CHECK-DAG: %[[C2:.*]] = constant 2 : index
// CHECK-DAG: %[[C3:.*]] = constant 3 : index
// CHECK-DAG: %[[C4:.*]] = constant 4 : index
-// CHECK: %[[T0:.*]] = dim %[[ARG0]], %[[C0]]
-// CHECK: %[[T1:.*]] = dim %[[ARG0]], %[[C1]]
-// CHECK: %[[T2:.*]] = dim %[[ARG1]], %[[C0]]
-// CHECK: %[[T3:.*]] = dim %[[ARG2]], %[[C1]]
-// CHECK: %[[T4:.*]] = dim %[[ARG2]], %[[C2]]
+// CHECK: %[[T0:.*]] = memref.dim %[[ARG0]], %[[C0]]
+// CHECK: %[[T1:.*]] = memref.dim %[[ARG0]], %[[C1]]
+// CHECK: %[[T2:.*]] = memref.dim %[[ARG1]], %[[C0]]
+// CHECK: %[[T3:.*]] = memref.dim %[[ARG2]], %[[C1]]
+// CHECK: %[[T4:.*]] = memref.dim %[[ARG2]], %[[C2]]
// CHECK: scf.for %[[ARG3:.*]] = %[[C0]] to %[[T2]] step %[[C2]]
// CHECK: scf.for %[[ARG4:.*]] = %[[C0]] to %[[T3]] step %[[C3]]
// CHECK: scf.for %[[ARG5:.*]] = %[[C0]] to %[[T4]] step %[[C4]]
-// CHECK: %[[T5:.*]] = dim %[[ARG1]], %[[C0]]
+// CHECK: %[[T5:.*]] = memref.dim %[[ARG1]], %[[C0]]
// CHECK: %[[T6:.*]] = affine.min #[[MAP0]](%[[ARG3]])[%[[T5]]]
-// CHECK: %[[T7:.*]] = dim %[[ARG1]], %[[C1]]
+// CHECK: %[[T7:.*]] = memref.dim %[[ARG1]], %[[C1]]
// CHECK: %[[T8:.*]] = affine.min #[[MAP1]](%[[ARG4]])[%[[T0]], %[[T7]]]
-// CHECK: %[[T9:.*]] = dim %[[ARG1]], %[[C2]]
+// CHECK: %[[T9:.*]] = memref.dim %[[ARG1]], %[[C2]]
// CHECK: %[[T10:.*]] = affine.min #[[MAP2]](%[[ARG5]])[%[[T1]], %[[T9]]]
-// CHECK: %[[T11:.*]] = dim %[[ARG1]], %[[C3]]
-// CHECK: %[[SV1:.*]] = subview %[[ARG1]][%[[ARG3]], %[[ARG4]], %[[ARG5]], 0]
+// CHECK: %[[T11:.*]] = memref.dim %[[ARG1]], %[[C3]]
+// CHECK: %[[SV1:.*]] = memref.subview %[[ARG1]][%[[ARG3]], %[[ARG4]], %[[ARG5]], 0]
// CHECK-SAME: [%[[T6]], %[[T8]], %[[T10]], %[[T11]]]
-// CHECK: %[[T13:.*]] = dim %[[ARG2]], %[[C0]]
+// CHECK: %[[T13:.*]] = memref.dim %[[ARG2]], %[[C0]]
// CHECK: %[[T14:.*]] = affine.min #[[MAP0]](%[[ARG3]])[%[[T13]]]
-// CHECK: %[[T15:.*]] = dim %[[ARG2]], %[[C1]]
+// CHECK: %[[T15:.*]] = memref.dim %[[ARG2]], %[[C1]]
// CHECK: %[[T16:.*]] = affine.min #[[MAP4]](%[[ARG4]])[%[[T15]]]
-// CHECK: %[[T17:.*]] = dim %[[ARG2]], %[[C2]]
+// CHECK: %[[T17:.*]] = memref.dim %[[ARG2]], %[[C2]]
// CHECK: %[[T18:.*]] = affine.min #[[MAP5]](%[[ARG5]])[%[[T17]]]
-// CHECK: %[[T19:.*]] = dim %[[ARG2]], %[[C3]]
-// CHECK: %[[SV2:.*]] = subview %[[ARG2]][%[[ARG3]], %[[ARG4]], %[[ARG5]], 0]
+// CHECK: %[[T19:.*]] = memref.dim %[[ARG2]], %[[C3]]
+// CHECK: %[[SV2:.*]] = memref.subview %[[ARG2]][%[[ARG3]], %[[ARG4]], %[[ARG5]], 0]
// CHECK-SAME: [%[[T14]], %[[T16]], %[[T18]], %[[T19]]]
// CHECK: linalg.conv(%[[ARG0]], %[[SV1]], %[[SV2]])
diff --git a/mlir/test/Dialect/Linalg/tile-tensors.mlir b/mlir/test/Dialect/Linalg/tile-tensors.mlir
index f8b996e1ae05..8a4478ee2812 100644
--- a/mlir/test/Dialect/Linalg/tile-tensors.mlir
+++ b/mlir/test/Dialect/Linalg/tile-tensors.mlir
@@ -34,9 +34,9 @@ func @generic_op_tensors(
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %arg0, %c0 : tensor<?x?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?x?xf32>
- %2 = dim %arg0, %c2 : tensor<?x?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?x?xf32>
+ %2 = memref.dim %arg0, %c2 : tensor<?x?x?xf32>
%3 = linalg.init_tensor [%0, %1, %2] : tensor<?x?x?xf32>
%4 = linalg.generic
{indexing_maps = [affine_map<(d0, d1, d2) -> (d0, d1, d2)>,
@@ -81,9 +81,9 @@ func @indexed_generic_op_tensors(
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = dim %arg0, %c0 : tensor<?x?x?xf32>
- %1 = dim %arg0, %c1 : tensor<?x?x?xf32>
- %2 = dim %arg0, %c2 : tensor<?x?x?xf32>
+ %0 = memref.dim %arg0, %c0 : tensor<?x?x?xf32>
+ %1 = memref.dim %arg0, %c1 : tensor<?x?x?xf32>
+ %2 = memref.dim %arg0, %c2 : tensor<?x?x?xf32>
%3 = linalg.init_tensor [%0, %1, %2] : tensor<?x?x?xf32>
%4 = linalg.indexed_generic
{indexing_maps = [affine_map<(d0, d1, d2) -> (d0, d1, d2)>,
diff --git a/mlir/test/Dialect/Linalg/tile.mlir b/mlir/test/Dialect/Linalg/tile.mlir
index 1c0a196823be..c761bd6cd57e 100644
--- a/mlir/test/Dialect/Linalg/tile.mlir
+++ b/mlir/test/Dialect/Linalg/tile.mlir
@@ -40,46 +40,46 @@ func @matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>,
// TILE-2-LABEL: func @matmul(
// TILE-2-DAG: %[[C0:.*]] = constant 0 : index
// TILE-2-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-2: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
// TILE-2: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[M]] step %{{.*}} {
-// TILE-2: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-2: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-2: %[[szM:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localM]]]
-// TILE-2: %[[K:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-2: %[[sAi:.*]] = subview %{{.*}}[%[[I]], 0] [%[[szM]], %[[K]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-2: %[[localK:.*]] = dim %{{.*}}, %c0
+// TILE-2: %[[K:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[sAi:.*]] = memref.subview %{{.*}}[%[[I]], 0] [%[[szM]], %[[K]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[localK:.*]] = memref.dim %{{.*}}, %c0
// TILE-2: %[[szK:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localK]]]
-// TILE-2: %[[N:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-2: %[[sCi:.*]] = subview %{{.*}}[%[[I]], 0] [%[[szK]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[N:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[sCi:.*]] = memref.subview %{{.*}}[%[[I]], 0] [%[[szK]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
// TILE-2: linalg.matmul ins(%[[sAi]]{{.*}} outs(%[[sCi]]
// TILE-02-LABEL: func @matmul(
// TILE-02-DAG: %[[C0:.*]] = constant 0 : index
// TILE-02-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-02: %[[N:.*]] = dim %arg1, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[N:.*]] = memref.dim %arg1, %c1 : memref<?x?xf32, #[[$strided2D]]>
// TILE-02: scf.for %[[J:.*]] = %{{.*}} to %[[N]] step %{{.*}} {
-// TILE-02: %[[K:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-02: %[[localN:.*]] = dim %{{.*}}, %c1
+// TILE-02: %[[K:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[localN:.*]] = memref.dim %{{.*}}, %c1
// TILE-02: %[[szN:.*]] = affine.min #[[$bound_map]](%[[J]])[%[[localN]]]
-// TILE-02: %[[sBj:.*]] = subview %{{.*}}[0, %[[J]]] [%[[K]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-02: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-02: %[[localK:.*]] = dim %{{.*}}, %c1
+// TILE-02: %[[sBj:.*]] = memref.subview %{{.*}}[0, %[[J]]] [%[[K]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[localK:.*]] = memref.dim %{{.*}}, %c1
// TILE-02: %[[szK:.*]] = affine.min #[[$bound_map]](%[[J]])[%[[localK]]]
-// TILE-02: %[[sCj:.*]] = subview %{{.*}}[0, %[[J]]] [%[[M]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[sCj:.*]] = memref.subview %{{.*}}[0, %[[J]]] [%[[M]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
// TILE-02: linalg.matmul ins(%{{.*}}, %[[sBj]]{{.*}} outs(%[[sCj]]
// TILE-002-LABEL: func @matmul(
// TILE-002-DAG: %[[C0:.*]] = constant 0 : index
// TILE-002-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-002: %[[ubK:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-002: %[[ubK:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
// TILE-002: scf.for %[[K:.*]] = %{{.*}}{{.*}} to %[[ubK]] step %{{.*}} {
-// TILE-002: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-002: %[[localK:.*]] = dim %{{.*}}, %c1
+// TILE-002: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-002: %[[localK:.*]] = memref.dim %{{.*}}, %c1
// TILE-002: %[[szK:.*]] = affine.min #[[$bound_map]](%[[K]])[%[[localK]]]
-// TILE-002: %[[sAj:.*]] = subview %{{.*}}[0, %[[K]]] [%[[M]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-002: %[[localK:.*]] = dim %{{.*}}, %c0
+// TILE-002: %[[sAj:.*]] = memref.subview %{{.*}}[0, %[[K]]] [%[[M]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-002: %[[localK:.*]] = memref.dim %{{.*}}, %c0
// TILE-002: %[[szK:.*]] = affine.min #[[$bound_map]](%[[K]])[%[[localK]]]
-// TILE-002: %[[N:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-002: %[[sBj:.*]] = subview %{{.*}}[%[[K]], 0] [%[[szK]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-002: %[[N:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-002: %[[sBj:.*]] = memref.subview %{{.*}}[%[[K]], 0] [%[[szK]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
// TILE-002: linalg.matmul ins(%[[sAj]], %[[sBj]]{{.*}} outs(%{{.*}}
// TILE-234-LABEL: func @matmul(
@@ -87,27 +87,27 @@ func @matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>,
// TILE-234-DAG: %[[C2:.*]] = constant 2 : index
// TILE-234-DAG: %[[C3:.*]] = constant 3 : index
// TILE-234-DAG: %[[C4:.*]] = constant 4 : index
-// TILE-234: %[[ubM:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[ubK:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[ubN:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[ubM:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[ubK:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[ubN:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
// TILE-234: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[ubM]] step %{{.*}} {
// TILE-234: scf.for %[[J:.*]] = %{{.*}}{{.*}} to %[[ubN]] step %{{.*}} {
// TILE-234: scf.for %[[K:.*]] = %{{.*}}{{.*}} to %[[ubK]] step %{{.*}} {
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[localK:.*]] = dim %{{.*}}, %c1
+// TILE-234: %[[localK:.*]] = memref.dim %{{.*}}, %c1
// TILE-234: %[[szK:.*]] = affine.min #[[$bound_map_4]](%[[K]])[%[[localK]]]
-// TILE-234: %[[sAik:.*]] = subview %{{.*}}[%[[I]], %[[K]]] [%[[szM]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[localK:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[sAik:.*]] = memref.subview %{{.*}}[%[[I]], %[[K]]] [%[[szM]], %[[szK]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[localK:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szK:.*]] = affine.min #[[$bound_map_4]](%[[K]])[%[[localK]]]
-// TILE-234: %[[localN:.*]] = dim %{{.*}}, %c1
+// TILE-234: %[[localN:.*]] = memref.dim %{{.*}}, %c1
// TILE-234: %[[szN:.*]] = affine.min #[[$bound_map_3]](%[[J]])[%[[localN]]]
-// TILE-234: %[[sBkj:.*]] = subview %{{.*}}[%[[K]], %[[J]]] [%[[szK]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[sBkj:.*]] = memref.subview %{{.*}}[%[[K]], %[[J]]] [%[[szK]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[localN:.*]] = dim %{{.*}}, %c1
+// TILE-234: %[[localN:.*]] = memref.dim %{{.*}}, %c1
// TILE-234: %[[szN:.*]] = affine.min #[[$bound_map_3]](%[[J]])[%[[localN]]]
-// TILE-234: %[[sCij:.*]] = subview %{{.*}}[%[[I]], %[[J]]] [%[[szM]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[sCij:.*]] = memref.subview %{{.*}}[%[[I]], %[[J]]] [%[[szM]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
//
// TILE-234: linalg.matmul ins(%[[sAik]], %[[sBkj]]{{.*}} outs(%[[sCij]]
@@ -133,9 +133,9 @@ func @matmul_static(%arg0: memref<10x16xf32, offset: ?, strides: [?, 1]>,
// TILE-2-DAG: %[[M:.*]] = constant 10 : index
// TILE-2: scf.for %[[I:.*]] = %{{.*}} to %[[M]] step %{{.*}} {
// TILE-2: %[[MIN2:.*]] = affine.min #[[$bound_map_static]](%[[I]])
-// TILE-2: %[[sAi:.*]] = subview %{{.*}}[%[[I]], 0] [%[[MIN2]], 16] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<?x16xf32, #[[$strided2D]]>
+// TILE-2: %[[sAi:.*]] = memref.subview %{{.*}}[%[[I]], 0] [%[[MIN2]], 16] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<?x16xf32, #[[$strided2D]]>
// TILE-2: %[[MIN22:.*]] = affine.min #[[$bound_map_static]](%[[I]])
-// TILE-2: %[[sCi:.*]] = subview %{{.*}}[%[[I]], 0] [%[[MIN22]], 12] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<?x12xf32, #[[$strided2D]]>
+// TILE-2: %[[sCi:.*]] = memref.subview %{{.*}}[%[[I]], 0] [%[[MIN22]], 12] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<?x12xf32, #[[$strided2D]]>
// TILE-2: linalg.matmul ins(%[[sAi]], %{{.*}}{{.*}} outs(%[[sCi]]
// TILE-02-LABEL: func @matmul_static(
@@ -144,9 +144,9 @@ func @matmul_static(%arg0: memref<10x16xf32, offset: ?, strides: [?, 1]>,
// TILE-02-DAG: %[[N:.*]] = constant 12 : index
// TILE-02: scf.for %[[J:.*]] = %{{.*}} to %[[N]] step %{{.*}} {
// TILE-02: %[[MIN2:.*]] = affine.min #[[$bound_map_static]](%[[J]])
-// TILE-02: %[[sBj:.*]] = subview %{{.*}}[0, %[[J]]] [16, %[[MIN2]]] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<16x?xf32, #[[$strided2D]]>
+// TILE-02: %[[sBj:.*]] = memref.subview %{{.*}}[0, %[[J]]] [16, %[[MIN2]]] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<16x?xf32, #[[$strided2D]]>
// TILE-02: %[[MIN22:.*]] = affine.min #[[$bound_map_static]](%[[J]])
-// TILE-02: %[[sCj:.*]] = subview %{{.*}}[0, %[[J]]] [10, %[[MIN22]]] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<10x?xf32, #[[$strided2D]]>
+// TILE-02: %[[sCj:.*]] = memref.subview %{{.*}}[0, %[[J]]] [10, %[[MIN22]]] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<10x?xf32, #[[$strided2D]]>
// TILE-02: linalg.matmul ins(%{{.*}}, %[[sBj]]{{.*}} outs(%[[sCj]]
// TILE-002-LABEL: func @matmul_static(
@@ -155,9 +155,9 @@ func @matmul_static(%arg0: memref<10x16xf32, offset: ?, strides: [?, 1]>,
// TILE-002-DAG: %[[C16:.*]] = constant 16 : index
// TILE-002: scf.for %[[K:.*]] = %{{.*}}{{.*}} to %[[C16]] step %{{.*}} {
// TILE-002: %[[MIN2:.*]] = affine.min #[[$bound_map_static]](%[[K]])
-// TILE-002: %[[sAj:.*]] = subview %{{.*}}[0, %[[K]]] [10, %[[MIN2]]] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<10x?xf32, #[[$strided2D]]>
+// TILE-002: %[[sAj:.*]] = memref.subview %{{.*}}[0, %[[K]]] [10, %[[MIN2]]] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<10x?xf32, #[[$strided2D]]>
// TILE-002: %[[MIN22:.*]] = affine.min #[[$bound_map_static]](%[[K]])
-// TILE-002: %[[sBj:.*]] = subview %{{.*}}[%[[K]], 0] [%[[MIN22]], 12] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<?x12xf32, #[[$strided2D]]>
+// TILE-002: %[[sBj:.*]] = memref.subview %{{.*}}[%[[K]], 0] [%[[MIN22]], 12] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<?x12xf32, #[[$strided2D]]>
// TILE-002: linalg.matmul ins(%[[sAj]], %[[sBj]]{{.*}} outs(%{{.*}}
// TILE-234-LABEL: func @matmul_static(
@@ -171,9 +171,9 @@ func @matmul_static(%arg0: memref<10x16xf32, offset: ?, strides: [?, 1]>,
// TILE-234: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[C10]] step %{{.*}} {
// TILE-234: scf.for %[[J:.*]] = %{{.*}}{{.*}} to %[[C12]] step %{{.*}} {
// TILE-234: scf.for %[[K:.*]] = %{{.*}}{{.*}} to %[[C16]] step %{{.*}} {
-// TILE-234: %[[sAik:.*]] = subview %{{.*}}[%[[I]], %[[K]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[sBkj:.*]] = subview %{{.*}}[%[[K]], %[[J]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[sCij:.*]] = subview %{{.*}}[%[[I]], %[[J]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[sAik:.*]] = memref.subview %{{.*}}[%[[I]], %[[K]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<10x16xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[sBkj:.*]] = memref.subview %{{.*}}[%[[K]], %[[J]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<16x12xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[sCij:.*]] = memref.subview %{{.*}}[%[[I]], %[[J]]] [%{{.*}}, %{{.*}}] [1, 1] : memref<10x12xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
//
// TILE-234: linalg.matmul ins(%[[sAik]], %[[sBkj]]{{.*}} outs(%[[sCij]]
@@ -190,15 +190,15 @@ func @matvec(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>, %arg1: memref<?
// TILE-2-SAME: %[[ARG2:[0-9a-zA-Z]*]]: memref
// TILE-2-DAG: %[[C0:.*]] = constant 0 : index
// TILE-2-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-2: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
// TILE-2: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[M]] step %{{.*}} {
-// TILE-2: %[[localM:.*]] = dim %[[ARG0]], %c0
+// TILE-2: %[[localM:.*]] = memref.dim %[[ARG0]], %c0
// TILE-2: %[[szM:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localM]]]
-// TILE-2: %[[N:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-2: %[[sAi:.*]] = subview %{{.*}}[%[[I]], 0] [%[[szM]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-2: %[[localN:.*]] = dim %{{.*}}, %c0
+// TILE-2: %[[N:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[sAi:.*]] = memref.subview %{{.*}}[%[[I]], 0] [%[[szM]], %[[N]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-2: %[[localN:.*]] = memref.dim %{{.*}}, %c0
// TILE-2: %[[szN:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localN]]]
-// TILE-2: %[[sCi:.*]] = subview %{{.*}}[%[[I]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-2: %[[sCi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
// TILE-2: linalg.matvec ins(%[[sAi]], %{{.*}} outs(%[[sCi]]
// TILE-02-LABEL: func @matvec(
@@ -207,15 +207,15 @@ func @matvec(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>, %arg1: memref<?
// TILE-02-SAME: %[[ARG2:[0-9a-zA-Z]*]]: memref
// TILE-02-DAG: %[[C0:.*]] = constant 0 : index
// TILE-02-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-02: %[[K:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[K:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
// TILE-02: scf.for %[[J:.*]] = %{{.*}}{{.*}} to %[[K]] step %{{.*}} {
-// TILE-02: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-02: %[[localN:.*]] = dim %{{.*}}, %c1
+// TILE-02: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[localN:.*]] = memref.dim %{{.*}}, %c1
// TILE-02: %[[szN:.*]] = affine.min #[[$bound_map]](%[[J]])[%[[localN]]]
-// TILE-02: %[[sAj:.*]] = subview %{{.*}}[0, %[[J]]] [%[[M]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-02: %[[localN:.*]] = dim %{{.*}}, %c0
+// TILE-02: %[[sAj:.*]] = memref.subview %{{.*}}[0, %[[J]]] [%[[M]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-02: %[[localN:.*]] = memref.dim %{{.*}}, %c0
// TILE-02: %[[szN:.*]] = affine.min #[[$bound_map]](%[[J]])[%[[localN]]]
-// TILE-02: %[[sBj:.*]] = subview %{{.*}}[%[[J]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-02: %[[sBj:.*]] = memref.subview %{{.*}}[%[[J]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
// TILE-02: linalg.matvec ins(%[[sAj]], %[[sBj]]{{.*}} outs(%{{.*}}
// TILE-002-LABEL: func @matvec(
@@ -231,21 +231,21 @@ func @matvec(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>, %arg1: memref<?
// TILE-234-DAG: %[[C0:.*]] = constant 0 : index
// TILE-234-DAG: %[[C2:.*]] = constant 2 : index
// TILE-234-DAG: %[[C3:.*]] = constant 3 : index
-// TILE-234: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[K:.*]] = dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[K:.*]] = memref.dim %{{.*}}, %c1 : memref<?x?xf32, #[[$strided2D]]>
// TILE-234: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[M]] step %{{.*}} {
// TILE-234: scf.for %[[J:.*]] = %{{.*}}{{.*}} to %[[K]] step %{{.*}} {
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[localN:.*]] = dim %{{.*}}, %c1
+// TILE-234: %[[localN:.*]] = memref.dim %{{.*}}, %c1
// TILE-234: %[[szN:.*]] = affine.min #[[$bound_map_3]](%[[J]])[%[[localN]]]
-// TILE-234: %[[sAij:.*]] = subview %{{.*}}[%[[I]], %[[J]]] [%[[szM]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
-// TILE-234: %[[localN:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[sAij:.*]] = memref.subview %{{.*}}[%[[I]], %[[J]]] [%[[szM]], %[[szN]]] [1, 1] : memref<?x?xf32, #[[$strided2D]]> to memref<?x?xf32, #[[$strided2D]]>
+// TILE-234: %[[localN:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szN:.*]] = affine.min #[[$bound_map_3]](%[[J]])[%[[localN]]]
-// TILE-234: %[[sBj:.*]] = subview %{{.*}}[%[[J]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[sBj:.*]] = memref.subview %{{.*}}[%[[J]]] [%[[szN]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[sCi:.*]] = subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-234: %[[sCi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
//
// TILE-234: linalg.matvec ins(%[[sAij]], %[[sBj]]{{.*}} outs(%[[sCi]]
@@ -258,14 +258,14 @@ func @dot(%arg0: memref<?xf32, offset: ?, strides: [1]>, %arg1: memref<?xf32, of
// TILE-2-LABEL: func @dot(
// TILE-2-DAG: %[[C0:.*]] = constant 0 : index
// TILE-2-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-2: %[[M:.*]] = dim %{{.*}}, %c0 : memref<?xf32, #[[$strided1D]]>
+// TILE-2: %[[M:.*]] = memref.dim %{{.*}}, %c0 : memref<?xf32, #[[$strided1D]]>
// TILE-2: scf.for %[[I:.*]] = %{{.*}}{{.*}} to %[[M]] step %{{.*}} {
-// TILE-2: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-2: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-2: %[[szM:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localM]]]
-// TILE-2: %[[sAi:.*]] = subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
-// TILE-2: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-2: %[[sAi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-2: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-2: %[[szM:.*]] = affine.min #[[$bound_map]](%[[I]])[%[[localM]]]
-// TILE-2: %[[sBi:.*]] = subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-2: %[[sBi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
// TILE-2: linalg.dot ins(%[[sAi]], %[[sBi]]{{.*}} outs(
// TILE-02-LABEL: func @dot(
@@ -277,14 +277,14 @@ func @dot(%arg0: memref<?xf32, offset: ?, strides: [1]>, %arg1: memref<?xf32, of
// TILE-234-LABEL: func @dot(
// TILE-234-DAG: %[[C0:.*]] = constant 0 : index
// TILE-234-DAG: %[[C2:.*]] = constant 2 : index
-// TILE-234: %[[ubK:.*]] = dim %{{.*}}, %c0 : memref<?xf32, #[[$strided1D]]>
+// TILE-234: %[[ubK:.*]] = memref.dim %{{.*}}, %c0 : memref<?xf32, #[[$strided1D]]>
// TILE-234: scf.for %[[I:.*]] = %{{.*}} to %[[ubK]] step %{{.*}} {
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[sAi:.*]] = subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
-// TILE-234: %[[localM:.*]] = dim %{{.*}}, %c0
+// TILE-234: %[[sAi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-234: %[[localM:.*]] = memref.dim %{{.*}}, %c0
// TILE-234: %[[szM:.*]] = affine.min #[[$bound_map_2]](%[[I]])[%[[localM]]]
-// TILE-234: %[[sBi:.*]] = subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
+// TILE-234: %[[sBi:.*]] = memref.subview %{{.*}}[%[[I]]] [%[[szM]]] [1] : memref<?xf32, #[[$strided1D]]> to memref<?xf32, #[[$strided1D]]>
// TILE-234: linalg.dot ins(%[[sAi]], %[[sBi]]{{.*}} outs(
func @fill_static(%arg0: memref<127x99xf32>, %arg1: f32) {
@@ -294,13 +294,13 @@ func @fill_static(%arg0: memref<127x99xf32>, %arg1: f32) {
// TILE-2-LABEL: func @fill_static
// TILE-2: for
// TILE-2-NOT: for
-// TILE-2: subview{{.*}} : memref<127x99xf32>
+// TILE-2: memref.subview{{.*}} : memref<127x99xf32>
// TILE-2: linalg.fill{{.*}} : memref<?x99xf32, #[[$stride_99_1_layout_map]]>, f32
// TILE-02-LABEL: func @fill_static
// TILE-02: for
// TILE-02-NOT: for
-// TILE-02: subview{{.*}} : memref<127x99xf32>
+// TILE-02: memref.subview{{.*}} : memref<127x99xf32>
// TILE-02: linalg.fill{{.*}} : memref<127x?xf32, #[[$stride_99_1_layout_map]]>, f32
// TILE-002-LABEL: func @fill_static
@@ -311,7 +311,7 @@ func @fill_static(%arg0: memref<127x99xf32>, %arg1: f32) {
// TILE-234: for
// TILE-234: for
// TILE-234-NOT: for
-// TILE-234: subview{{.*}} : memref<127x99xf32>
+// TILE-234: memref.subview{{.*}} : memref<127x99xf32>
// TILE-234: linalg.fill{{.*}} : memref<?x?xf32, #[[$stride_99_1_layout_map]]>, f32
diff --git a/mlir/test/Dialect/Linalg/transform-patterns.mlir b/mlir/test/Dialect/Linalg/transform-patterns.mlir
index 32d2e01a9731..a70816984c00 100644
--- a/mlir/test/Dialect/Linalg/transform-patterns.mlir
+++ b/mlir/test/Dialect/Linalg/transform-patterns.mlir
@@ -216,17 +216,17 @@ func @promote_subview_matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>,
%c4000 = constant 4000 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %1 = dim %arg0, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %2 = dim %arg1, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %1 = memref.dim %arg0, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %2 = memref.dim %arg1, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
scf.for %arg3 = %c0 to %0 step %c2000 {
scf.for %arg4 = %c0 to %2 step %c3000 {
scf.for %arg5 = %c0 to %1 step %c4000 {
- %3 = subview %arg0[%arg3, %arg5][%c2000, %c4000][%c1, %c1] :
+ %3 = memref.subview %arg0[%arg3, %arg5][%c2000, %c4000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %4 = subview %arg1[%arg5, %arg4][%c4000, %c3000][%c1, %c1] :
+ %4 = memref.subview %arg1[%arg5, %arg4][%c4000, %c3000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %5 = subview %arg2[%arg3, %arg4][%c2000, %c3000][%c1, %c1] :
+ %5 = memref.subview %arg2[%arg3, %arg4][%c2000, %c3000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul {__internal_linalg_transform__ = "_promote_views_"}
ins(%3, %4: memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -245,20 +245,20 @@ func @promote_subview_matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>,
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c2000]] {
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c3000]] {
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c4000]] {
-// CHECK: %[[s0:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[s1:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[s2:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[a0:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK: %[[v0:.*]] = std.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[l0:.*]] = subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
+// CHECK: %[[s0:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[s1:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[s2:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[a0:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK: %[[v0:.*]] = memref.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[l0:.*]] = memref.subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
// CHECK-SAME: memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
-// CHECK: %[[a1:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK: %[[v1:.*]] = std.view %[[a1]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[l1:.*]] = subview %[[v1]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
+// CHECK: %[[a1:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK: %[[v1:.*]] = memref.view %[[a1]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[l1:.*]] = memref.subview %[[v1]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
// CHECK-SAME: memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
-// CHECK: %[[a2:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK: %[[v2:.*]] = std.view %[[a2]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[l2:.*]] = subview %[[v2]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
+// CHECK: %[[a2:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK: %[[v2:.*]] = memref.view %[[a2]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[l2:.*]] = memref.subview %[[v2]][0, 0] [%{{.*}}, %{{.*}}] [1, 1]
// CHECK-SAME: memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
// CHECK: linalg.copy(%[[s0]], %[[l0]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>
// CHECK: linalg.copy(%[[s1]], %[[l1]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>
@@ -275,17 +275,17 @@ func @promote_first_subview_matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?
%c4000 = constant 4000 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %1 = dim %arg0, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
- %2 = dim %arg1, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %0 = memref.dim %arg0, %c0 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %1 = memref.dim %arg0, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
+ %2 = memref.dim %arg1, %c1 : memref<?x?xf32, offset: ?, strides: [?, 1]>
scf.for %arg3 = %c0 to %0 step %c2000 {
scf.for %arg4 = %c0 to %2 step %c3000 {
scf.for %arg5 = %c0 to %1 step %c4000 {
- %3 = std.subview %arg0[%arg3, %arg5][%c2000, %c4000][%c1, %c1] :
+ %3 = memref.subview %arg0[%arg3, %arg5][%c2000, %c4000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %4 = std.subview %arg1[%arg5, %arg4][%c4000, %c3000][%c1, %c1] :
+ %4 = memref.subview %arg1[%arg5, %arg4][%c4000, %c3000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %5 = std.subview %arg2[%arg3, %arg4][%c2000, %c3000][%c1, %c1] :
+ %5 = memref.subview %arg2[%arg3, %arg4][%c2000, %c3000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.matmul {__internal_linalg_transform__ = "_promote_first_view_"}
ins(%3, %4: memref<?x?xf32, offset: ?, strides: [?, ?]>,
@@ -304,18 +304,18 @@ func @promote_first_subview_matmul(%arg0: memref<?x?xf32, offset: ?, strides: [?
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c2000]] {
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c3000]] {
// CHECK: scf.for {{.*}} = %[[c0]] to {{.*}} step %[[c4000]] {
-// CHECK: %[[s0:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[s1:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[s2:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[a0:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK: %[[v0:.*]] = std.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[l0:.*]] = subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
-// CHECK-NOT: %[[a1:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK-NOT: %[[v1:.*]] = std.view %[[a1]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK-NOT: %[[l0:.*]] = subview %[[v1]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
-// CHECK-NOT: %[[a2:.*]] = alloc({{%.*}}) : memref<?xi8>
-// CHECK-NOT: %[[v2:.*]] = std.view %[[a2]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK-NOT: %[[l0:.*]] = subview %[[v2]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
+// CHECK: %[[s0:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[s1:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[s2:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[a0:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK: %[[v0:.*]] = memref.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[l0:.*]] = memref.subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
+// CHECK-NOT: %[[a1:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK-NOT: %[[v1:.*]] = memref.view %[[a1]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK-NOT: %[[l0:.*]] = memref.subview %[[v1]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
+// CHECK-NOT: %[[a2:.*]] = memref.alloc({{%.*}}) : memref<?xi8>
+// CHECK-NOT: %[[v2:.*]] = memref.view %[[a2]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK-NOT: %[[l0:.*]] = memref.subview %[[v2]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
// CHECK: linalg.copy(%[[s0]], %[[l0]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>
// CHECK-NOT: linalg.copy(%[[s1]], %[[l1]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>
// CHECK-NOT: linalg.copy(%[[s2]], %[[l2]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>^
@@ -329,7 +329,7 @@ func @aligned_promote_fill(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%cf = constant 1.0 : f32
- %3 = std.subview %arg0[%c0, %c0][%c2000, %c4000][%c1, %c1] :
+ %3 = memref.subview %arg0[%c0, %c0][%c2000, %c4000][%c1, %c1] :
memref<?x?xf32, offset: ?, strides: [?, 1]> to memref<?x?xf32, offset: ?, strides: [?, ?]>
linalg.fill(%3, %cf) { __internal_linalg_transform__ = "_promote_views_aligned_"}
: memref<?x?xf32, offset: ?, strides: [?, ?]>, f32
@@ -337,10 +337,10 @@ func @aligned_promote_fill(%arg0: memref<?x?xf32, offset: ?, strides: [?, 1]>) {
}
// CHECK-LABEL: func @aligned_promote_fill
// CHECK: %[[cf:.*]] = constant {{.*}} : f32
-// CHECK: %[[s0:.*]] = subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
-// CHECK: %[[a0:.*]] = alloc({{%.*}}) {alignment = 32 : i64} : memref<?xi8>
-// CHECK: %[[v0:.*]] = std.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
-// CHECK: %[[l0:.*]] = subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
+// CHECK: %[[s0:.*]] = memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] [{{%.*}}, {{%.*}}] : memref<?x?xf32, #map{{.*}}> to memref<?x?xf32, #map{{.*}}>
+// CHECK: %[[a0:.*]] = memref.alloc({{%.*}}) {alignment = 32 : i64} : memref<?xi8>
+// CHECK: %[[v0:.*]] = memref.view %[[a0]][{{.*}}][{{%.*}}, {{%.*}}] : memref<?xi8> to memref<?x?xf32>
+// CHECK: %[[l0:.*]] = memref.subview %[[v0]][0, 0] [%{{.*}}, %{{.*}}] [1, 1] : memref<?x?xf32> to memref<?x?xf32, #[[$STRIDED_2D_u_1]]>
// CHECK: linalg.fill(%[[v0]], {{%.*}}) : memref<?x?xf32>, f32
// CHECK: linalg.copy(%[[s0]], %[[l0]]) : memref<?x?xf32, #map{{.*}}>, memref<?x?xf32, #map{{.*}}>
// CHECK: linalg.fill(%[[v0]], %[[cf]]) : memref<?x?xf32>, f32
@@ -361,9 +361,9 @@ func @tile_permute_parallel_loop(%arg0: memref<?x?xf32>,
// CHECK-DAG: %[[C8:.*]] = constant 8 : index
// CHECK-DAG: %[[C4:.*]] = constant 4 : index
// CHECK-DAG: %[[C0:.*]] = constant 0 : index
-// CHECK-DAG: %[[D0:.*]] = dim %[[ARG0]], %c0
-// CHECK-DAG: %[[D1:.*]] = dim %[[ARG0]], %c1
-// CHECK-DAG: %[[D2:.*]] = dim %[[ARG1]], %c1
+// CHECK-DAG: %[[D0:.*]] = memref.dim %[[ARG0]], %c0
+// CHECK-DAG: %[[D1:.*]] = memref.dim %[[ARG0]], %c1
+// CHECK-DAG: %[[D2:.*]] = memref.dim %[[ARG1]], %c1
// CHECK: scf.parallel (%{{.*}}) = (%[[C0]]) to (%[[D2]]) step (%[[C8]])
// CHECK: scf.for %{{.*}} = %[[C0]] to %[[D1]] step %[[C4]]
// CHECK: scf.parallel (%{{.*}}) = (%[[C0]]) to (%[[D0]]) step (%[[C16]])
diff --git a/mlir/test/Dialect/Linalg/vectorization.mlir b/mlir/test/Dialect/Linalg/vectorization.mlir
index 436d513e637a..c43bf07d775d 100644
--- a/mlir/test/Dialect/Linalg/vectorization.mlir
+++ b/mlir/test/Dialect/Linalg/vectorization.mlir
@@ -166,7 +166,7 @@ func @test_vectorize_copy(%A : memref<8x16xf32>, %B : memref<8x16xf32>) {
// CHECK-LABEL: func @test_vectorize_copy_scalar
func @test_vectorize_copy_scalar(%A : memref<f32>, %B : memref<f32>) {
- // CHECK: %[[V:.*]] = load {{.*}} : memref<f32>
+ // CHECK: %[[V:.*]] = memref.load {{.*}} : memref<f32>
// CHECK: store %[[V]], {{.*}} : memref<f32>
linalg.copy(%A, %B) : memref<f32>, memref<f32>
return
diff --git a/mlir/test/Dialect/OpenACC/invalid.mlir b/mlir/test/Dialect/OpenACC/invalid.mlir
index df52b0cd9a60..9d2afb7d3a08 100644
--- a/mlir/test/Dialect/OpenACC/invalid.mlir
+++ b/mlir/test/Dialect/OpenACC/invalid.mlir
@@ -82,21 +82,21 @@ acc.update
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{wait_devnum cannot appear without waitOperands}}
acc.update wait_devnum(%cst: index) host(%value: memref<10xf32>)
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{async attribute cannot appear with asyncOperand}}
acc.update async(%cst: index) host(%value: memref<10xf32>) attributes {async}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{wait attribute cannot appear with waitOperands}}
acc.update wait(%cst: index) host(%value: memref<10xf32>) attributes {wait}
@@ -162,14 +162,14 @@ acc.exit_data attributes {async}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{async attribute cannot appear with asyncOperand}}
acc.exit_data async(%cst: index) delete(%value : memref<10xf32>) attributes {async}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{wait_devnum cannot appear without waitOperands}}
acc.exit_data wait_devnum(%cst: index) delete(%value : memref<10xf32>)
@@ -181,20 +181,20 @@ acc.enter_data attributes {async}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{async attribute cannot appear with asyncOperand}}
acc.enter_data async(%cst: index) create(%value : memref<10xf32>) attributes {async}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{wait attribute cannot appear with waitOperands}}
acc.enter_data wait(%cst: index) create(%value : memref<10xf32>) attributes {wait}
// -----
%cst = constant 1 : index
-%value = alloc() : memref<10xf32>
+%value = memref.alloc() : memref<10xf32>
// expected-error at +1 {{wait_devnum cannot appear without waitOperands}}
acc.enter_data wait_devnum(%cst: index) create(%value : memref<10xf32>)
diff --git a/mlir/test/Dialect/OpenACC/ops.mlir b/mlir/test/Dialect/OpenACC/ops.mlir
index 6e5336727a65..6dddf4085354 100644
--- a/mlir/test/Dialect/OpenACC/ops.mlir
+++ b/mlir/test/Dialect/OpenACC/ops.mlir
@@ -15,12 +15,12 @@ func @compute1(%A: memref<10x10xf32>, %B: memref<10x10xf32>, %C: memref<10x10xf3
scf.for %arg3 = %c0 to %c10 step %c1 {
scf.for %arg4 = %c0 to %c10 step %c1 {
scf.for %arg5 = %c0 to %c10 step %c1 {
- %a = load %A[%arg3, %arg5] : memref<10x10xf32>
- %b = load %B[%arg5, %arg4] : memref<10x10xf32>
- %cij = load %C[%arg3, %arg4] : memref<10x10xf32>
+ %a = memref.load %A[%arg3, %arg5] : memref<10x10xf32>
+ %b = memref.load %B[%arg5, %arg4] : memref<10x10xf32>
+ %cij = memref.load %C[%arg3, %arg4] : memref<10x10xf32>
%p = mulf %a, %b : f32
%co = addf %cij, %p : f32
- store %co, %C[%arg3, %arg4] : memref<10x10xf32>
+ memref.store %co, %C[%arg3, %arg4] : memref<10x10xf32>
}
}
}
@@ -42,12 +42,12 @@ func @compute1(%A: memref<10x10xf32>, %B: memref<10x10xf32>, %C: memref<10x10xf3
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
// CHECK-NEXT: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
// CHECK-NEXT: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
-// CHECK-NEXT: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: memref.store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
@@ -70,12 +70,12 @@ func @compute2(%A: memref<10x10xf32>, %B: memref<10x10xf32>, %C: memref<10x10xf3
scf.for %arg3 = %c0 to %c10 step %c1 {
scf.for %arg4 = %c0 to %c10 step %c1 {
scf.for %arg5 = %c0 to %c10 step %c1 {
- %a = load %A[%arg3, %arg5] : memref<10x10xf32>
- %b = load %B[%arg5, %arg4] : memref<10x10xf32>
- %cij = load %C[%arg3, %arg4] : memref<10x10xf32>
+ %a = memref.load %A[%arg3, %arg5] : memref<10x10xf32>
+ %b = memref.load %B[%arg5, %arg4] : memref<10x10xf32>
+ %cij = memref.load %C[%arg3, %arg4] : memref<10x10xf32>
%p = mulf %a, %b : f32
%co = addf %cij, %p : f32
- store %co, %C[%arg3, %arg4] : memref<10x10xf32>
+ memref.store %co, %C[%arg3, %arg4] : memref<10x10xf32>
}
}
}
@@ -96,12 +96,12 @@ func @compute2(%A: memref<10x10xf32>, %B: memref<10x10xf32>, %C: memref<10x10xf3
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
// CHECK-NEXT: scf.for %{{.*}} = %{{.*}} to %{{.*}} step %{{.*}} {
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
// CHECK-NEXT: %{{.*}} = mulf %{{.*}}, %{{.*}} : f32
// CHECK-NEXT: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
-// CHECK-NEXT: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: memref.store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
@@ -127,10 +127,10 @@ func @compute3(%a: memref<10x10xf32>, %b: memref<10x10xf32>, %c: memref<10xf32>,
scf.for %x = %lb to %c10 step %st {
acc.loop worker {
scf.for %y = %lb to %c10 step %st {
- %axy = load %a[%x, %y] : memref<10x10xf32>
- %bxy = load %b[%x, %y] : memref<10x10xf32>
+ %axy = memref.load %a[%x, %y] : memref<10x10xf32>
+ %bxy = memref.load %b[%x, %y] : memref<10x10xf32>
%tmp = addf %axy, %bxy : f32
- store %tmp, %c[%y] : memref<10xf32>
+ memref.store %tmp, %c[%y] : memref<10xf32>
}
acc.yield
}
@@ -139,10 +139,10 @@ func @compute3(%a: memref<10x10xf32>, %b: memref<10x10xf32>, %c: memref<10xf32>,
// for i = 0 to 10 step 1
// d[x] += c[i]
scf.for %i = %lb to %c10 step %st {
- %ci = load %c[%i] : memref<10xf32>
- %dx = load %d[%x] : memref<10xf32>
+ %ci = memref.load %c[%i] : memref<10xf32>
+ %dx = memref.load %d[%x] : memref<10xf32>
%z = addf %ci, %dx : f32
- store %z, %d[%x] : memref<10xf32>
+ memref.store %z, %d[%x] : memref<10xf32>
}
acc.yield
} attributes {seq}
@@ -169,19 +169,19 @@ func @compute3(%a: memref<10x10xf32>, %b: memref<10x10xf32>, %c: memref<10xf32>,
// CHECK-NEXT: scf.for %{{.*}} = [[C0]] to [[C10]] step [[C1]] {
// CHECK-NEXT: acc.loop worker {
// CHECK-NEXT: scf.for %{{.*}} = [[C0]] to [[C10]] step [[C1]] {
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<10x10xf32>
// CHECK-NEXT: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
-// CHECK-NEXT: store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
+// CHECK-NEXT: memref.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: acc.yield
// CHECK-NEXT: }
// CHECK-NEXT: acc.loop {
// CHECK-NEXT: scf.for %{{.*}} = [[C0]] to [[C10]] step [[C1]] {
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}] : memref<10xf32>
-// CHECK-NEXT: %{{.*}} = load %{{.*}}[%{{.*}}] : memref<10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}] : memref<10xf32>
+// CHECK-NEXT: %{{.*}} = memref.load %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: %{{.*}} = addf %{{.*}}, %{{.*}} : f32
-// CHECK-NEXT: store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
+// CHECK-NEXT: memref.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: acc.yield
// CHECK-NEXT: } attributes {seq}
diff --git a/mlir/test/Dialect/SCF/bufferize.mlir b/mlir/test/Dialect/SCF/bufferize.mlir
index 7f42e15107d6..39a0fefea1a5 100644
--- a/mlir/test/Dialect/SCF/bufferize.mlir
+++ b/mlir/test/Dialect/SCF/bufferize.mlir
@@ -5,13 +5,13 @@
// CHECK-SAME: %[[TRUE_TENSOR:.*]]: tensor<?xf32>,
// CHECK-SAME: %[[FALSE_TENSOR:.*]]: tensor<?xf32>) -> tensor<?xf32> {
// CHECK: %[[RESULT_MEMREF:.*]] = scf.if %[[PRED]] -> (memref<?xf32>) {
-// CHECK: %[[TRUE_MEMREF:.*]] = tensor_to_memref %[[TRUE_TENSOR]] : memref<?xf32>
+// CHECK: %[[TRUE_MEMREF:.*]] = memref.buffer_cast %[[TRUE_TENSOR]] : memref<?xf32>
// CHECK: scf.yield %[[TRUE_MEMREF]] : memref<?xf32>
// CHECK: } else {
-// CHECK: %[[FALSE_MEMREF:.*]] = tensor_to_memref %[[FALSE_TENSOR]] : memref<?xf32>
+// CHECK: %[[FALSE_MEMREF:.*]] = memref.buffer_cast %[[FALSE_TENSOR]] : memref<?xf32>
// CHECK: scf.yield %[[FALSE_MEMREF]] : memref<?xf32>
// CHECK: }
-// CHECK: %[[RESULT_TENSOR:.*]] = tensor_load %[[RESULT_MEMREF:.*]] : memref<?xf32>
+// CHECK: %[[RESULT_TENSOR:.*]] = memref.tensor_load %[[RESULT_MEMREF:.*]] : memref<?xf32>
// CHECK: return %[[RESULT_TENSOR]] : tensor<?xf32>
// CHECK: }
func @if(%pred: i1, %true_val: tensor<?xf32>, %false_val: tensor<?xf32>) -> tensor<?xf32> {
@@ -27,13 +27,13 @@ func @if(%pred: i1, %true_val: tensor<?xf32>, %false_val: tensor<?xf32>) -> tens
// CHECK-SAME: %[[TENSOR:.*]]: tensor<f32>,
// CHECK-SAME: %[[LB:.*]]: index, %[[UB:.*]]: index,
// CHECK-SAME: %[[STEP:.*]]: index) -> tensor<f32> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
// CHECK: %[[RESULT_MEMREF:.*]] = scf.for %[[VAL_6:.*]] = %[[LB]] to %[[UB]] step %[[STEP]] iter_args(%[[ITER:.*]] = %[[MEMREF]]) -> (memref<f32>) {
-// CHECK: %[[TENSOR_ITER:.*]] = tensor_load %[[ITER]] : memref<f32>
-// CHECK: %[[MEMREF_YIELDED:.*]] = tensor_to_memref %[[TENSOR_ITER]] : memref<f32>
+// CHECK: %[[TENSOR_ITER:.*]] = memref.tensor_load %[[ITER]] : memref<f32>
+// CHECK: %[[MEMREF_YIELDED:.*]] = memref.buffer_cast %[[TENSOR_ITER]] : memref<f32>
// CHECK: scf.yield %[[MEMREF_YIELDED]] : memref<f32>
// CHECK: }
-// CHECK: %[[VAL_8:.*]] = tensor_load %[[VAL_9:.*]] : memref<f32>
+// CHECK: %[[VAL_8:.*]] = memref.tensor_load %[[VAL_9:.*]] : memref<f32>
// CHECK: return %[[VAL_8]] : tensor<f32>
// CHECK: }
func @for(%arg0: tensor<f32>, %lb: index, %ub: index, %step: index) -> tensor<f32> {
@@ -62,14 +62,14 @@ func @if_correct_recursive_legalization_behavior(%pred: i1, %tensor: tensor<f32>
// CHECK-LABEL: func @for_correct_recursive_legalization_behavior(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<f32>,
// CHECK-SAME: %[[INDEX:.*]]: index) -> tensor<f32> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
// CHECK: %[[RESULT:.*]] = scf.for %[[IV:.*]] = %[[INDEX]] to %[[INDEX]] step %[[INDEX]] iter_args(%[[MEMREF_ITER:.*]] = %[[MEMREF]]) -> (memref<f32>) {
-// CHECK: %[[TENSOR_ITER:.*]] = tensor_load %[[MEMREF_ITER]] : memref<f32>
+// CHECK: %[[TENSOR_ITER:.*]] = memref.tensor_load %[[MEMREF_ITER]] : memref<f32>
// CHECK: %[[TENSOR_MUNGED:.*]] = "test.munge_tensor"(%[[TENSOR_ITER]]) : (tensor<f32>) -> tensor<f32>
-// CHECK: %[[MEMREF_MUNGED:.*]] = tensor_to_memref %[[TENSOR_MUNGED]] : memref<f32>
+// CHECK: %[[MEMREF_MUNGED:.*]] = memref.buffer_cast %[[TENSOR_MUNGED]] : memref<f32>
// CHECK: scf.yield %[[MEMREF_MUNGED]] : memref<f32>
// CHECK: }
-// CHECK: %[[TENSOR:.*]] = tensor_load %[[RESULT:.*]] : memref<f32>
+// CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[RESULT:.*]] : memref<f32>
// CHECK: return %[[TENSOR]] : tensor<f32>
// CHECK: }
func @for_correct_recursive_legalization_behavior(%arg0: tensor<f32>, %index: index) -> tensor<f32> {
diff --git a/mlir/test/Dialect/SCF/canonicalize.mlir b/mlir/test/Dialect/SCF/canonicalize.mlir
index 0d7c4eefae25..8f76926bdff0 100644
--- a/mlir/test/Dialect/SCF/canonicalize.mlir
+++ b/mlir/test/Dialect/SCF/canonicalize.mlir
@@ -13,7 +13,7 @@ func @single_iteration(%A: memref<?x?x?xi32>) {
%c10 = constant 10 : index
scf.parallel (%i0, %i1, %i2) = (%c0, %c3, %c7) to (%c1, %c6, %c10) step (%c1, %c2, %c3) {
%c42 = constant 42 : i32
- store %c42, %A[%i0, %i1, %i2] : memref<?x?x?xi32>
+ memref.store %c42, %A[%i0, %i1, %i2] : memref<?x?x?xi32>
scf.yield
}
return
@@ -28,7 +28,7 @@ func @single_iteration(%A: memref<?x?x?xi32>) {
// CHECK: [[C7:%.*]] = constant 7 : index
// CHECK: [[C42:%.*]] = constant 42 : i32
// CHECK: scf.parallel ([[V0:%.*]]) = ([[C3]]) to ([[C6]]) step ([[C2]]) {
-// CHECK: store [[C42]], [[ARG0]]{{\[}}[[C0]], [[V0]], [[C7]]] : memref<?x?x?xi32>
+// CHECK: memref.store [[C42]], [[ARG0]]{{\[}}[[C0]], [[V0]], [[C7]]] : memref<?x?x?xi32>
// CHECK: scf.yield
// CHECK: }
// CHECK: return
@@ -348,12 +348,12 @@ func @last_value(%t0: tensor<128x128xf32>, %t1: tensor<128x128xf32>,
%lb : index, %ub : index, %step : index)
-> (tensor<128x128xf32>, tensor<128x128xf32>, tensor<128x128xf32>)
{
- // CHECK-NEXT: %[[M1:.*]] = tensor_to_memref %[[T1]] : memref<128x128xf32>
+ // CHECK-NEXT: %[[M1:.*]] = memref.buffer_cast %[[T1]] : memref<128x128xf32>
// CHECK-NEXT: %[[FOR_RES:.*]] = scf.for {{.*}} iter_args(%[[BBARG_T2:.*]] = %[[T2]]) -> (tensor<128x128xf32>) {
%0:3 = scf.for %arg0 = %lb to %ub step %step iter_args(%arg1 = %t0, %arg2 = %t1, %arg3 = %t2)
-> (tensor<128x128xf32>, tensor<128x128xf32>, tensor<128x128xf32>)
{
- %m1 = tensor_to_memref %arg2 : memref<128x128xf32>
+ %m1 = memref.buffer_cast %arg2 : memref<128x128xf32>
// CHECK-NEXT: call @process(%[[M0]]) : (memref<128x128xf32>) -> ()
call @process(%m0) : (memref<128x128xf32>) -> ()
@@ -363,13 +363,13 @@ func @last_value(%t0: tensor<128x128xf32>, %t1: tensor<128x128xf32>,
// This does not hoist (fails the bbArg has at most a single check).
// CHECK-NEXT: %[[T:.*]] = call @process_tensor(%[[BBARG_T2]]) : (tensor<128x128xf32>) -> memref<128x128xf32>
- // CHECK-NEXT: %[[YIELD_T:.*]] = tensor_load %[[T:.*]]
+ // CHECK-NEXT: %[[YIELD_T:.*]] = memref.tensor_load %[[T:.*]]
%m2 = call @process_tensor(%arg3): (tensor<128x128xf32>) -> memref<128x128xf32>
- %3 = tensor_load %m2 : memref<128x128xf32>
+ %3 = memref.tensor_load %m2 : memref<128x128xf32>
// All this stuff goes away, incrementally
- %1 = tensor_load %m0 : memref<128x128xf32>
- %2 = tensor_load %m1 : memref<128x128xf32>
+ %1 = memref.tensor_load %m0 : memref<128x128xf32>
+ %2 = memref.tensor_load %m1 : memref<128x128xf32>
// CHECK-NEXT: scf.yield %[[YIELD_T]] : tensor<128x128xf32>
scf.yield %1, %2, %3 : tensor<128x128xf32>, tensor<128x128xf32>, tensor<128x128xf32>
@@ -377,8 +377,8 @@ func @last_value(%t0: tensor<128x128xf32>, %t1: tensor<128x128xf32>,
// CHECK-NEXT: }
}
- // CHECK-NEXT: %[[R0:.*]] = tensor_load %[[M0]] : memref<128x128xf32>
- // CHECK-NEXT: %[[R1:.*]] = tensor_load %[[M1]] : memref<128x128xf32>
+ // CHECK-NEXT: %[[R0:.*]] = memref.tensor_load %[[M0]] : memref<128x128xf32>
+ // CHECK-NEXT: %[[R1:.*]] = memref.tensor_load %[[M1]] : memref<128x128xf32>
// CHECK-NEXT: return %[[R0]], %[[R1]], %[[FOR_RES]] : tensor<128x128xf32>, tensor<128x128xf32>, tensor<128x128xf32>
return %0#0, %0#1, %0#2 : tensor<128x128xf32>, tensor<128x128xf32>, tensor<128x128xf32>
}
diff --git a/mlir/test/Dialect/SCF/for-loop-specialization.mlir b/mlir/test/Dialect/SCF/for-loop-specialization.mlir
index 473b8232e351..e5bd752ded6c 100644
--- a/mlir/test/Dialect/SCF/for-loop-specialization.mlir
+++ b/mlir/test/Dialect/SCF/for-loop-specialization.mlir
@@ -7,13 +7,13 @@ func @for(%outer: index, %A: memref<?xf32>, %B: memref<?xf32>,
%C: memref<?xf32>, %result: memref<?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %d0 = dim %A, %c0 : memref<?xf32>
+ %d0 = memref.dim %A, %c0 : memref<?xf32>
%b0 = affine.min #map0()[%d0, %outer]
scf.for %i0 = %c0 to %b0 step %c1 {
- %B_elem = load %B[%i0] : memref<?xf32>
- %C_elem = load %C[%i0] : memref<?xf32>
+ %B_elem = memref.load %B[%i0] : memref<?xf32>
+ %C_elem = memref.load %C[%i0] : memref<?xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %result[%i0] : memref<?xf32>
+ memref.store %sum_elem, %result[%i0] : memref<?xf32>
}
return
}
@@ -22,17 +22,17 @@ func @for(%outer: index, %A: memref<?xf32>, %B: memref<?xf32>,
// CHECK-SAME: [[ARG0:%.*]]: index, [[ARG1:%.*]]: memref<?xf32>, [[ARG2:%.*]]: memref<?xf32>, [[ARG3:%.*]]: memref<?xf32>, [[ARG4:%.*]]: memref<?xf32>) {
// CHECK: [[CST_0:%.*]] = constant 0 : index
// CHECK: [[CST_1:%.*]] = constant 1 : index
-// CHECK: [[DIM_0:%.*]] = dim [[ARG1]], [[CST_0]] : memref<?xf32>
+// CHECK: [[DIM_0:%.*]] = memref.dim [[ARG1]], [[CST_0]] : memref<?xf32>
// CHECK: [[MIN:%.*]] = affine.min #map(){{\[}}[[DIM_0]], [[ARG0]]]
// CHECK: [[CST_1024:%.*]] = constant 1024 : index
// CHECK: [[PRED:%.*]] = cmpi eq, [[MIN]], [[CST_1024]] : index
// CHECK: scf.if [[PRED]] {
// CHECK: scf.for [[IDX0:%.*]] = [[CST_0]] to [[CST_1024]] step [[CST_1]] {
-// CHECK: store
+// CHECK: memref.store
// CHECK: }
// CHECK: } else {
// CHECK: scf.for [[IDX0:%.*]] = [[CST_0]] to [[MIN]] step [[CST_1]] {
-// CHECK: store
+// CHECK: memref.store
// CHECK: }
// CHECK: }
// CHECK: return
diff --git a/mlir/test/Dialect/SCF/loop-unroll.mlir b/mlir/test/Dialect/SCF/loop-unroll.mlir
index 0b6e178ed0aa..a824a7df51ce 100644
--- a/mlir/test/Dialect/SCF/loop-unroll.mlir
+++ b/mlir/test/Dialect/SCF/loop-unroll.mlir
@@ -8,7 +8,7 @@ func @dynamic_loop_unroll(%arg0 : index, %arg1 : index, %arg2 : index,
%arg3: memref<?xf32>) {
%0 = constant 7.0 : f32
scf.for %i0 = %arg0 to %arg1 step %arg2 {
- store %0, %arg3[%i0] : memref<?xf32>
+ memref.store %0, %arg3[%i0] : memref<?xf32>
}
return
}
@@ -34,14 +34,14 @@ func @dynamic_loop_unroll(%arg0 : index, %arg1 : index, %arg2 : index,
// Compute step of unrolled loop in V8.
// UNROLL-BY-2-DAG: %[[V8:.*]] = muli %[[STEP]], %[[C2]] : index
// UNROLL-BY-2: scf.for %[[IV:.*]] = %[[LB]] to %[[V7]] step %[[V8]] {
-// UNROLL-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-2-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-BY-2-NEXT: %[[V9:.*]] = muli %[[STEP]], %[[C1_IV]] : index
// UNROLL-BY-2-NEXT: %[[V10:.*]] = addi %[[IV]], %[[V9]] : index
-// UNROLL-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[V10]]] : memref<?xf32>
+// UNROLL-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V10]]] : memref<?xf32>
// UNROLL-BY-2-NEXT: }
// UNROLL-BY-2-NEXT: scf.for %[[IV:.*]] = %[[V7]] to %[[UB]] step %[[STEP]] {
-// UNROLL-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-2-NEXT: }
// UNROLL-BY-2-NEXT: return
@@ -67,18 +67,18 @@ func @dynamic_loop_unroll(%arg0 : index, %arg1 : index, %arg2 : index,
// Compute step of unrolled loop in V8.
// UNROLL-BY-3-DAG: %[[V8:.*]] = muli %[[STEP]], %[[C3]] : index
// UNROLL-BY-3: scf.for %[[IV:.*]] = %[[LB]] to %[[V7]] step %[[V8]] {
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-BY-3-NEXT: %[[V9:.*]] = muli %[[STEP]], %[[C1_IV]] : index
// UNROLL-BY-3-NEXT: %[[V10:.*]] = addi %[[IV]], %[[V9]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V10]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V10]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C2_IV:.*]] = constant 2 : index
// UNROLL-BY-3-NEXT: %[[V11:.*]] = muli %[[STEP]], %[[C2_IV]] : index
// UNROLL-BY-3-NEXT: %[[V12:.*]] = addi %[[IV]], %[[V11]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V12]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V12]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: }
// UNROLL-BY-3-NEXT: scf.for %[[IV:.*]] = %[[V7]] to %[[UB]] step %[[STEP]] {
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: }
// UNROLL-BY-3-NEXT: return
@@ -88,7 +88,7 @@ func @dynamic_loop_unroll_outer_by_2(
%0 = constant 7.0 : f32
scf.for %i0 = %arg0 to %arg1 step %arg2 {
scf.for %i1 = %arg3 to %arg4 step %arg5 {
- store %0, %arg6[%i1] : memref<?xf32>
+ memref.store %0, %arg6[%i1] : memref<?xf32>
}
}
return
@@ -104,15 +104,15 @@ func @dynamic_loop_unroll_outer_by_2(
//
// UNROLL-OUTER-BY-2: scf.for %[[IV0:.*]] = %[[LB0]] to %{{.*}} step %{{.*}} {
// UNROLL-OUTER-BY-2-NEXT: scf.for %[[IV1:.*]] = %[[LB1]] to %[[UB1]] step %[[STEP1]] {
-// UNROLL-OUTER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
+// UNROLL-OUTER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
// UNROLL-OUTER-BY-2-NEXT: }
// UNROLL-OUTER-BY-2-NEXT: scf.for %[[IV1:.*]] = %[[LB1]] to %[[UB1]] step %[[STEP1]] {
-// UNROLL-OUTER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
+// UNROLL-OUTER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
// UNROLL-OUTER-BY-2-NEXT: }
// UNROLL-OUTER-BY-2-NEXT: }
// UNROLL-OUTER-BY-2-NEXT: scf.for %[[IV0:.*]] = %{{.*}} to %[[UB0]] step %[[STEP0]] {
// UNROLL-OUTER-BY-2-NEXT: scf.for %[[IV1:.*]] = %[[LB1]] to %[[UB1]] step %[[STEP1]] {
-// UNROLL-OUTER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
+// UNROLL-OUTER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
// UNROLL-OUTER-BY-2-NEXT: }
// UNROLL-OUTER-BY-2-NEXT: }
// UNROLL-OUTER-BY-2-NEXT: return
@@ -123,7 +123,7 @@ func @dynamic_loop_unroll_inner_by_2(
%0 = constant 7.0 : f32
scf.for %i0 = %arg0 to %arg1 step %arg2 {
scf.for %i1 = %arg3 to %arg4 step %arg5 {
- store %0, %arg6[%i1] : memref<?xf32>
+ memref.store %0, %arg6[%i1] : memref<?xf32>
}
}
return
@@ -139,14 +139,14 @@ func @dynamic_loop_unroll_inner_by_2(
//
// UNROLL-INNER-BY-2: scf.for %[[IV0:.*]] = %[[LB0]] to %[[UB0]] step %[[STEP0]] {
// UNROLL-INNER-BY-2: scf.for %[[IV1:.*]] = %[[LB1]] to %{{.*}} step %{{.*}} {
-// UNROLL-INNER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
+// UNROLL-INNER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
// UNROLL-INNER-BY-2-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-INNER-BY-2-NEXT: %[[V0:.*]] = muli %[[STEP1]], %[[C1_IV]] : index
// UNROLL-INNER-BY-2-NEXT: %[[V1:.*]] = addi %[[IV1]], %[[V0]] : index
-// UNROLL-INNER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
+// UNROLL-INNER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
// UNROLL-INNER-BY-2-NEXT: }
// UNROLL-INNER-BY-2-NEXT: scf.for %[[IV1:.*]] = %{{.*}} to %[[UB1]] step %[[STEP1]] {
-// UNROLL-INNER-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
+// UNROLL-INNER-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV1]]] : memref<?xf32>
// UNROLL-INNER-BY-2-NEXT: }
// UNROLL-INNER-BY-2-NEXT: }
// UNROLL-INNER-BY-2-NEXT: return
@@ -159,7 +159,7 @@ func @static_loop_unroll_by_2(%arg0 : memref<?xf32>) {
%ub = constant 20 : index
%step = constant 1 : index
scf.for %i0 = %lb to %ub step %step {
- store %0, %arg0[%i0] : memref<?xf32>
+ memref.store %0, %arg0[%i0] : memref<?xf32>
}
return
}
@@ -171,11 +171,11 @@ func @static_loop_unroll_by_2(%arg0 : memref<?xf32>) {
// UNROLL-BY-2-DAG: %[[C20:.*]] = constant 20 : index
// UNROLL-BY-2-DAG: %[[C2:.*]] = constant 2 : index
// UNROLL-BY-2: scf.for %[[IV:.*]] = %[[C0]] to %[[C20]] step %[[C2]] {
-// UNROLL-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-2-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-BY-2-NEXT: %[[V0:.*]] = muli %[[C1]], %[[C1_IV]] : index
// UNROLL-BY-2-NEXT: %[[V1:.*]] = addi %[[IV]], %[[V0]] : index
-// UNROLL-BY-2-NEXT: store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
+// UNROLL-BY-2-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
// UNROLL-BY-2-NEXT: }
// UNROLL-BY-2-NEXT: return
@@ -187,7 +187,7 @@ func @static_loop_unroll_by_3(%arg0 : memref<?xf32>) {
%ub = constant 20 : index
%step = constant 1 : index
scf.for %i0 = %lb to %ub step %step {
- store %0, %arg0[%i0] : memref<?xf32>
+ memref.store %0, %arg0[%i0] : memref<?xf32>
}
return
}
@@ -201,18 +201,18 @@ func @static_loop_unroll_by_3(%arg0 : memref<?xf32>) {
// UNROLL-BY-3-DAG: %[[C18:.*]] = constant 18 : index
// UNROLL-BY-3-DAG: %[[C3:.*]] = constant 3 : index
// UNROLL-BY-3: scf.for %[[IV:.*]] = %[[C0]] to %[[C18]] step %[[C3]] {
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-BY-3-NEXT: %[[V0:.*]] = muli %[[C1]], %[[C1_IV]] : index
// UNROLL-BY-3-NEXT: %[[V1:.*]] = addi %[[IV]], %[[V0]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C2_IV:.*]] = constant 2 : index
// UNROLL-BY-3-NEXT: %[[V2:.*]] = muli %[[C1]], %[[C2_IV]] : index
// UNROLL-BY-3-NEXT: %[[V3:.*]] = addi %[[IV]], %[[V2]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V3]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V3]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: }
// UNROLL-BY-3-NEXT: scf.for %[[IV:.*]] = %[[C18]] to %[[C20]] step %[[C1]] {
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: }
// UNROLL-BY-3-NEXT: return
@@ -224,7 +224,7 @@ func @static_loop_unroll_by_3_promote_epilogue(%arg0 : memref<?xf32>) {
%ub = constant 10 : index
%step = constant 1 : index
scf.for %i0 = %lb to %ub step %step {
- store %0, %arg0[%i0] : memref<?xf32>
+ memref.store %0, %arg0[%i0] : memref<?xf32>
}
return
}
@@ -237,17 +237,17 @@ func @static_loop_unroll_by_3_promote_epilogue(%arg0 : memref<?xf32>) {
// UNROLL-BY-3-DAG: %[[C9:.*]] = constant 9 : index
// UNROLL-BY-3-DAG: %[[C3:.*]] = constant 3 : index
// UNROLL-BY-3: scf.for %[[IV:.*]] = %[[C0]] to %[[C9]] step %[[C3]] {
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[IV]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C1_IV:.*]] = constant 1 : index
// UNROLL-BY-3-NEXT: %[[V0:.*]] = muli %[[C1]], %[[C1_IV]] : index
// UNROLL-BY-3-NEXT: %[[V1:.*]] = addi %[[IV]], %[[V0]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V1]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: %[[C2_IV:.*]] = constant 2 : index
// UNROLL-BY-3-NEXT: %[[V2:.*]] = muli %[[C1]], %[[C2_IV]] : index
// UNROLL-BY-3-NEXT: %[[V3:.*]] = addi %[[IV]], %[[V2]] : index
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[V3]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[V3]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: }
-// UNROLL-BY-3-NEXT: store %{{.*}}, %[[MEM]][%[[C9]]] : memref<?xf32>
+// UNROLL-BY-3-NEXT: memref.store %{{.*}}, %[[MEM]][%[[C9]]] : memref<?xf32>
// UNROLL-BY-3-NEXT: return
// Test unroll-up-to functionality.
diff --git a/mlir/test/Dialect/SCF/ops.mlir b/mlir/test/Dialect/SCF/ops.mlir
index 8192653787c2..35cbc88c2950 100644
--- a/mlir/test/Dialect/SCF/ops.mlir
+++ b/mlir/test/Dialect/SCF/ops.mlir
@@ -208,7 +208,7 @@ func @conditional_reduce(%buffer: memref<1024xf32>, %lb: index, %ub: index, %ste
%sum_0 = constant 0.0 : f32
%c0 = constant 0.0 : f32
%sum = scf.for %iv = %lb to %ub step %step iter_args(%sum_iter = %sum_0) -> (f32) {
- %t = load %buffer[%iv] : memref<1024xf32>
+ %t = memref.load %buffer[%iv] : memref<1024xf32>
%cond = cmpf ugt, %t, %c0 : f32
%sum_next = scf.if %cond -> (f32) {
%new_sum = addf %sum_iter, %t : f32
@@ -229,7 +229,7 @@ func @conditional_reduce(%buffer: memref<1024xf32>, %lb: index, %ub: index, %ste
// CHECK-NEXT: %[[ZERO:.*]] = constant
// CHECK-NEXT: %[[RESULT:.*]] = scf.for %[[IV:.*]] = %[[ARG1]] to %[[ARG2]] step %[[ARG3]]
// CHECK-SAME: iter_args(%[[ITER:.*]] = %[[INIT]]) -> (f32) {
-// CHECK-NEXT: %[[T:.*]] = load %[[ARG0]][%[[IV]]]
+// CHECK-NEXT: %[[T:.*]] = memref.load %[[ARG0]][%[[IV]]]
// CHECK-NEXT: %[[COND:.*]] = cmpf ugt, %[[T]], %[[ZERO]]
// CHECK-NEXT: %[[IFRES:.*]] = scf.if %[[COND]] -> (f32) {
// CHECK-NEXT: %[[THENRES:.*]] = addf %[[ITER]], %[[T]]
diff --git a/mlir/test/Dialect/SCF/parallel-loop-fusion.mlir b/mlir/test/Dialect/SCF/parallel-loop-fusion.mlir
index 971e1c8041b8..2c2541cf9d10 100644
--- a/mlir/test/Dialect/SCF/parallel-loop-fusion.mlir
+++ b/mlir/test/Dialect/SCF/parallel-loop-fusion.mlir
@@ -29,22 +29,22 @@ func @fuse_two(%A: memref<2x2xf32>, %B: memref<2x2xf32>,
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %sum = alloc() : memref<2x2xf32>
+ %sum = memref.alloc() : memref<2x2xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %B_elem = load %B[%i, %j] : memref<2x2xf32>
- %C_elem = load %C[%i, %j] : memref<2x2xf32>
+ %B_elem = memref.load %B[%i, %j] : memref<2x2xf32>
+ %C_elem = memref.load %C[%i, %j] : memref<2x2xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
+ memref.store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %sum_elem = load %sum[%i, %j] : memref<2x2xf32>
- %A_elem = load %A[%i, %j] : memref<2x2xf32>
+ %sum_elem = memref.load %sum[%i, %j] : memref<2x2xf32>
+ %A_elem = memref.load %A[%i, %j] : memref<2x2xf32>
%product_elem = mulf %sum_elem, %A_elem : f32
- store %product_elem, %result[%i, %j] : memref<2x2xf32>
+ memref.store %product_elem, %result[%i, %j] : memref<2x2xf32>
scf.yield
}
- dealloc %sum : memref<2x2xf32>
+ memref.dealloc %sum : memref<2x2xf32>
return
}
// CHECK-LABEL: func @fuse_two
@@ -53,20 +53,20 @@ func @fuse_two(%A: memref<2x2xf32>, %B: memref<2x2xf32>,
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C0:%.*]] = constant 0 : index
// CHECK: [[C1:%.*]] = constant 1 : index
-// CHECK: [[SUM:%.*]] = alloc()
+// CHECK: [[SUM:%.*]] = memref.alloc()
// CHECK: scf.parallel ([[I:%.*]], [[J:%.*]]) = ([[C0]], [[C0]])
// CHECK-SAME: to ([[C2]], [[C2]]) step ([[C1]], [[C1]]) {
-// CHECK: [[B_ELEM:%.*]] = load [[B]]{{\[}}[[I]], [[J]]]
-// CHECK: [[C_ELEM:%.*]] = load [[C]]{{\[}}[[I]], [[J]]]
+// CHECK: [[B_ELEM:%.*]] = memref.load [[B]]{{\[}}[[I]], [[J]]]
+// CHECK: [[C_ELEM:%.*]] = memref.load [[C]]{{\[}}[[I]], [[J]]]
// CHECK: [[SUM_ELEM:%.*]] = addf [[B_ELEM]], [[C_ELEM]]
-// CHECK: store [[SUM_ELEM]], [[SUM]]{{\[}}[[I]], [[J]]]
-// CHECK: [[SUM_ELEM_:%.*]] = load [[SUM]]{{\[}}[[I]], [[J]]]
-// CHECK: [[A_ELEM:%.*]] = load [[A]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[SUM_ELEM]], [[SUM]]{{\[}}[[I]], [[J]]]
+// CHECK: [[SUM_ELEM_:%.*]] = memref.load [[SUM]]{{\[}}[[I]], [[J]]]
+// CHECK: [[A_ELEM:%.*]] = memref.load [[A]]{{\[}}[[I]], [[J]]]
// CHECK: [[PRODUCT_ELEM:%.*]] = mulf [[SUM_ELEM_]], [[A_ELEM]]
-// CHECK: store [[PRODUCT_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[PRODUCT_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
// CHECK: scf.yield
// CHECK: }
-// CHECK: dealloc [[SUM]]
+// CHECK: memref.dealloc [[SUM]]
// -----
@@ -76,28 +76,28 @@ func @fuse_three(%lhs: memref<100x10xf32>, %rhs: memref<100xf32>,
%c10 = constant 10 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %broadcast_rhs = alloc() : memref<100x10xf32>
- %
diff = alloc() : memref<100x10xf32>
+ %broadcast_rhs = memref.alloc() : memref<100x10xf32>
+ %
diff = memref.alloc() : memref<100x10xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c100, %c10) step (%c1, %c1) {
- %rhs_elem = load %rhs[%i] : memref<100xf32>
- store %rhs_elem, %broadcast_rhs[%i, %j] : memref<100x10xf32>
+ %rhs_elem = memref.load %rhs[%i] : memref<100xf32>
+ memref.store %rhs_elem, %broadcast_rhs[%i, %j] : memref<100x10xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c100, %c10) step (%c1, %c1) {
- %lhs_elem = load %lhs[%i, %j] : memref<100x10xf32>
- %broadcast_rhs_elem = load %broadcast_rhs[%i, %j] : memref<100x10xf32>
+ %lhs_elem = memref.load %lhs[%i, %j] : memref<100x10xf32>
+ %broadcast_rhs_elem = memref.load %broadcast_rhs[%i, %j] : memref<100x10xf32>
%
diff _elem = subf %lhs_elem, %broadcast_rhs_elem : f32
- store %
diff _elem, %
diff [%i, %j] : memref<100x10xf32>
+ memref.store %
diff _elem, %
diff [%i, %j] : memref<100x10xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c100, %c10) step (%c1, %c1) {
- %
diff _elem = load %
diff [%i, %j] : memref<100x10xf32>
+ %
diff _elem = memref.load %
diff [%i, %j] : memref<100x10xf32>
%exp_elem = math.exp %
diff _elem : f32
- store %exp_elem, %result[%i, %j] : memref<100x10xf32>
+ memref.store %exp_elem, %result[%i, %j] : memref<100x10xf32>
scf.yield
}
- dealloc %broadcast_rhs : memref<100x10xf32>
- dealloc %
diff : memref<100x10xf32>
+ memref.dealloc %broadcast_rhs : memref<100x10xf32>
+ memref.dealloc %
diff : memref<100x10xf32>
return
}
// CHECK-LABEL: func @fuse_three
@@ -107,23 +107,23 @@ func @fuse_three(%lhs: memref<100x10xf32>, %rhs: memref<100xf32>,
// CHECK: [[C10:%.*]] = constant 10 : index
// CHECK: [[C0:%.*]] = constant 0 : index
// CHECK: [[C1:%.*]] = constant 1 : index
-// CHECK: [[BROADCAST_RHS:%.*]] = alloc()
-// CHECK: [[DIFF:%.*]] = alloc()
+// CHECK: [[BROADCAST_RHS:%.*]] = memref.alloc()
+// CHECK: [[DIFF:%.*]] = memref.alloc()
// CHECK: scf.parallel ([[I:%.*]], [[J:%.*]]) = ([[C0]], [[C0]])
// CHECK-SAME: to ([[C100]], [[C10]]) step ([[C1]], [[C1]]) {
-// CHECK: [[RHS_ELEM:%.*]] = load [[RHS]]{{\[}}[[I]]]
-// CHECK: store [[RHS_ELEM]], [[BROADCAST_RHS]]{{\[}}[[I]], [[J]]]
-// CHECK: [[LHS_ELEM:%.*]] = load [[LHS]]{{\[}}[[I]], [[J]]]
-// CHECK: [[BROADCAST_RHS_ELEM:%.*]] = load [[BROADCAST_RHS]]
+// CHECK: [[RHS_ELEM:%.*]] = memref.load [[RHS]]{{\[}}[[I]]]
+// CHECK: memref.store [[RHS_ELEM]], [[BROADCAST_RHS]]{{\[}}[[I]], [[J]]]
+// CHECK: [[LHS_ELEM:%.*]] = memref.load [[LHS]]{{\[}}[[I]], [[J]]]
+// CHECK: [[BROADCAST_RHS_ELEM:%.*]] = memref.load [[BROADCAST_RHS]]
// CHECK: [[DIFF_ELEM:%.*]] = subf [[LHS_ELEM]], [[BROADCAST_RHS_ELEM]]
-// CHECK: store [[DIFF_ELEM]], [[DIFF]]{{\[}}[[I]], [[J]]]
-// CHECK: [[DIFF_ELEM_:%.*]] = load [[DIFF]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[DIFF_ELEM]], [[DIFF]]{{\[}}[[I]], [[J]]]
+// CHECK: [[DIFF_ELEM_:%.*]] = memref.load [[DIFF]]{{\[}}[[I]], [[J]]]
// CHECK: [[EXP_ELEM:%.*]] = math.exp [[DIFF_ELEM_]]
-// CHECK: store [[EXP_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[EXP_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
// CHECK: scf.yield
// CHECK: }
-// CHECK: dealloc [[BROADCAST_RHS]]
-// CHECK: dealloc [[DIFF]]
+// CHECK: memref.dealloc [[BROADCAST_RHS]]
+// CHECK: memref.dealloc [[DIFF]]
// -----
@@ -196,7 +196,7 @@ func @do_not_fuse_loops_with_side_effecting_ops_in_between() {
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
scf.yield
}
- %buffer = alloc() : memref<2x2xf32>
+ %buffer = memref.alloc() : memref<2x2xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
scf.yield
}
@@ -233,23 +233,23 @@ func @do_not_fuse_unmatching_write_read_patterns(
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %common_buf = alloc() : memref<2x2xf32>
+ %common_buf = memref.alloc() : memref<2x2xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %B_elem = load %B[%i, %j] : memref<2x2xf32>
- %C_elem = load %C[%i, %j] : memref<2x2xf32>
+ %B_elem = memref.load %B[%i, %j] : memref<2x2xf32>
+ %C_elem = memref.load %C[%i, %j] : memref<2x2xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %common_buf[%i, %j] : memref<2x2xf32>
+ memref.store %sum_elem, %common_buf[%i, %j] : memref<2x2xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
%k = addi %i, %c1 : index
- %sum_elem = load %common_buf[%k, %j] : memref<2x2xf32>
- %A_elem = load %A[%i, %j] : memref<2x2xf32>
+ %sum_elem = memref.load %common_buf[%k, %j] : memref<2x2xf32>
+ %A_elem = memref.load %A[%i, %j] : memref<2x2xf32>
%product_elem = mulf %sum_elem, %A_elem : f32
- store %product_elem, %result[%i, %j] : memref<2x2xf32>
+ memref.store %product_elem, %result[%i, %j] : memref<2x2xf32>
scf.yield
}
- dealloc %common_buf : memref<2x2xf32>
+ memref.dealloc %common_buf : memref<2x2xf32>
return
}
// CHECK-LABEL: func @do_not_fuse_unmatching_write_read_patterns
@@ -263,23 +263,23 @@ func @do_not_fuse_unmatching_read_write_patterns(
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %sum = alloc() : memref<2x2xf32>
+ %sum = memref.alloc() : memref<2x2xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %B_elem = load %B[%i, %j] : memref<2x2xf32>
- %C_elem = load %common_buf[%i, %j] : memref<2x2xf32>
+ %B_elem = memref.load %B[%i, %j] : memref<2x2xf32>
+ %C_elem = memref.load %common_buf[%i, %j] : memref<2x2xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
+ memref.store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
%k = addi %i, %c1 : index
- %sum_elem = load %sum[%k, %j] : memref<2x2xf32>
- %A_elem = load %A[%i, %j] : memref<2x2xf32>
+ %sum_elem = memref.load %sum[%k, %j] : memref<2x2xf32>
+ %A_elem = memref.load %A[%i, %j] : memref<2x2xf32>
%product_elem = mulf %sum_elem, %A_elem : f32
- store %product_elem, %common_buf[%j, %i] : memref<2x2xf32>
+ memref.store %product_elem, %common_buf[%j, %i] : memref<2x2xf32>
scf.yield
}
- dealloc %sum : memref<2x2xf32>
+ memref.dealloc %sum : memref<2x2xf32>
return
}
// CHECK-LABEL: func @do_not_fuse_unmatching_read_write_patterns
@@ -292,14 +292,14 @@ func @do_not_fuse_loops_with_memref_defined_in_loop_bodies() {
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %buffer = alloc() : memref<2x2xf32>
+ %buffer = memref.alloc() : memref<2x2xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %A = subview %buffer[%c0, %c0][%c2, %c2][%c1, %c1]
+ %A = memref.subview %buffer[%c0, %c0][%c2, %c2][%c1, %c1]
: memref<2x2xf32> to memref<?x?xf32, offset: ?, strides:[?, ?]>
- %A_elem = load %A[%i, %j] : memref<?x?xf32, offset: ?, strides:[?, ?]>
+ %A_elem = memref.load %A[%i, %j] : memref<?x?xf32, offset: ?, strides:[?, ?]>
scf.yield
}
return
@@ -315,24 +315,24 @@ func @nested_fuse(%A: memref<2x2xf32>, %B: memref<2x2xf32>,
%c2 = constant 2 : index
%c0 = constant 0 : index
%c1 = constant 1 : index
- %sum = alloc() : memref<2x2xf32>
+ %sum = memref.alloc() : memref<2x2xf32>
scf.parallel (%k) = (%c0) to (%c2) step (%c1) {
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %B_elem = load %B[%i, %j] : memref<2x2xf32>
- %C_elem = load %C[%i, %j] : memref<2x2xf32>
+ %B_elem = memref.load %B[%i, %j] : memref<2x2xf32>
+ %C_elem = memref.load %C[%i, %j] : memref<2x2xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
+ memref.store %sum_elem, %sum[%i, %j] : memref<2x2xf32>
scf.yield
}
scf.parallel (%i, %j) = (%c0, %c0) to (%c2, %c2) step (%c1, %c1) {
- %sum_elem = load %sum[%i, %j] : memref<2x2xf32>
- %A_elem = load %A[%i, %j] : memref<2x2xf32>
+ %sum_elem = memref.load %sum[%i, %j] : memref<2x2xf32>
+ %A_elem = memref.load %A[%i, %j] : memref<2x2xf32>
%product_elem = mulf %sum_elem, %A_elem : f32
- store %product_elem, %result[%i, %j] : memref<2x2xf32>
+ memref.store %product_elem, %result[%i, %j] : memref<2x2xf32>
scf.yield
}
}
- dealloc %sum : memref<2x2xf32>
+ memref.dealloc %sum : memref<2x2xf32>
return
}
// CHECK-LABEL: func @nested_fuse
@@ -341,19 +341,19 @@ func @nested_fuse(%A: memref<2x2xf32>, %B: memref<2x2xf32>,
// CHECK: [[C2:%.*]] = constant 2 : index
// CHECK: [[C0:%.*]] = constant 0 : index
// CHECK: [[C1:%.*]] = constant 1 : index
-// CHECK: [[SUM:%.*]] = alloc()
+// CHECK: [[SUM:%.*]] = memref.alloc()
// CHECK: scf.parallel
// CHECK: scf.parallel ([[I:%.*]], [[J:%.*]]) = ([[C0]], [[C0]])
// CHECK-SAME: to ([[C2]], [[C2]]) step ([[C1]], [[C1]]) {
-// CHECK: [[B_ELEM:%.*]] = load [[B]]{{\[}}[[I]], [[J]]]
-// CHECK: [[C_ELEM:%.*]] = load [[C]]{{\[}}[[I]], [[J]]]
+// CHECK: [[B_ELEM:%.*]] = memref.load [[B]]{{\[}}[[I]], [[J]]]
+// CHECK: [[C_ELEM:%.*]] = memref.load [[C]]{{\[}}[[I]], [[J]]]
// CHECK: [[SUM_ELEM:%.*]] = addf [[B_ELEM]], [[C_ELEM]]
-// CHECK: store [[SUM_ELEM]], [[SUM]]{{\[}}[[I]], [[J]]]
-// CHECK: [[SUM_ELEM_:%.*]] = load [[SUM]]{{\[}}[[I]], [[J]]]
-// CHECK: [[A_ELEM:%.*]] = load [[A]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[SUM_ELEM]], [[SUM]]{{\[}}[[I]], [[J]]]
+// CHECK: [[SUM_ELEM_:%.*]] = memref.load [[SUM]]{{\[}}[[I]], [[J]]]
+// CHECK: [[A_ELEM:%.*]] = memref.load [[A]]{{\[}}[[I]], [[J]]]
// CHECK: [[PRODUCT_ELEM:%.*]] = mulf [[SUM_ELEM_]], [[A_ELEM]]
-// CHECK: store [[PRODUCT_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
+// CHECK: memref.store [[PRODUCT_ELEM]], [[RESULT]]{{\[}}[[I]], [[J]]]
// CHECK: scf.yield
// CHECK: }
// CHECK: }
-// CHECK: dealloc [[SUM]]
+// CHECK: memref.dealloc [[SUM]]
diff --git a/mlir/test/Dialect/SCF/parallel-loop-specialization.mlir b/mlir/test/Dialect/SCF/parallel-loop-specialization.mlir
index 959b2f8dabc1..3f23b5077fbe 100644
--- a/mlir/test/Dialect/SCF/parallel-loop-specialization.mlir
+++ b/mlir/test/Dialect/SCF/parallel-loop-specialization.mlir
@@ -7,15 +7,15 @@ func @parallel_loop(%outer_i0: index, %outer_i1: index, %A: memref<?x?xf32>, %B:
%C: memref<?x?xf32>, %result: memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %d0 = dim %A, %c0 : memref<?x?xf32>
- %d1 = dim %A, %c1 : memref<?x?xf32>
+ %d0 = memref.dim %A, %c0 : memref<?x?xf32>
+ %d1 = memref.dim %A, %c1 : memref<?x?xf32>
%b0 = affine.min #map0()[%d0, %outer_i0]
%b1 = affine.min #map1()[%d1, %outer_i1]
scf.parallel (%i0, %i1) = (%c0, %c0) to (%b0, %b1) step (%c1, %c1) {
- %B_elem = load %B[%i0, %i1] : memref<?x?xf32>
- %C_elem = load %C[%i0, %i1] : memref<?x?xf32>
+ %B_elem = memref.load %B[%i0, %i1] : memref<?x?xf32>
+ %C_elem = memref.load %C[%i0, %i1] : memref<?x?xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %result[%i0, %i1] : memref<?x?xf32>
+ memref.store %sum_elem, %result[%i0, %i1] : memref<?x?xf32>
}
return
}
@@ -24,8 +24,8 @@ func @parallel_loop(%outer_i0: index, %outer_i1: index, %A: memref<?x?xf32>, %B:
// CHECK-SAME: [[VAL_0:%.*]]: index, [[VAL_1:%.*]]: index, [[VAL_2:%.*]]: memref<?x?xf32>, [[VAL_3:%.*]]: memref<?x?xf32>, [[VAL_4:%.*]]: memref<?x?xf32>, [[VAL_5:%.*]]: memref<?x?xf32>) {
// CHECK: [[VAL_6:%.*]] = constant 0 : index
// CHECK: [[VAL_7:%.*]] = constant 1 : index
-// CHECK: [[VAL_8:%.*]] = dim [[VAL_2]], [[VAL_6]] : memref<?x?xf32>
-// CHECK: [[VAL_9:%.*]] = dim [[VAL_2]], [[VAL_7]] : memref<?x?xf32>
+// CHECK: [[VAL_8:%.*]] = memref.dim [[VAL_2]], [[VAL_6]] : memref<?x?xf32>
+// CHECK: [[VAL_9:%.*]] = memref.dim [[VAL_2]], [[VAL_7]] : memref<?x?xf32>
// CHECK: [[VAL_10:%.*]] = affine.min #map0(){{\[}}[[VAL_8]], [[VAL_0]]]
// CHECK: [[VAL_11:%.*]] = affine.min #map1(){{\[}}[[VAL_9]], [[VAL_1]]]
// CHECK: [[VAL_12:%.*]] = constant 1024 : index
@@ -35,11 +35,11 @@ func @parallel_loop(%outer_i0: index, %outer_i1: index, %A: memref<?x?xf32>, %B:
// CHECK: [[VAL_16:%.*]] = and [[VAL_13]], [[VAL_15]] : i1
// CHECK: scf.if [[VAL_16]] {
// CHECK: scf.parallel ([[VAL_17:%.*]], [[VAL_18:%.*]]) = ([[VAL_6]], [[VAL_6]]) to ([[VAL_12]], [[VAL_14]]) step ([[VAL_7]], [[VAL_7]]) {
-// CHECK: store
+// CHECK: memref.store
// CHECK: }
// CHECK: } else {
// CHECK: scf.parallel ([[VAL_22:%.*]], [[VAL_23:%.*]]) = ([[VAL_6]], [[VAL_6]]) to ([[VAL_10]], [[VAL_11]]) step ([[VAL_7]], [[VAL_7]]) {
-// CHECK: store
+// CHECK: memref.store
// CHECK: }
// CHECK: }
// CHECK: return
diff --git a/mlir/test/Dialect/SCF/parallel-loop-tiling.mlir b/mlir/test/Dialect/SCF/parallel-loop-tiling.mlir
index 5d3a676f58ab..5a8f765da421 100644
--- a/mlir/test/Dialect/SCF/parallel-loop-tiling.mlir
+++ b/mlir/test/Dialect/SCF/parallel-loop-tiling.mlir
@@ -5,10 +5,10 @@ func @parallel_loop(%arg0 : index, %arg1 : index, %arg2 : index,
%A: memref<?x?xf32>, %B: memref<?x?xf32>,
%C: memref<?x?xf32>, %result: memref<?x?xf32>) {
scf.parallel (%i0, %i1) = (%arg0, %arg1) to (%arg2, %arg3) step (%arg4, %arg5) {
- %B_elem = load %B[%i0, %i1] : memref<?x?xf32>
- %C_elem = load %C[%i0, %i1] : memref<?x?xf32>
+ %B_elem = memref.load %B[%i0, %i1] : memref<?x?xf32>
+ %C_elem = memref.load %C[%i0, %i1] : memref<?x?xf32>
%sum_elem = addf %B_elem, %C_elem : f32
- store %sum_elem, %result[%i0, %i1] : memref<?x?xf32>
+ memref.store %sum_elem, %result[%i0, %i1] : memref<?x?xf32>
}
return
}
@@ -27,10 +27,10 @@ func @parallel_loop(%arg0 : index, %arg1 : index, %arg2 : index,
// CHECK: scf.parallel ([[V7:%.*]], [[V8:%.*]]) = ([[C0]], [[C0]]) to ([[V5]], [[V6]]) step ([[ARG5]], [[ARG6]]) {
// CHECK: [[V9:%.*]] = addi [[V7]], [[V3]] : index
// CHECK: [[V10:%.*]] = addi [[V8]], [[V4]] : index
-// CHECK: [[V11:%.*]] = load [[ARG8]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
-// CHECK: [[V12:%.*]] = load [[ARG9]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
+// CHECK: [[V11:%.*]] = memref.load [[ARG8]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
+// CHECK: [[V12:%.*]] = memref.load [[ARG9]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
// CHECK: [[V13:%.*]] = addf [[V11]], [[V12]] : f32
-// CHECK: store [[V13]], [[ARG10]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
+// CHECK: memref.store [[V13]], [[ARG10]]{{\[}}[[V9]], [[V10]]] : memref<?x?xf32>
// CHECK: }
// CHECK: }
// CHECK: return
diff --git a/mlir/test/Dialect/Shape/bufferize.mlir b/mlir/test/Dialect/Shape/bufferize.mlir
index cb65a5d42d4b..a912d5f5dc82 100644
--- a/mlir/test/Dialect/Shape/bufferize.mlir
+++ b/mlir/test/Dialect/Shape/bufferize.mlir
@@ -6,10 +6,10 @@
// CHECK: %[[WTRUE:.*]] = shape.const_witness true
// CHECK: %[[MEMREF:.*]] = shape.assuming %[[WTRUE]] -> (memref<2xf16>) {
// CHECK: %[[TENSOR_VAL:.*]] = "test.source"() : () -> tensor<2xf16>
-// CHECK: %[[YIELDED_MEMREF:.*]] = tensor_to_memref %[[TENSOR_VAL]] : memref<2xf16>
+// CHECK: %[[YIELDED_MEMREF:.*]] = memref.buffer_cast %[[TENSOR_VAL]] : memref<2xf16>
// CHECK: shape.assuming_yield %[[YIELDED_MEMREF]] : memref<2xf16>
// CHECK: }
-// CHECK: %[[TENSOR:.*]] = tensor_load %[[MEMREF:.*]] : memref<2xf16>
+// CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[MEMREF:.*]] : memref<2xf16>
// CHECK: "test.sink"(%[[TENSOR]]) : (tensor<2xf16>) -> ()
// CHECK: return
// CHECK: }
diff --git a/mlir/test/Dialect/Standard/bufferize.mlir b/mlir/test/Dialect/Standard/bufferize.mlir
index 10310542f138..7ed51ca9293f 100644
--- a/mlir/test/Dialect/Standard/bufferize.mlir
+++ b/mlir/test/Dialect/Standard/bufferize.mlir
@@ -3,11 +3,11 @@
// CHECK-LABEL: func @dim(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<f32>,
// CHECK-SAME: %[[INDEX:.*]]: index) -> index {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
-// CHECK: %[[EXTENT:.*]] = dim %[[MEMREF]], %[[INDEX]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
+// CHECK: %[[EXTENT:.*]] = memref.dim %[[MEMREF]], %[[INDEX]] : memref<f32>
// CHECK: return %[[EXTENT]] : index
func @dim(%arg0: tensor<f32>, %arg1: index) -> index {
- %0 = dim %arg0, %arg1 : tensor<f32>
+ %0 = memref.dim %arg0, %arg1 : tensor<f32>
return %0 : index
}
@@ -15,10 +15,10 @@ func @dim(%arg0: tensor<f32>, %arg1: index) -> index {
// CHECK-SAME: %[[PRED:.*]]: i1,
// CHECK-SAME: %[[TRUE_VAL:.*]]: tensor<f32>,
// CHECK-SAME: %[[FALSE_VAL:.*]]: tensor<f32>) -> tensor<f32> {
-// CHECK: %[[TRUE_VAL_MEMREF:.*]] = tensor_to_memref %[[TRUE_VAL]] : memref<f32>
-// CHECK: %[[FALSE_VAL_MEMREF:.*]] = tensor_to_memref %[[FALSE_VAL]] : memref<f32>
+// CHECK: %[[TRUE_VAL_MEMREF:.*]] = memref.buffer_cast %[[TRUE_VAL]] : memref<f32>
+// CHECK: %[[FALSE_VAL_MEMREF:.*]] = memref.buffer_cast %[[FALSE_VAL]] : memref<f32>
// CHECK: %[[RET_MEMREF:.*]] = select %[[PRED]], %[[TRUE_VAL_MEMREF]], %[[FALSE_VAL_MEMREF]] : memref<f32>
-// CHECK: %[[RET:.*]] = tensor_load %[[RET_MEMREF]] : memref<f32>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[RET_MEMREF]] : memref<f32>
// CHECK: return %[[RET]] : tensor<f32>
func @select(%arg0: i1, %arg1: tensor<f32>, %arg2: tensor<f32>) -> tensor<f32> {
%0 = select %arg0, %arg1, %arg2 : tensor<f32>
diff --git a/mlir/test/Dialect/Standard/canonicalize.mlir b/mlir/test/Dialect/Standard/canonicalize.mlir
index 2d722eaaaa07..41ae2248b299 100644
--- a/mlir/test/Dialect/Standard/canonicalize.mlir
+++ b/mlir/test/Dialect/Standard/canonicalize.mlir
@@ -2,25 +2,25 @@
// -----
-// Test case: Basic folding of tensor_load(tensor_to_memref(t)) -> t
-// CHECK-LABEL: func @tensor_load_of_tensor_to_memref(
+// Test case: Basic folding of memref.tensor_load(memref.buffer_cast(t)) -> t
+// CHECK-LABEL: func @tensor_load_of_buffer_cast(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<?xf32>) -> tensor<?xf32> {
// CHECK: return %[[TENSOR]]
-func @tensor_load_of_tensor_to_memref(%arg0: tensor<?xf32>) -> tensor<?xf32> {
- %0 = tensor_to_memref %arg0 : memref<?xf32>
- %1 = tensor_load %0 : memref<?xf32>
+func @tensor_load_of_buffer_cast(%arg0: tensor<?xf32>) -> tensor<?xf32> {
+ %0 = memref.buffer_cast %arg0 : memref<?xf32>
+ %1 = memref.tensor_load %0 : memref<?xf32>
return %1 : tensor<?xf32>
}
// -----
-// Test case: Basic folding of tensor_to_memref(tensor_load(m)) -> m
-// CHECK-LABEL: func @tensor_to_memref_of_tensor_load(
+// Test case: Basic folding of memref.buffer_cast(memref.tensor_load(m)) -> m
+// CHECK-LABEL: func @buffer_cast_of_tensor_load(
// CHECK-SAME: %[[MEMREF:.*]]: memref<?xf32>) -> memref<?xf32> {
// CHECK: return %[[MEMREF]]
-func @tensor_to_memref_of_tensor_load(%arg0: memref<?xf32>) -> memref<?xf32> {
- %0 = tensor_load %arg0 : memref<?xf32>
- %1 = tensor_to_memref %0 : memref<?xf32>
+func @buffer_cast_of_tensor_load(%arg0: memref<?xf32>) -> memref<?xf32> {
+ %0 = memref.tensor_load %arg0 : memref<?xf32>
+ %1 = memref.buffer_cast %0 : memref<?xf32>
return %1 : memref<?xf32>
}
@@ -29,14 +29,14 @@ func @tensor_to_memref_of_tensor_load(%arg0: memref<?xf32>) -> memref<?xf32> {
// Test case: If the memrefs are not the same type, don't fold them.
// Test case: If the memrefs are not cast-compatible (e.g.
diff erent address space),
// don't canonicalize them either.
-// CHECK-LABEL: func @no_fold_tensor_to_memref_of_tensor_load(
+// CHECK-LABEL: func @no_fold_buffer_cast_of_tensor_load(
// CHECK-SAME: %[[MEMREF_ADDRSPACE2:.*]]: memref<?xf32, 2>) -> memref<?xf32, 7> {
-// CHECK: %[[TENSOR:.*]] = tensor_load %[[MEMREF_ADDRSPACE2]] : memref<?xf32, 2>
-// CHECK: %[[MEMREF_ADDRSPACE7:.*]] = tensor_to_memref %[[TENSOR]] : memref<?xf32, 7>
+// CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[MEMREF_ADDRSPACE2]] : memref<?xf32, 2>
+// CHECK: %[[MEMREF_ADDRSPACE7:.*]] = memref.buffer_cast %[[TENSOR]] : memref<?xf32, 7>
// CHECK: return %[[MEMREF_ADDRSPACE7]]
-func @no_fold_tensor_to_memref_of_tensor_load(%arg0: memref<?xf32, 2>) -> memref<?xf32, 7> {
- %0 = tensor_load %arg0 : memref<?xf32, 2>
- %1 = tensor_to_memref %0 : memref<?xf32, 7>
+func @no_fold_buffer_cast_of_tensor_load(%arg0: memref<?xf32, 2>) -> memref<?xf32, 7> {
+ %0 = memref.tensor_load %arg0 : memref<?xf32, 2>
+ %1 = memref.buffer_cast %0 : memref<?xf32, 7>
return %1 : memref<?xf32, 7>
}
@@ -46,57 +46,57 @@ func @no_fold_tensor_to_memref_of_tensor_load(%arg0: memref<?xf32, 2>) -> memref
// CHECK-DAG: #[[$OFF_UNK:[a-z0-9]+]] = affine_map<(d0)[s0] -> (d0 + s0)>
// Test case: If the memrefs are cast-compatible, canonicalize.
-// CHECK-LABEL: func @canonicalize_tensor_to_memref_of_tensor_load(
+// CHECK-LABEL: func @canonicalize_buffer_cast_of_tensor_load(
// CHECK-SAME: %[[M:.*]]: memref<?xf32, #[[$OFF_3]]>) -> memref<?xf32, #[[$OFF_UNK]]> {
-// CHECK-NOT: tensor_load
-// CHECK-NOT: tensor_to_memref
-// CHECK: %[[R:.*]] = memref_cast %[[M]] : memref<?xf32, #[[$OFF_3]]> to memref<?xf32, #[[$OFF_UNK]]>
+// CHECK-NOT: memref.tensor_load
+// CHECK-NOT: memref.buffer_cast
+// CHECK: %[[R:.*]] = memref.cast %[[M]] : memref<?xf32, #[[$OFF_3]]> to memref<?xf32, #[[$OFF_UNK]]>
// CHECK: return %[[R]]
-func @canonicalize_tensor_to_memref_of_tensor_load(%arg0: memref<?xf32, offset: 3, strides: [1]>)
+func @canonicalize_buffer_cast_of_tensor_load(%arg0: memref<?xf32, offset: 3, strides: [1]>)
-> memref<?xf32, offset: ?, strides: [1]>
{
- %0 = tensor_load %arg0 : memref<?xf32, offset: 3, strides: [1]>
- %1 = tensor_to_memref %0 : memref<?xf32, offset: ?, strides: [1]>
+ %0 = memref.tensor_load %arg0 : memref<?xf32, offset: 3, strides: [1]>
+ %1 = memref.buffer_cast %0 : memref<?xf32, offset: ?, strides: [1]>
return %1 : memref<?xf32, offset: ?, strides: [1]>
}
// -----
-// Test case: Basic folding of dim(tensor_load(m)) -> dim(m).
+// Test case: Basic folding of memref.dim(memref.tensor_load(m)) -> memref.dim(m).
// CHECK-LABEL: func @dim_of_tensor_load(
// CHECK-SAME: %[[MEMREF:[0-9a-z]*]]: memref<?xf32>
// CHECK: %[[C0:.*]] = constant 0
-// CHECK: %[[D:.*]] = dim %[[MEMREF]], %[[C0]]
+// CHECK: %[[D:.*]] = memref.dim %[[MEMREF]], %[[C0]]
// CHECK: return %[[D]] : index
func @dim_of_tensor_load(%arg0: memref<?xf32>) -> index {
%c0 = constant 0 : index
- %0 = tensor_load %arg0 : memref<?xf32>
- %1 = dim %0, %c0 : tensor<?xf32>
+ %0 = memref.tensor_load %arg0 : memref<?xf32>
+ %1 = memref.dim %0, %c0 : tensor<?xf32>
return %1 : index
}
// -----
-// Test case: Folding of load(tensor_to_memref(%v, %idxs))
+// Test case: Folding of memref.load(memref.buffer_cast(%v, %idxs))
// -> tensor.extract(%v, %idx)
-// CHECK-LABEL: func @load_from_tensor_to_memref(
+// CHECK-LABEL: func @load_from_buffer_cast(
// CHECK-SAME: %[[IDX0:[0-9a-z]+]]: index, %[[IDX1:[0-9a-z]+]]: index
// CHECK-SAME: %[[TENSOR:[0-9a-z]+]]: tensor<?x?xf32>
// CHECK: %[[RES:.*]] = tensor.extract %[[TENSOR]][%[[IDX0]], %[[IDX1]]]
-// CHECK-NOT: load
+// CHECK-NOT: memref.load
// CHECK: return %[[RES]] : f32
-func @load_from_tensor_to_memref(%arg0: index, %arg1: index, %arg2: tensor<?x?xf32>) -> f32 {
- %0 = tensor_to_memref %arg2 : memref<?x?xf32>
- %1 = load %0[%arg0, %arg1] : memref<?x?xf32>
+func @load_from_buffer_cast(%arg0: index, %arg1: index, %arg2: tensor<?x?xf32>) -> f32 {
+ %0 = memref.buffer_cast %arg2 : memref<?x?xf32>
+ %1 = memref.load %0[%arg0, %arg1] : memref<?x?xf32>
return %1 : f32
}
// -----
-// Test case: Folding of dim(tensor.generate %idx) -> %idx
+// Test case: Folding of memref.dim(tensor.generate %idx) -> %idx
// CHECK-LABEL: func @dim_of_tensor.generate(
// CHECK-SAME: %[[IDX0:[0-9a-z]+]]: index, %[[IDX1:[0-9a-z]+]]: index
-// CHECK-NOT: dim
+// CHECK-NOT: memref.dim
// CHECK: return %[[IDX1]] : index
func @dim_of_tensor.generate(%arg0: index, %arg1: index) -> index {
%c3 = constant 3 : index
@@ -104,7 +104,7 @@ func @dim_of_tensor.generate(%arg0: index, %arg1: index) -> index {
^bb0(%arg2: index, %arg3: index, %arg4: index, %arg5: index, %arg6: index):
tensor.yield %c3 : index
} : tensor<2x?x4x?x5xindex>
- %1 = dim %0, %c3 : tensor<2x?x4x?x5xindex>
+ %1 = memref.dim %0, %c3 : tensor<2x?x4x?x5xindex>
return %1 : index
}
@@ -134,41 +134,41 @@ func @cmpi_equal_operands(%arg0: i64)
// -----
-// Test case: Folding of dim(memref_reshape %v %shp, %idx) -> load %shp[%idx]
+// Test case: Folding of memref.dim(memref.reshape %v %shp, %idx) -> memref.load %shp[%idx]
// CHECK-LABEL: func @dim_of_memref_reshape(
// CHECK-SAME: %[[MEM:[0-9a-z]+]]: memref<*xf32>,
// CHECK-SAME: %[[SHP:[0-9a-z]+]]: memref<?xindex>
// CHECK-NEXT: %[[IDX:.*]] = constant 3
-// CHECK-NEXT: %[[DIM:.*]] = load %[[SHP]][%[[IDX]]]
-// CHECK-NEXT: store
-// CHECK-NOT: dim
+// CHECK-NEXT: %[[DIM:.*]] = memref.load %[[SHP]][%[[IDX]]]
+// CHECK-NEXT: memref.store
+// CHECK-NOT: memref.dim
// CHECK: return %[[DIM]] : index
func @dim_of_memref_reshape(%arg0: memref<*xf32>, %arg1: memref<?xindex>)
-> index {
%c3 = constant 3 : index
- %0 = memref_reshape %arg0(%arg1)
+ %0 = memref.reshape %arg0(%arg1)
: (memref<*xf32>, memref<?xindex>) -> memref<*xf32>
// Update the shape to test that he load ends up in the right place.
- store %c3, %arg1[%c3] : memref<?xindex>
- %1 = dim %0, %c3 : memref<*xf32>
+ memref.store %c3, %arg1[%c3] : memref<?xindex>
+ %1 = memref.dim %0, %c3 : memref<*xf32>
return %1 : index
}
// -----
-// Test case: Folding dim(tensor.cast %0, %idx) -> dim %0, %idx
+// Test case: Folding memref.dim(tensor.cast %0, %idx) -> memref.dim %0, %idx
// CHECK-LABEL: func @fold_dim_of_tensor.cast
// CHECK-SAME: %[[ARG0:.[a-z0-9A-Z_]+]]: tensor<4x?xf32>
// CHECK-DAG: %[[C1:.+]] = constant 1 : index
// CHECK-DAG: %[[C4:.+]] = constant 4 : index
-// CHECK: %[[T0:.+]] = dim %[[ARG0]], %[[C1]]
+// CHECK: %[[T0:.+]] = memref.dim %[[ARG0]], %[[C1]]
// CHECK-NEXT: return %[[C4]], %[[T0]]
func @fold_dim_of_tensor.cast(%arg0 : tensor<4x?xf32>) -> (index, index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%0 = tensor.cast %arg0 : tensor<4x?xf32> to tensor<?x?xf32>
- %1 = dim %0, %c0 : tensor<?x?xf32>
- %2 = dim %0, %c1 : tensor<?x?xf32>
+ %1 = memref.dim %0, %c0 : tensor<?x?xf32>
+ %2 = memref.dim %0, %c1 : tensor<?x?xf32>
return %1, %2: index, index
}
@@ -176,13 +176,13 @@ func @fold_dim_of_tensor.cast(%arg0 : tensor<4x?xf32>) -> (index, index) {
// CHECK-LABEL: func @tensor_cast_to_memref
// CHECK-SAME: %[[ARG0:.+]]: tensor<4x6x16x32xi8>
-// CHECK: %[[M:.+]] = tensor_to_memref %[[ARG0]] : memref<4x6x16x32xi8>
-// CHECK: %[[M1:.+]] = memref_cast %[[M]] : memref<4x6x16x32xi8> to memref<?x?x16x32xi8>
+// CHECK: %[[M:.+]] = memref.buffer_cast %[[ARG0]] : memref<4x6x16x32xi8>
+// CHECK: %[[M1:.+]] = memref.cast %[[M]] : memref<4x6x16x32xi8> to memref<?x?x16x32xi8>
// CHECK: return %[[M1]] : memref<?x?x16x32xi8>
func @tensor_cast_to_memref(%arg0 : tensor<4x6x16x32xi8>) ->
memref<?x?x16x32xi8> {
%0 = tensor.cast %arg0 : tensor<4x6x16x32xi8> to tensor<?x?x16x32xi8>
- %1 = tensor_to_memref %0 : memref<?x?x16x32xi8>
+ %1 = memref.buffer_cast %0 : memref<?x?x16x32xi8>
return %1 : memref<?x?x16x32xi8>
}
@@ -190,13 +190,13 @@ func @tensor_cast_to_memref(%arg0 : tensor<4x6x16x32xi8>) ->
// CHECK-LABEL: func @subview_of_memcast
// CHECK-SAME: %[[ARG0:.[a-z0-9A-Z_]+]]: memref<4x6x16x32xi8>
-// CHECK: %[[S:.+]] = subview %arg0[0, 1, 0] [1, 1, 16] [1, 1, 1] : memref<4x6x16x32xi8> to memref<16x32xi8, #{{.*}}>
-// CHECK: %[[M:.+]] = memref_cast %[[S]] : memref<16x32xi8, #{{.*}}> to memref<16x32xi8, #{{.*}}>
+// CHECK: %[[S:.+]] = memref.subview %arg0[0, 1, 0] [1, 1, 16] [1, 1, 1] : memref<4x6x16x32xi8> to memref<16x32xi8, #{{.*}}>
+// CHECK: %[[M:.+]] = memref.cast %[[S]] : memref<16x32xi8, #{{.*}}> to memref<16x32xi8, #{{.*}}>
// CHECK: return %[[M]] : memref<16x32xi8, #{{.*}}>
func @subview_of_memcast(%arg : memref<4x6x16x32xi8>) ->
memref<16x32xi8, affine_map<(d0, d1)[s0] -> (d0 * 32 + d1 + s0)>>{
- %0 = memref_cast %arg : memref<4x6x16x32xi8> to memref<?x?x16x32xi8>
- %1 = subview %0[0, 1, 0] [1, 1, 16] [1, 1, 1] :
+ %0 = memref.cast %arg : memref<4x6x16x32xi8> to memref<?x?x16x32xi8>
+ %1 = memref.subview %0[0, 1, 0] [1, 1, 16] [1, 1, 1] :
memref<?x?x16x32xi8> to
memref<16x32xi8, affine_map<(d0, d1)[s0] -> (d0 * 32 + d1 + s0)>>
return %1 : memref<16x32xi8, affine_map<(d0, d1)[s0] -> (d0 * 32 + d1 + s0)>>
@@ -206,10 +206,10 @@ func @subview_of_memcast(%arg : memref<4x6x16x32xi8>) ->
// CHECK-LABEL: func @subview_of_static_full_size
// CHECK-SAME: %[[ARG0:.+]]: memref<4x6x16x32xi8>
-// CHECK-NOT: subview
+// CHECK-NOT: memref.subview
// CHECK: return %[[ARG0]] : memref<4x6x16x32xi8>
func @subview_of_static_full_size(%arg0 : memref<4x6x16x32xi8>) -> memref<4x6x16x32xi8> {
- %0 = subview %arg0[0, 0, 0, 0] [4, 6, 16, 32] [1, 1, 1, 1] : memref<4x6x16x32xi8> to memref<4x6x16x32xi8>
+ %0 = memref.subview %arg0[0, 0, 0, 0] [4, 6, 16, 32] [1, 1, 1, 1] : memref<4x6x16x32xi8> to memref<4x6x16x32xi8>
return %0 : memref<4x6x16x32xi8>
}
@@ -272,7 +272,7 @@ func @subtensor_canonicalize(%arg0 : tensor<2x?xi32>, %arg1 : tensor<i32>,
%c1 = constant 1 : index
%c2 = constant 2 : index
%c8 = constant 8 : index
- %0 = dim %arg0, %c1 : tensor<2x?xi32>
+ %0 = memref.dim %arg0, %c1 : tensor<2x?xi32>
%1 = tensor.extract %arg1[] : tensor<i32>
%2 = tensor.generate %arg2, %c8 {
^bb0(%arg4: index, %arg5: index):
diff --git a/mlir/test/Dialect/Standard/expand-ops.mlir b/mlir/test/Dialect/Standard/expand-ops.mlir
index b28c73ea469a..f91f9b8ff373 100644
--- a/mlir/test/Dialect/Standard/expand-ops.mlir
+++ b/mlir/test/Dialect/Standard/expand-ops.mlir
@@ -85,7 +85,7 @@ func @floordivi(%arg0: i32, %arg1: i32) -> (i32) {
// CHECK-LABEL: func @memref_reshape(
func @memref_reshape(%input: memref<*xf32>,
%shape: memref<3xi32>) -> memref<?x?x?xf32> {
- %result = memref_reshape %input(%shape)
+ %result = memref.reshape %input(%shape)
: (memref<*xf32>, memref<3xi32>) -> memref<?x?x?xf32>
return %result : memref<?x?x?xf32>
}
@@ -94,20 +94,20 @@ func @memref_reshape(%input: memref<*xf32>,
// CHECK: [[C1:%.*]] = constant 1 : index
// CHECK: [[C2:%.*]] = constant 2 : index
-// CHECK: [[DIM_2:%.*]] = load [[SHAPE]]{{\[}}[[C2]]] : memref<3xi32>
+// CHECK: [[DIM_2:%.*]] = memref.load [[SHAPE]]{{\[}}[[C2]]] : memref<3xi32>
// CHECK: [[SIZE_2:%.*]] = index_cast [[DIM_2]] : i32 to index
// CHECK: [[STRIDE_1:%.*]] = muli [[C1]], [[SIZE_2]] : index
// CHECK: [[C1_:%.*]] = constant 1 : index
-// CHECK: [[DIM_1:%.*]] = load [[SHAPE]]{{\[}}[[C1_]]] : memref<3xi32>
+// CHECK: [[DIM_1:%.*]] = memref.load [[SHAPE]]{{\[}}[[C1_]]] : memref<3xi32>
// CHECK: [[SIZE_1:%.*]] = index_cast [[DIM_1]] : i32 to index
// CHECK: [[STRIDE_0:%.*]] = muli [[STRIDE_1]], [[SIZE_1]] : index
// CHECK: [[C0:%.*]] = constant 0 : index
-// CHECK: [[DIM_0:%.*]] = load [[SHAPE]]{{\[}}[[C0]]] : memref<3xi32>
+// CHECK: [[DIM_0:%.*]] = memref.load [[SHAPE]]{{\[}}[[C0]]] : memref<3xi32>
// CHECK: [[SIZE_0:%.*]] = index_cast [[DIM_0]] : i32 to index
-// CHECK: [[RESULT:%.*]] = memref_reinterpret_cast [[SRC]]
+// CHECK: [[RESULT:%.*]] = memref.reinterpret_cast [[SRC]]
// CHECK-SAME: to offset: [0], sizes: {{\[}}[[SIZE_0]], [[SIZE_1]], [[SIZE_2]]],
// CHECK-SAME: strides: {{\[}}[[STRIDE_0]], [[STRIDE_1]], [[C1]]]
// CHECK-SAME: : memref<*xf32> to memref<?x?x?xf32>
diff --git a/mlir/test/Dialect/Standard/func-bufferize.mlir b/mlir/test/Dialect/Standard/func-bufferize.mlir
index de2f75c4a293..e4d6a8cf1de1 100644
--- a/mlir/test/Dialect/Standard/func-bufferize.mlir
+++ b/mlir/test/Dialect/Standard/func-bufferize.mlir
@@ -2,8 +2,8 @@
// CHECK-LABEL: func @identity(
// CHECK-SAME: %[[ARG:.*]]: memref<f32>) -> memref<f32> {
-// CHECK: %[[TENSOR:.*]] = tensor_load %[[ARG]] : memref<f32>
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
+// CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[ARG]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
// CHECK: return %[[MEMREF]] : memref<f32>
func @identity(%arg0: tensor<f32>) -> tensor<f32> {
return %arg0 : tensor<f32>
@@ -11,12 +11,12 @@ func @identity(%arg0: tensor<f32>) -> tensor<f32> {
// CHECK-LABEL: func @block_arguments(
// CHECK-SAME: %[[ARG:.*]]: memref<f32>) -> memref<f32> {
-// CHECK: %[[T1:.*]] = tensor_load %[[ARG]] : memref<f32>
-// CHECK: %[[M1:.*]] = tensor_to_memref %[[T1]] : memref<f32>
+// CHECK: %[[T1:.*]] = memref.tensor_load %[[ARG]] : memref<f32>
+// CHECK: %[[M1:.*]] = memref.buffer_cast %[[T1]] : memref<f32>
// CHECK: br ^bb1(%[[M1]] : memref<f32>)
// CHECK: ^bb1(%[[BBARG:.*]]: memref<f32>):
-// CHECK: %[[T2:.*]] = tensor_load %[[BBARG]] : memref<f32>
-// CHECK: %[[M2:.*]] = tensor_to_memref %[[T2]] : memref<f32>
+// CHECK: %[[T2:.*]] = memref.tensor_load %[[BBARG]] : memref<f32>
+// CHECK: %[[M2:.*]] = memref.buffer_cast %[[T2]] : memref<f32>
// CHECK: return %[[M2]] : memref<f32>
func @block_arguments(%arg0: tensor<f32>) -> tensor<f32> {
br ^bb1(%arg0: tensor<f32>)
@@ -35,8 +35,8 @@ func @call_source() -> tensor<f32> {
}
// CHECK-LABEL: func @call_sink(
// CHECK-SAME: %[[ARG:.*]]: memref<f32>) {
-// CHECK: %[[TENSOR:.*]] = tensor_load %[[ARG]] : memref<f32>
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
+// CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[ARG]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
// CHECK: call @sink(%[[MEMREF]]) : (memref<f32>) -> ()
// CHECK: return
func private @sink(tensor<f32>)
@@ -47,7 +47,7 @@ func @call_sink(%arg0: tensor<f32>) {
// CHECK-LABEL: func @unconverted_op_in_body() -> memref<f32> {
// CHECK: %[[TENSOR:.*]] = "test.source"() : () -> tensor<f32>
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<f32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<f32>
// CHECK: return %[[MEMREF]] : memref<f32>
func @unconverted_op_in_body() -> tensor<f32> {
%0 = "test.source"() : () -> tensor<f32>
diff --git a/mlir/test/Dialect/Standard/invalid.mlir b/mlir/test/Dialect/Standard/invalid.mlir
index c6f11a871848..bb8bf3374002 100644
--- a/mlir/test/Dialect/Standard/invalid.mlir
+++ b/mlir/test/Dialect/Standard/invalid.mlir
@@ -18,28 +18,28 @@ func @test_index_cast_tensor_error(%arg0 : tensor<index>) -> i64 {
func @transpose_not_permutation(%v : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>) {
// expected-error @+1 {{expected a permutation map}}
- transpose %v (i, j) -> (i, i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
+ memref.transpose %v (i, j) -> (i, i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
}
// -----
func @transpose_bad_rank(%v : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>) {
// expected-error @+1 {{expected a permutation map of same rank as the input}}
- transpose %v (i) -> (i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
+ memref.transpose %v (i) -> (i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
}
// -----
func @transpose_wrong_type(%v : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>) {
// expected-error @+1 {{output type 'memref<?x?xf32, affine_map<(d0, d1)[s0, s1] -> (d0 * s1 + s0 + d1)>>' does not match transposed input type 'memref<?x?xf32, affine_map<(d0, d1)[s0, s1] -> (d0 * s1 + s0 + d1)>>'}}
- transpose %v (i, j) -> (j, i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
+ memref.transpose %v (i, j) -> (j, i) : memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>> to memref<?x?xf32, affine_map<(i, j)[off, M]->(off + M * i + j)>>
}
// -----
func @memref_reinterpret_cast_too_many_offsets(%in: memref<?xf32>) {
// expected-error @+1 {{expected <= 1 offset values}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [0, 0], sizes: [10, 10], strides: [10, 1]
: memref<?xf32> to memref<10x10xf32, offset: 0, strides: [10, 1]>
return
@@ -49,7 +49,7 @@ func @memref_reinterpret_cast_too_many_offsets(%in: memref<?xf32>) {
func @memref_reinterpret_cast_incompatible_element_types(%in: memref<*xf32>) {
// expected-error @+1 {{
diff erent element types specified}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [0], sizes: [10], strides: [1]
: memref<*xf32> to memref<10xi32, offset: 0, strides: [1]>
return
@@ -59,7 +59,7 @@ func @memref_reinterpret_cast_incompatible_element_types(%in: memref<*xf32>) {
func @memref_reinterpret_cast_incompatible_memory_space(%in: memref<*xf32>) {
// expected-error @+1 {{
diff erent memory spaces specified}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [0], sizes: [10], strides: [1]
: memref<*xf32> to memref<10xi32, offset: 0, strides: [1], 2>
return
@@ -69,7 +69,7 @@ func @memref_reinterpret_cast_incompatible_memory_space(%in: memref<*xf32>) {
func @memref_reinterpret_cast_offset_mismatch(%in: memref<?xf32>) {
// expected-error @+1 {{expected result type with offset = 2 instead of 1}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [1], sizes: [10], strides: [1]
: memref<?xf32> to memref<10xf32, offset: 2, strides: [1]>
return
@@ -79,7 +79,7 @@ func @memref_reinterpret_cast_offset_mismatch(%in: memref<?xf32>) {
func @memref_reinterpret_cast_size_mismatch(%in: memref<*xf32>) {
// expected-error @+1 {{expected result type with size = 10 instead of 1 in dim = 0}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [0], sizes: [10], strides: [1]
: memref<*xf32> to memref<1xf32, offset: 0, strides: [1]>
return
@@ -89,7 +89,7 @@ func @memref_reinterpret_cast_size_mismatch(%in: memref<*xf32>) {
func @memref_reinterpret_cast_offset_mismatch(%in: memref<?xf32>) {
// expected-error @+1 {{expected result type with stride = 2 instead of 1 in dim = 0}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [2], sizes: [10], strides: [2]
: memref<?xf32> to memref<10xf32, offset: 2, strides: [1]>
return
@@ -101,7 +101,7 @@ func @memref_reinterpret_cast_offset_mismatch(%in: memref<?xf32>) {
%c0 = constant 0 : index
%c10 = constant 10 : index
// expected-error @+1 {{expected result type with size = 10 instead of -1 in dim = 0}}
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [%c0], sizes: [10, %c10], strides: [%c10, 1]
: memref<?xf32> to memref<?x?xf32, offset: ?, strides: [?, 1]>
return
@@ -112,7 +112,7 @@ func @memref_reinterpret_cast_offset_mismatch(%in: memref<?xf32>) {
func @memref_reshape_element_type_mismatch(
%buf: memref<*xf32>, %shape: memref<1xi32>) {
// expected-error @+1 {{element types of source and destination memref types should be the same}}
- memref_reshape %buf(%shape) : (memref<*xf32>, memref<1xi32>) -> memref<?xi32>
+ memref.reshape %buf(%shape) : (memref<*xf32>, memref<1xi32>) -> memref<?xi32>
}
// -----
@@ -120,7 +120,7 @@ func @memref_reshape_element_type_mismatch(
func @memref_reshape_dst_ranked_shape_unranked(
%buf: memref<*xf32>, %shape: memref<?xi32>) {
// expected-error @+1 {{cannot use shape operand with dynamic length to reshape to statically-ranked memref type}}
- memref_reshape %buf(%shape) : (memref<*xf32>, memref<?xi32>) -> memref<?xf32>
+ memref.reshape %buf(%shape) : (memref<*xf32>, memref<?xi32>) -> memref<?xf32>
}
// -----
@@ -128,7 +128,7 @@ func @memref_reshape_dst_ranked_shape_unranked(
func @memref_reshape_dst_shape_rank_mismatch(
%buf: memref<*xf32>, %shape: memref<1xi32>) {
// expected-error @+1 {{length of shape operand
diff ers from the result's memref rank}}
- memref_reshape %buf(%shape)
+ memref.reshape %buf(%shape)
: (memref<*xf32>, memref<1xi32>) -> memref<?x?xf32>
}
@@ -138,7 +138,7 @@ func @memref_reshape_src_affine_map_is_not_identity(
%buf: memref<4x4xf32, offset: 0, strides: [3, 2]>,
%shape: memref<1xi32>) {
// expected-error @+1 {{source memref type should have identity affine map}}
- memref_reshape %buf(%shape)
+ memref.reshape %buf(%shape)
: (memref<4x4xf32, offset: 0, strides: [3, 2]>, memref<1xi32>)
-> memref<8xf32>
}
@@ -148,67 +148,67 @@ func @memref_reshape_src_affine_map_is_not_identity(
func @memref_reshape_result_affine_map_is_not_identity(
%buf: memref<4x4xf32>, %shape: memref<1xi32>) {
// expected-error @+1 {{result memref type should have identity affine map}}
- memref_reshape %buf(%shape)
+ memref.reshape %buf(%shape)
: (memref<4x4xf32>, memref<1xi32>) -> memref<8xf32, offset: 0, strides: [2]>
}
// -----
// expected-error @+1 {{type should be static shaped memref}}
-global_memref @foo : i32
+memref.global @foo : i32
// -----
// expected-error @+1 {{type should be static shaped memref}}
-global_memref @foo : i32 = 5
+memref.global @foo : i32 = 5
// -----
// expected-error @+1 {{type should be static shaped memref}}
-global_memref @foo : memref<*xf32>
+memref.global @foo : memref<*xf32>
// -----
// expected-error @+1 {{type should be static shaped memref}}
-global_memref @foo : memref<?x?xf32>
+memref.global @foo : memref<?x?xf32>
// -----
// expected-error @+1 {{initial value should be a unit or elements attribute}}
-global_memref @foo : memref<2x2xf32> = "foo"
+memref.global @foo : memref<2x2xf32> = "foo"
// -----
// expected-error @+1 {{inferred shape of elements literal ([2]) does not match type ([2, 2])}}
-global_memref @foo : memref<2x2xf32> = dense<[0.0, 1.0]>
+memref.global @foo : memref<2x2xf32> = dense<[0.0, 1.0]>
// -----
// expected-error @+1 {{expected valid '@'-identifier for symbol name}}
-global_memref "private" "public" @foo : memref<2x2xf32> = "foo"
+memref.global "private" "public" @foo : memref<2x2xf32> = "foo"
// -----
// expected-error @+1 {{expected valid '@'-identifier for symbol name}}
-global_memref constant external @foo : memref<2x2xf32> = "foo"
+memref.global constant external @foo : memref<2x2xf32> = "foo"
// -----
// constant qualifier must be after visibility.
// expected-error @+1 {{expected valid '@'-identifier for symbol name}}
-global_memref constant "private" @foo : memref<2x2xf32> = "foo"
+memref.global constant "private" @foo : memref<2x2xf32> = "foo"
// -----
// expected-error @+1 {{op visibility expected to be one of ["public", "private", "nested"], but got "priate"}}
-global_memref "priate" constant @memref5 : memref<2xf32> = uninitialized
+memref.global "priate" constant @memref5 : memref<2xf32> = uninitialized
// -----
func @nonexistent_global_memref() {
// expected-error @+1 {{'gv' does not reference a valid global memref}}
- %0 = get_global_memref @gv : memref<3xf32>
+ %0 = memref.get_global @gv : memref<3xf32>
return
}
@@ -218,17 +218,17 @@ func @foo()
func @nonexistent_global_memref() {
// expected-error @+1 {{'foo' does not reference a valid global memref}}
- %0 = get_global_memref @foo : memref<3xf32>
+ %0 = memref.get_global @foo : memref<3xf32>
return
}
// -----
-global_memref @gv : memref<3xi32>
+memref.global @gv : memref<3xi32>
func @mismatched_types() {
// expected-error @+1 {{result type 'memref<3xf32>' does not match type 'memref<3xi32>' of the global memref @gv}}
- %0 = get_global_memref @gv : memref<3xf32>
+ %0 = memref.get_global @gv : memref<3xf32>
return
}
diff --git a/mlir/test/Dialect/Standard/ops.mlir b/mlir/test/Dialect/Standard/ops.mlir
index 80fc9341e8dd..02ec47f96ac6 100644
--- a/mlir/test/Dialect/Standard/ops.mlir
+++ b/mlir/test/Dialect/Standard/ops.mlir
@@ -19,10 +19,10 @@ func @test_index_cast_tensor_reverse(%arg0 : tensor<i64>) -> tensor<index> {
return %0 : tensor<index>
}
-// CHECK-LABEL: test_tensor_to_memref
-func @test_tensor_to_memref(%arg0: tensor<?xi64>, %arg1: tensor<*xi64>) -> (memref<?xi64, affine_map<(d0) -> (d0 + 7)>>, memref<*xi64, 1>) {
- %0 = tensor_to_memref %arg0 : memref<?xi64, affine_map<(d0) -> (d0 + 7)>>
- %1 = tensor_to_memref %arg1 : memref<*xi64, 1>
+// CHECK-LABEL: test_buffer_cast
+func @test_buffer_cast(%arg0: tensor<?xi64>, %arg1: tensor<*xi64>) -> (memref<?xi64, affine_map<(d0) -> (d0 + 7)>>, memref<*xi64, 1>) {
+ %0 = memref.buffer_cast %arg0 : memref<?xi64, affine_map<(d0) -> (d0 + 7)>>
+ %1 = memref.buffer_cast %arg1 : memref<*xi64, 1>
return %0, %1 : memref<?xi64, affine_map<(d0) -> (d0 + 7)>>, memref<*xi64, 1>
}
@@ -49,7 +49,7 @@ func @memref_reinterpret_cast(%in: memref<?xf32>)
-> memref<10x?xf32, offset: ?, strides: [?, 1]> {
%c0 = constant 0 : index
%c10 = constant 10 : index
- %out = memref_reinterpret_cast %in to
+ %out = memref.reinterpret_cast %in to
offset: [%c0], sizes: [10, %c10], strides: [%c10, 1]
: memref<?xf32> to memref<10x?xf32, offset: ?, strides: [?, 1]>
return %out : memref<10x?xf32, offset: ?, strides: [?, 1]>
@@ -58,41 +58,41 @@ func @memref_reinterpret_cast(%in: memref<?xf32>)
// CHECK-LABEL: func @memref_reshape(
func @memref_reshape(%unranked: memref<*xf32>, %shape1: memref<1xi32>,
%shape2: memref<2xi32>, %shape3: memref<?xi32>) -> memref<*xf32> {
- %dyn_vec = memref_reshape %unranked(%shape1)
+ %dyn_vec = memref.reshape %unranked(%shape1)
: (memref<*xf32>, memref<1xi32>) -> memref<?xf32>
- %dyn_mat = memref_reshape %dyn_vec(%shape2)
+ %dyn_mat = memref.reshape %dyn_vec(%shape2)
: (memref<?xf32>, memref<2xi32>) -> memref<?x?xf32>
- %new_unranked = memref_reshape %dyn_mat(%shape3)
+ %new_unranked = memref.reshape %dyn_mat(%shape3)
: (memref<?x?xf32>, memref<?xi32>) -> memref<*xf32>
return %new_unranked : memref<*xf32>
}
-// CHECK-LABEL: global_memref @memref0 : memref<2xf32>
-global_memref @memref0 : memref<2xf32>
+// CHECK-LABEL: memref.global @memref0 : memref<2xf32>
+memref.global @memref0 : memref<2xf32>
-// CHECK-LABEL: global_memref constant @memref1 : memref<2xf32> = dense<[0.000000e+00, 1.000000e+00]>
-global_memref constant @memref1 : memref<2xf32> = dense<[0.0, 1.0]>
+// CHECK-LABEL: memref.global constant @memref1 : memref<2xf32> = dense<[0.000000e+00, 1.000000e+00]>
+memref.global constant @memref1 : memref<2xf32> = dense<[0.0, 1.0]>
-// CHECK-LABEL: global_memref @memref2 : memref<2xf32> = uninitialized
-global_memref @memref2 : memref<2xf32> = uninitialized
+// CHECK-LABEL: memref.global @memref2 : memref<2xf32> = uninitialized
+memref.global @memref2 : memref<2xf32> = uninitialized
-// CHECK-LABEL: global_memref "private" @memref3 : memref<2xf32> = uninitialized
-global_memref "private" @memref3 : memref<2xf32> = uninitialized
+// CHECK-LABEL: memref.global "private" @memref3 : memref<2xf32> = uninitialized
+memref.global "private" @memref3 : memref<2xf32> = uninitialized
-// CHECK-LABEL: global_memref "private" constant @memref4 : memref<2xf32> = uninitialized
-global_memref "private" constant @memref4 : memref<2xf32> = uninitialized
+// CHECK-LABEL: memref.global "private" constant @memref4 : memref<2xf32> = uninitialized
+memref.global "private" constant @memref4 : memref<2xf32> = uninitialized
// CHECK-LABEL: func @write_global_memref
func @write_global_memref() {
- %0 = get_global_memref @memref0 : memref<2xf32>
+ %0 = memref.get_global @memref0 : memref<2xf32>
%1 = constant dense<[1.0, 2.0]> : tensor<2xf32>
- tensor_store %1, %0 : memref<2xf32>
+ memref.tensor_store %1, %0 : memref<2xf32>
return
}
// CHECK-LABEL: func @read_global_memref
func @read_global_memref() {
- %0 = get_global_memref @memref0 : memref<2xf32>
- %1 = tensor_load %0 : memref<2xf32>
+ %0 = memref.get_global @memref0 : memref<2xf32>
+ %1 = memref.tensor_load %0 : memref<2xf32>
return
}
diff --git a/mlir/test/Dialect/Standard/tensor-constant-bufferize.mlir b/mlir/test/Dialect/Standard/tensor-constant-bufferize.mlir
index fe897da670da..16a7d7d6d002 100644
--- a/mlir/test/Dialect/Standard/tensor-constant-bufferize.mlir
+++ b/mlir/test/Dialect/Standard/tensor-constant-bufferize.mlir
@@ -1,13 +1,13 @@
-// RUN: mlir-opt %s -tensor-constant-bufferize -split-input-file
+// RUN: mlir-opt %s -tensor-constant-bufferize -split-input-file | FileCheck %s
// CHECK-LABEL: module {
// We check the debug name too since we put some effort into making that readable.
// The name isn't load-bearing though.
-// CHECK: global_memref "private" constant @__constant_3x4xf32 : memref<3x4xf32> = dense<7.000000e+00>
+// CHECK: memref.global "private" constant @__constant_3x4xf32 : memref<3x4xf32> = dense<7.000000e+00>
// CHECK: @basic
func @basic() -> tensor<3x4xf32> {
- // CHECK: %[[MEMREF:.*]] = get_global_memref @__constant_3x4xf32 : memref<3x4xf32>
- // CHECK: %[[TENSOR:.*]] = tensor_load %[[MEMREF]]
+ // CHECK: %[[MEMREF:.*]] = memref.get_global @__constant_3x4xf32 : memref<3x4xf32>
+ // CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[MEMREF]]
%0 = constant dense<7.0> : tensor<3x4xf32>
// CHECK: return %[[TENSOR]]
return %0 : tensor<3x4xf32>
@@ -20,8 +20,8 @@ func @basic() -> tensor<3x4xf32> {
// CHECK-LABEL: module {
// Only one global is created.
-// CHECK: global_memref
-// CHECK-NOT: global_memref
+// CHECK: memref.global
+// CHECK-NOT: memref.global
func @duplicate_constants() -> (tensor<3x4xf32>, tensor<3x4xf32>) {
%0 = constant dense<7.0> : tensor<3x4xf32>
%1 = constant dense<7.0> : tensor<3x4xf32>
@@ -35,9 +35,9 @@ func @duplicate_constants() -> (tensor<3x4xf32>, tensor<3x4xf32>) {
// CHECK-LABEL: module {
// Two globals are created.
-// CHECK: global_memref
-// CHECK: global_memref
-// CHECK-NOT: global_memref
+// CHECK: memref.global
+// CHECK: memref.global
+// CHECK-NOT: memref.global
func @multiple_constants() -> (tensor<3x4xf32>, tensor<3x4xf32>) {
%0 = constant dense<7.0> : tensor<3x4xf32>
%1 = constant dense<8.0> : tensor<3x4xf32>
@@ -50,7 +50,7 @@ func @multiple_constants() -> (tensor<3x4xf32>, tensor<3x4xf32>) {
// CHECK-LABEL: module {
// We don't convert non-tensor globals.
-// CHECK-NOT: global_memref
+// CHECK-NOT: memref.global
func @non_tensor() {
%0 = constant 7 : i32
return
diff --git a/mlir/test/Dialect/Tensor/bufferize.mlir b/mlir/test/Dialect/Tensor/bufferize.mlir
index abc7d2af5676..36da47938309 100644
--- a/mlir/test/Dialect/Tensor/bufferize.mlir
+++ b/mlir/test/Dialect/Tensor/bufferize.mlir
@@ -2,9 +2,9 @@
// CHECK-LABEL: func @tensor.cast(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<?xindex>) -> tensor<2xindex> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]]
-// CHECK: %[[CASTED:.*]] = memref_cast %[[MEMREF]] : memref<?xindex> to memref<2xindex>
-// CHECK: %[[RET:.*]] = tensor_load %[[CASTED]]
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]]
+// CHECK: %[[CASTED:.*]] = memref.cast %[[MEMREF]] : memref<?xindex> to memref<2xindex>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[CASTED]]
// CHECK: return %[[RET]] : tensor<2xindex>
func @tensor.cast(%arg0: tensor<?xindex>) -> tensor<2xindex> {
%0 = tensor.cast %arg0 : tensor<?xindex> to tensor<2xindex>
@@ -13,9 +13,9 @@ func @tensor.cast(%arg0: tensor<?xindex>) -> tensor<2xindex> {
// CHECK-LABEL: func @tensor.cast_from_unranked(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<*xf32>) -> tensor<2xf32> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<*xf32>
-// CHECK: %[[CASTED_MEMREF:.*]] = memref_cast %[[MEMREF]] : memref<*xf32> to memref<2xf32>
-// CHECK: %[[RET:.*]] = tensor_load %[[CASTED_MEMREF]] : memref<2xf32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<*xf32>
+// CHECK: %[[CASTED_MEMREF:.*]] = memref.cast %[[MEMREF]] : memref<*xf32> to memref<2xf32>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[CASTED_MEMREF]] : memref<2xf32>
// CHECK: return %[[RET]] : tensor<2xf32>
func @tensor.cast_from_unranked(%arg0: tensor<*xf32>) -> tensor<2xf32> {
%0 = tensor.cast %arg0 : tensor<*xf32> to tensor<2xf32>
@@ -24,9 +24,9 @@ func @tensor.cast_from_unranked(%arg0: tensor<*xf32>) -> tensor<2xf32> {
// CHECK-LABEL: func @tensor.cast_to_unranked(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<2xf32>) -> tensor<*xf32> {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<2xf32>
-// CHECK: %[[CASTED_MEMREF:.*]] = memref_cast %[[MEMREF]] : memref<2xf32> to memref<*xf32>
-// CHECK: %[[RET:.*]] = tensor_load %[[CASTED_MEMREF]] : memref<*xf32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<2xf32>
+// CHECK: %[[CASTED_MEMREF:.*]] = memref.cast %[[MEMREF]] : memref<2xf32> to memref<*xf32>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[CASTED_MEMREF]] : memref<*xf32>
// CHECK: return %[[RET]] : tensor<*xf32>
func @tensor.cast_to_unranked(%arg0: tensor<2xf32>) -> tensor<*xf32> {
%0 = tensor.cast %arg0 : tensor<2xf32> to tensor<*xf32>
@@ -36,8 +36,8 @@ func @tensor.cast_to_unranked(%arg0: tensor<2xf32>) -> tensor<*xf32> {
// CHECK-LABEL: func @tensor.extract(
// CHECK-SAME: %[[TENSOR:.*]]: tensor<?xf32>,
// CHECK-SAME: %[[IDX:.*]]: index) -> f32 {
-// CHECK: %[[MEMREF:.*]] = tensor_to_memref %[[TENSOR]] : memref<?xf32>
-// CHECK: %[[RET:.*]] = load %[[MEMREF]][%[[IDX]]] : memref<?xf32>
+// CHECK: %[[MEMREF:.*]] = memref.buffer_cast %[[TENSOR]] : memref<?xf32>
+// CHECK: %[[RET:.*]] = memref.load %[[MEMREF]][%[[IDX]]] : memref<?xf32>
// CHECK: return %[[RET]] : f32
// CHECK: }
func @tensor.extract(%arg0: tensor<?xf32>, %arg1: index) -> f32 {
@@ -48,12 +48,12 @@ func @tensor.extract(%arg0: tensor<?xf32>, %arg1: index) -> f32 {
// CHECK-LABEL: func @tensor.from_elements(
// CHECK-SAME: %[[ELEM0:.*]]: index,
// CHECK-SAME: %[[ELEM1:.*]]: index) -> tensor<2xindex> {
-// CHECK: %[[MEMREF:.*]] = alloc()
+// CHECK: %[[MEMREF:.*]] = memref.alloc()
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: store %[[ELEM0]], %[[MEMREF]][%[[C0]]]
// CHECK: %[[C1:.*]] = constant 1 : index
// CHECK: store %[[ELEM1]], %[[MEMREF]][%[[C1]]]
-// CHECK: %[[RET:.*]] = tensor_load %[[MEMREF]]
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[MEMREF]]
// CHECK: return %[[RET]] : tensor<2xindex>
func @tensor.from_elements(%arg0: index, %arg1: index) -> tensor<2xindex> {
%0 = tensor.from_elements %arg0, %arg1 : tensor<2xindex>
@@ -63,21 +63,21 @@ func @tensor.from_elements(%arg0: index, %arg1: index) -> tensor<2xindex> {
// CHECK-LABEL: func @tensor.generate(
// CHECK-SAME: %[[ARG:.*]]: tensor<*xf32>,
// CHECK-SAME: %[[DYNAMIC_EXTENT:.*]]: index) -> tensor<?xindex> {
-// CHECK: %[[MEMREF:.*]] = alloc(%[[DYNAMIC_EXTENT]]) : memref<?xindex>
+// CHECK: %[[MEMREF:.*]] = memref.alloc(%[[DYNAMIC_EXTENT]]) : memref<?xindex>
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C1:.*]] = constant 1 : index
// CHECK: scf.parallel (%[[I:.*]]) = (%[[C0]]) to (%[[DYNAMIC_EXTENT]]) step (%[[C1]]) {
-// CHECK: %[[ELEM:.*]] = dim %[[ARG]], %[[I]] : tensor<*xf32>
+// CHECK: %[[ELEM:.*]] = memref.dim %[[ARG]], %[[I]] : tensor<*xf32>
// CHECK: store %[[ELEM]], %[[MEMREF]][%[[I]]] : memref<?xindex>
// CHECK: scf.yield
// CHECK: }
-// CHECK: %[[RET:.*]] = tensor_load %[[MEMREF]] : memref<?xindex>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[MEMREF]] : memref<?xindex>
// CHECK: return %[[RET]] : tensor<?xindex>
// CHECK: }
func @tensor.generate(%arg: tensor<*xf32>, %dynamic_extent: index) -> tensor<?xindex> {
%result = tensor.generate %dynamic_extent {
^bb0(%i : index):
- %elem = dim %arg, %i : tensor<*xf32>
+ %elem = memref.dim %arg, %i : tensor<*xf32>
tensor.yield %elem : index
} : tensor<?xindex>
return %result : tensor<?xindex>
@@ -88,7 +88,7 @@ func @tensor.generate(%arg: tensor<*xf32>, %dynamic_extent: index) -> tensor<?xi
//
// CHECK-LABEL: func @tensor.generate_static_and_dynamic(
// CHECK-SAME: %[[DYNAMIC_EXTENT:.*]]: index) -> tensor<16x?xindex> {
-// CHECK: %[[MEMREF:.*]] = alloc(%[[DYNAMIC_EXTENT]]) : memref<16x?xindex>
+// CHECK: %[[MEMREF:.*]] = memref.alloc(%[[DYNAMIC_EXTENT]]) : memref<16x?xindex>
// CHECK: %[[C0:.*]] = constant 0 : index
// CHECK: %[[C1:.*]] = constant 1 : index
// CHECK: %[[C16:.*]] = constant 16 : index
@@ -97,7 +97,7 @@ func @tensor.generate(%arg: tensor<*xf32>, %dynamic_extent: index) -> tensor<?xi
// CHECK: store %[[VAL_7]], %[[MEMREF]][%[[I]], %[[J]]] : memref<16x?xindex>
// CHECK: scf.yield
// CHECK: }
-// CHECK: %[[RET:.*]] = tensor_load %[[MEMREF]] : memref<16x?xindex>
+// CHECK: %[[RET:.*]] = memref.tensor_load %[[MEMREF]] : memref<16x?xindex>
// CHECK: return %[[RET]] : tensor<16x?xindex>
// CHECK: }
func @tensor.generate_static_and_dynamic(%arg0: index) -> tensor<16x?xindex> {
diff --git a/mlir/test/Dialect/Tensor/canonicalize.mlir b/mlir/test/Dialect/Tensor/canonicalize.mlir
index f975a51280e5..c274b6f8b1c9 100644
--- a/mlir/test/Dialect/Tensor/canonicalize.mlir
+++ b/mlir/test/Dialect/Tensor/canonicalize.mlir
@@ -171,10 +171,10 @@ func @extract_oob_from_tensor.from_elements(%element : index) -> index {
// CHECK-SAME: %[[IDX:.*]]: index, %[[TENSOR:.*]]: tensor<*xf32>
func @extract_from_tensor.generate(%idx: index, %tensor: tensor<*xf32>) -> index {
%size = rank %tensor : tensor<*xf32>
- // CHECK-NEXT: %[[RES:.*]] = dim %[[TENSOR]], %[[IDX]]
+ // CHECK-NEXT: %[[RES:.*]] = memref.dim %[[TENSOR]], %[[IDX]]
%0 = tensor.generate %size {
^bb0(%arg0: index):
- %1 = dim %tensor, %arg0 : tensor<*xf32>
+ %1 = memref.dim %tensor, %arg0 : tensor<*xf32>
tensor.yield %1 : index
} : tensor<?xindex>
%1 = tensor.extract %0[%idx] : tensor<?xindex>
@@ -188,13 +188,13 @@ func @extract_from_tensor.generate(%idx: index, %tensor: tensor<*xf32>) -> index
// CHECK-SAME: %[[IDX0:.*]]: index, %[[IDX1:.*]]: index, %[[TENSOR:.*]]: tensor<*xf32>
func @extract_from_tensor.generate_2d(%idx0: index, %idx1: index, %tensor: tensor<*xf32>) -> index {
%size = rank %tensor : tensor<*xf32>
- // CHECK-NEXT: %[[DIM0:.*]] = dim %[[TENSOR]], %[[IDX0]]
- // CHECK-NEXT: %[[DIM1:.*]] = dim %[[TENSOR]], %[[IDX1]]
+ // CHECK-NEXT: %[[DIM0:.*]] = memref.dim %[[TENSOR]], %[[IDX0]]
+ // CHECK-NEXT: %[[DIM1:.*]] = memref.dim %[[TENSOR]], %[[IDX1]]
// CHECK-NEXT: %[[RES:.*]] = addi %[[DIM0]], %[[DIM1]]
%0 = tensor.generate %size, %size {
^bb0(%arg0: index, %arg1: index):
- %1 = dim %tensor, %arg0 : tensor<*xf32>
- %2 = dim %tensor, %arg1 : tensor<*xf32>
+ %1 = memref.dim %tensor, %arg0 : tensor<*xf32>
+ %2 = memref.dim %tensor, %arg1 : tensor<*xf32>
%3 = addi %1, %2 : index
tensor.yield %3 : index
} : tensor<?x?xindex>
@@ -209,12 +209,12 @@ func @extract_from_tensor.generate_2d(%idx0: index, %idx1: index, %tensor: tenso
// CHECK-SAME: %[[IDX:.*]]: index
func @extract_from_tensor.generate_sideeffects(%idx: index, %tensor: tensor<*xf32>) -> index {
%size = rank %tensor : tensor<*xf32>
- %mem = alloc(%size) : memref<?xindex>
+ %mem = memref.alloc(%size) : memref<?xindex>
// CHECK: %[[DTENSOR:.*]] = tensor.generate
%0 = tensor.generate %size {
^bb0(%arg0: index):
- %1 = dim %tensor, %arg0 : tensor<*xf32>
- store %1, %mem[%arg0] : memref<?xindex>
+ %1 = memref.dim %tensor, %arg0 : tensor<*xf32>
+ memref.store %1, %mem[%arg0] : memref<?xindex>
tensor.yield %1 : index
} : tensor<?xindex>
// CHECK: %[[RES:.*]] = tensor.extract %[[DTENSOR]][%[[IDX]]]
diff --git a/mlir/test/Dialect/Vector/canonicalize.mlir b/mlir/test/Dialect/Vector/canonicalize.mlir
index 17da655f72cb..a68c7fba8e1c 100644
--- a/mlir/test/Dialect/Vector/canonicalize.mlir
+++ b/mlir/test/Dialect/Vector/canonicalize.mlir
@@ -255,7 +255,7 @@ func @transpose_3D_sequence(%arg : vector<4x3x2xf32>) -> vector<4x3x2xf32> {
func @cast_transfers(%A: memref<4x8xf32>) -> (vector<4x8xf32>) {
%c0 = constant 0 : index
%f0 = constant 0.0 : f32
- %0 = memref_cast %A : memref<4x8xf32> to memref<?x?xf32>
+ %0 = memref.cast %A : memref<4x8xf32> to memref<?x?xf32>
// CHECK: vector.transfer_read %{{.*}} {masked = [false, false]} : memref<4x8xf32>, vector<4x8xf32>
%1 = vector.transfer_read %0[%c0, %c0], %f0 : memref<?x?xf32>, vector<4x8xf32>
diff --git a/mlir/test/Dialect/Vector/vector-contract-matvec-transforms.mlir b/mlir/test/Dialect/Vector/vector-contract-matvec-transforms.mlir
index c02de28b427e..6d84f64d37d8 100644
--- a/mlir/test/Dialect/Vector/vector-contract-matvec-transforms.mlir
+++ b/mlir/test/Dialect/Vector/vector-contract-matvec-transforms.mlir
@@ -49,9 +49,9 @@
// CHECK-SAME: %[[A:.*0]]: memref<vector<2x2xf32>>
// CHECK-SAME: %[[B:.*1]]: memref<vector<2xf32>>
// CHECK-SAME: %[[C:.*2]]: memref<vector<2xf32>>
-// CHECK: %[[T0:.*]] = load %[[A]][] : memref<vector<2x2xf32>>
-// CHECK: %[[T1:.*]] = load %[[B]][] : memref<vector<2xf32>>
-// CHECK: %[[T2:.*]] = load %[[C]][] : memref<vector<2xf32>>
+// CHECK: %[[T0:.*]] = memref.load %[[A]][] : memref<vector<2x2xf32>>
+// CHECK: %[[T1:.*]] = memref.load %[[B]][] : memref<vector<2xf32>>
+// CHECK: %[[T2:.*]] = memref.load %[[C]][] : memref<vector<2xf32>>
// CHECK: %[[T3:.*]] = vector.transpose %[[T0]], [1, 0] : vector<2x2xf32> to vector<2x2xf32>
// CHECK: %[[T4:.*]] = vector.extract %[[T3]][0] : vector<2x2xf32>
// CHECK: %[[T5:.*]] = vector.extract %[[T1]][0] : vector<2xf32>
@@ -59,15 +59,15 @@
// CHECK: %[[T7:.*]] = vector.extract %[[T3]][1] : vector<2x2xf32>
// CHECK: %[[T8:.*]] = vector.extract %[[T1]][1] : vector<2xf32>
// CHECK: %[[T9:.*]] = vector.outerproduct %[[T7]], %[[T8]], %[[T6]] {kind = #vector.kind<add>} : vector<2xf32>, f32
-// CHECK: store %[[T9]], %[[C]][] : memref<vector<2xf32>>
+// CHECK: memref.store %[[T9]], %[[C]][] : memref<vector<2xf32>>
// CHECK: return
func @matvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
%arg2: memref<vector<2xf32>>) {
- %A = load %arg0[] : memref<vector<2x2xf32>>
- %x = load %arg1[] : memref<vector<2xf32>>
- %b = load %arg2[] : memref<vector<2xf32>>
+ %A = memref.load %arg0[] : memref<vector<2x2xf32>>
+ %x = memref.load %arg1[] : memref<vector<2xf32>>
+ %b = memref.load %arg2[] : memref<vector<2xf32>>
%0 = vector.contract #matvec_trait %A, %x, %b : vector<2x2xf32>, vector<2xf32> into vector<2xf32>
- store %0, %arg2[] : memref<vector<2xf32>>
+ memref.store %0, %arg2[] : memref<vector<2xf32>>
return
}
@@ -75,9 +75,9 @@ func @matvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
// CHECK-SAME: %[[A:.*0]]: memref<vector<2x2xf32>>
// CHECK-SAME: %[[B:.*1]]: memref<vector<2xf32>>
// CHECK-SAME: %[[C:.*2]]: memref<vector<2xf32>>
-// CHECK: %[[T0:.*]] = load %[[A]][] : memref<vector<2x2xf32>>
-// CHECK: %[[T1:.*]] = load %[[B]][] : memref<vector<2xf32>>
-// CHECK: %[[T2:.*]] = load %[[C]][] : memref<vector<2xf32>>
+// CHECK: %[[T0:.*]] = memref.load %[[A]][] : memref<vector<2x2xf32>>
+// CHECK: %[[T1:.*]] = memref.load %[[B]][] : memref<vector<2xf32>>
+// CHECK: %[[T2:.*]] = memref.load %[[C]][] : memref<vector<2xf32>>
// CHECK: %[[T3:.*]] = vector.transpose %[[T0]], [1, 0] : vector<2x2xf32> to vector<2x2xf32>
// CHECK: %[[T4:.*]] = vector.extract %[[T3]][0] : vector<2x2xf32>
// CHECK: %[[T5:.*]] = vector.extract %[[T1]][0] : vector<2xf32>
@@ -85,15 +85,15 @@ func @matvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
// CHECK: %[[T7:.*]] = vector.extract %[[T3]][1] : vector<2x2xf32>
// CHECK: %[[T8:.*]] = vector.extract %[[T1]][1] : vector<2xf32>
// CHECK: %[[T9:.*]] = vector.outerproduct %[[T7]], %[[T8]], %[[T6]] {kind = #vector.kind<max>} : vector<2xf32>, f32
-// CHECK: store %[[T9]], %[[C]][] : memref<vector<2xf32>>
+// CHECK: memref.store %[[T9]], %[[C]][] : memref<vector<2xf32>>
// CHECK: return
func @matvecmax2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
%arg2: memref<vector<2xf32>>) {
- %A = load %arg0[] : memref<vector<2x2xf32>>
- %x = load %arg1[] : memref<vector<2xf32>>
- %b = load %arg2[] : memref<vector<2xf32>>
+ %A = memref.load %arg0[] : memref<vector<2x2xf32>>
+ %x = memref.load %arg1[] : memref<vector<2xf32>>
+ %b = memref.load %arg2[] : memref<vector<2xf32>>
%0 = vector.contract #matvecmax_trait %A, %x, %b : vector<2x2xf32>, vector<2xf32> into vector<2xf32>
- store %0, %arg2[] : memref<vector<2xf32>>
+ memref.store %0, %arg2[] : memref<vector<2xf32>>
return
}
@@ -101,24 +101,24 @@ func @matvecmax2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
// CHECK-SAME: %[[A:.*0]]: memref<vector<2x2xf32>>
// CHECK-SAME: %[[B:.*1]]: memref<vector<2xf32>>
// CHECK-SAME: %[[C:.*2]]: memref<vector<2xf32>>
-// CHECK: %[[T0:.*]] = load %[[A]][] : memref<vector<2x2xf32>>
-// CHECK: %[[T1:.*]] = load %[[B]][] : memref<vector<2xf32>>
-// CHECK: %[[T2:.*]] = load %[[C]][] : memref<vector<2xf32>>
+// CHECK: %[[T0:.*]] = memref.load %[[A]][] : memref<vector<2x2xf32>>
+// CHECK: %[[T1:.*]] = memref.load %[[B]][] : memref<vector<2xf32>>
+// CHECK: %[[T2:.*]] = memref.load %[[C]][] : memref<vector<2xf32>>
// CHECK: %[[T3:.*]] = vector.extract %[[T0]][0] : vector<2x2xf32>
// CHECK: %[[T4:.*]] = vector.extract %[[T1]][0] : vector<2xf32>
// CHECK: %[[T5:.*]] = vector.outerproduct %[[T3]], %[[T4]], %[[T2]] {kind = #vector.kind<add>} : vector<2xf32>, f32
// CHECK: %[[T6:.*]] = vector.extract %[[T0]][1] : vector<2x2xf32>
// CHECK: %[[T7:.*]] = vector.extract %[[T1]][1] : vector<2xf32>
// CHECK: %[[T8:.*]] = vector.outerproduct %[[T6]], %[[T7]], %[[T5]] {kind = #vector.kind<add>} : vector<2xf32>, f32
-// CHECK: store %[[T8]], %[[C]][] : memref<vector<2xf32>>
+// CHECK: memref.store %[[T8]], %[[C]][] : memref<vector<2xf32>>
// CHECK: return
func @mattransvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
%arg2: memref<vector<2xf32>>) {
- %A = load %arg0[] : memref<vector<2x2xf32>>
- %x = load %arg1[] : memref<vector<2xf32>>
- %b = load %arg2[] : memref<vector<2xf32>>
+ %A = memref.load %arg0[] : memref<vector<2x2xf32>>
+ %x = memref.load %arg1[] : memref<vector<2xf32>>
+ %b = memref.load %arg2[] : memref<vector<2xf32>>
%0 = vector.contract #mattransvec_trait %A, %x, %b : vector<2x2xf32>, vector<2xf32> into vector<2xf32>
- store %0, %arg2[] : memref<vector<2xf32>>
+ memref.store %0, %arg2[] : memref<vector<2xf32>>
return
}
@@ -126,9 +126,9 @@ func @mattransvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>
// CHECK-SAME: %[[A:.*0]]: memref<vector<2x2xf32>>
// CHECK-SAME: %[[B:.*1]]: memref<vector<2xf32>>
// CHECK-SAME: %[[C:.*2]]: memref<vector<2xf32>>
-// CHECK: %[[T0:.*]] = load %[[A]][] : memref<vector<2x2xf32>>
-// CHECK: %[[T1:.*]] = load %[[B]][] : memref<vector<2xf32>>
-// CHECK: %[[T2:.*]] = load %[[C]][] : memref<vector<2xf32>>
+// CHECK: %[[T0:.*]] = memref.load %[[A]][] : memref<vector<2x2xf32>>
+// CHECK: %[[T1:.*]] = memref.load %[[B]][] : memref<vector<2xf32>>
+// CHECK: %[[T2:.*]] = memref.load %[[C]][] : memref<vector<2xf32>>
// CHECK: %[[T3:.*]] = vector.transpose %[[T0]], [1, 0] : vector<2x2xf32> to vector<2x2xf32>
// CHECK: %[[T4:.*]] = vector.extract %[[T3]][0] : vector<2x2xf32>
// CHECK: %[[T5:.*]] = vector.extract %[[T1]][0] : vector<2xf32>
@@ -136,15 +136,15 @@ func @mattransvec2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>
// CHECK: %[[T7:.*]] = vector.extract %[[T3]][1] : vector<2x2xf32>
// CHECK: %[[T8:.*]] = vector.extract %[[T1]][1] : vector<2xf32>
// CHECK: %[[T9:.*]] = vector.outerproduct %[[T7]], %[[T8]], %[[T6]] {kind = #vector.kind<add>} : vector<2xf32>, f32
-// CHECK: store %[[T9]], %[[C]][] : memref<vector<2xf32>>
+// CHECK: memref.store %[[T9]], %[[C]][] : memref<vector<2xf32>>
// CHECK: return
func @vecmat2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
%arg2: memref<vector<2xf32>>) {
- %A = load %arg0[] : memref<vector<2x2xf32>>
- %x = load %arg1[] : memref<vector<2xf32>>
- %b = load %arg2[] : memref<vector<2xf32>>
+ %A = memref.load %arg0[] : memref<vector<2x2xf32>>
+ %x = memref.load %arg1[] : memref<vector<2xf32>>
+ %b = memref.load %arg2[] : memref<vector<2xf32>>
%0 = vector.contract #vecmat_trait %x, %A, %b : vector<2xf32>, vector<2x2xf32> into vector<2xf32>
- store %0, %arg2[] : memref<vector<2xf32>>
+ memref.store %0, %arg2[] : memref<vector<2xf32>>
return
}
@@ -152,23 +152,23 @@ func @vecmat2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
// CHECK-SAME: %[[A:.*0]]: memref<vector<2x2xf32>>
// CHECK-SAME: %[[B:.*1]]: memref<vector<2xf32>>
// CHECK-SAME: %[[C:.*2]]: memref<vector<2xf32>>
-// CHECK: %[[T0:.*]] = load %[[A]][] : memref<vector<2x2xf32>>
-// CHECK: %[[T1:.*]] = load %[[B]][] : memref<vector<2xf32>>
-// CHECK: %[[T2:.*]] = load %[[C]][] : memref<vector<2xf32>>
+// CHECK: %[[T0:.*]] = memref.load %[[A]][] : memref<vector<2x2xf32>>
+// CHECK: %[[T1:.*]] = memref.load %[[B]][] : memref<vector<2xf32>>
+// CHECK: %[[T2:.*]] = memref.load %[[C]][] : memref<vector<2xf32>>
// CHECK: %[[T3:.*]] = vector.extract %[[T0]][0] : vector<2x2xf32>
// CHECK: %[[T4:.*]] = vector.extract %[[T1]][0] : vector<2xf32>
// CHECK: %[[T5:.*]] = vector.outerproduct %[[T3]], %[[T4]], %[[T2]] {kind = #vector.kind<add>} : vector<2xf32>, f32
// CHECK: %[[T6:.*]] = vector.extract %[[T0]][1] : vector<2x2xf32>
// CHECK: %[[T7:.*]] = vector.extract %[[T1]][1] : vector<2xf32>
// CHECK: %[[T8:.*]] = vector.outerproduct %[[T6]], %[[T7]], %[[T5]] {kind = #vector.kind<add>} : vector<2xf32>, f32
-// CHECK: store %[[T8]], %[[C]][] : memref<vector<2xf32>>
+// CHECK: memref.store %[[T8]], %[[C]][] : memref<vector<2xf32>>
// CHECK: return
func @vecmattrans2x2(%arg0: memref<vector<2x2xf32>>, %arg1: memref<vector<2xf32>>,
%arg2: memref<vector<2xf32>>) {
- %A = load %arg0[] : memref<vector<2x2xf32>>
- %x = load %arg1[] : memref<vector<2xf32>>
- %b = load %arg2[] : memref<vector<2xf32>>
+ %A = memref.load %arg0[] : memref<vector<2x2xf32>>
+ %x = memref.load %arg1[] : memref<vector<2xf32>>
+ %b = memref.load %arg2[] : memref<vector<2xf32>>
%0 = vector.contract #vecmattrans_trait %x, %A, %b : vector<2xf32>, vector<2x2xf32> into vector<2xf32>
- store %0, %arg2[] : memref<vector<2xf32>>
+ memref.store %0, %arg2[] : memref<vector<2xf32>>
return
}
diff --git a/mlir/test/Dialect/Vector/vector-transfer-full-partial-split.mlir b/mlir/test/Dialect/Vector/vector-transfer-full-partial-split.mlir
index 4757b8305c69..21d749cc088f 100644
--- a/mlir/test/Dialect/Vector/vector-transfer-full-partial-split.mlir
+++ b/mlir/test/Dialect/Vector/vector-transfer-full-partial-split.mlir
@@ -29,10 +29,10 @@ func @split_vector_transfer_read_2d(%A: memref<?x8xf32>, %i: index, %j: index) -
// CHECK-DAG: %[[c8:.*]] = constant 8 : index
// CHECK-DAG: %[[cst:.*]] = constant 0.000000e+00 : f32
// alloca for boundary full tile
- // CHECK: %[[alloc:.*]] = alloca() {alignment = 32 : i64} : memref<4x8xf32>
+ // CHECK: %[[alloc:.*]] = memref.alloca() {alignment = 32 : i64} : memref<4x8xf32>
// %i + 4 <= dim(%A, 0)
// CHECK: %[[idx0:.*]] = affine.apply #[[$map_p4]]()[%[[i]]]
- // CHECK: %[[d0:.*]] = dim %[[A]], %[[c0]] : memref<?x8xf32>
+ // CHECK: %[[d0:.*]] = memref.dim %[[A]], %[[c0]] : memref<?x8xf32>
// CHECK: %[[cmp0:.*]] = cmpi sle, %[[idx0]], %[[d0]] : index
// %j + 8 <= dim(%A, 1)
// CHECK: %[[idx1:.*]] = affine.apply #[[$map_p8]]()[%[[j]]]
@@ -49,7 +49,7 @@ func @split_vector_transfer_read_2d(%A: memref<?x8xf32>, %i: index, %j: index) -
// CHECK: %[[cast_alloc:.*]] = vector.type_cast %[[alloc]] :
// CHECK-SAME: memref<4x8xf32> to memref<vector<4x8xf32>>
// CHECK: store %[[slow]], %[[cast_alloc]][] : memref<vector<4x8xf32>>
- // CHECK: %[[yielded:.*]] = memref_cast %[[alloc]] :
+ // CHECK: %[[yielded:.*]] = memref.cast %[[alloc]] :
// CHECK-SAME: memref<4x8xf32> to memref<?x8xf32>
// CHECK: scf.yield %[[yielded]], %[[c0]], %[[c0]] :
// CHECK-SAME: memref<?x8xf32>, index, index
@@ -62,10 +62,10 @@ func @split_vector_transfer_read_2d(%A: memref<?x8xf32>, %i: index, %j: index) -
// LINALG-DAG: %[[c8:.*]] = constant 8 : index
// LINALG-DAG: %[[cst:.*]] = constant 0.000000e+00 : f32
// alloca for boundary full tile
- // LINALG: %[[alloc:.*]] = alloca() {alignment = 32 : i64} : memref<4x8xf32>
+ // LINALG: %[[alloc:.*]] = memref.alloca() {alignment = 32 : i64} : memref<4x8xf32>
// %i + 4 <= dim(%A, 0)
// LINALG: %[[idx0:.*]] = affine.apply #[[$map_p4]]()[%[[i]]]
- // LINALG: %[[d0:.*]] = dim %[[A]], %[[c0]] : memref<?x8xf32>
+ // LINALG: %[[d0:.*]] = memref.dim %[[A]], %[[c0]] : memref<?x8xf32>
// LINALG: %[[cmp0:.*]] = cmpi sle, %[[idx0]], %[[d0]] : index
// %j + 8 <= dim(%A, 1)
// LINALG: %[[idx1:.*]] = affine.apply #[[$map_p8]]()[%[[j]]]
@@ -78,13 +78,13 @@ func @split_vector_transfer_read_2d(%A: memref<?x8xf32>, %i: index, %j: index) -
// LINALG: } else {
// slow path, fill tmp alloc and yield a memref_casted version of it
// LINALG: linalg.fill(%[[alloc]], %[[cst]]) : memref<4x8xf32>, f32
- // LINALG: %[[d0:.*]] = dim %[[A]], %[[c0]] : memref<?x8xf32>
+ // LINALG: %[[d0:.*]] = memref.dim %[[A]], %[[c0]] : memref<?x8xf32>
// LINALG: %[[sv0:.*]] = affine.min #[[$bounds_map_4]](%[[d0]], %[[i]], %[[c4]])
// LINALG: %[[sv1:.*]] = affine.min #[[$bounds_map_8]](%[[c8]], %[[j]], %[[c8]])
- // LINALG: %[[sv:.*]] = subview %[[A]][%[[i]], %[[j]]] [%[[sv0]], %[[sv1]]] [1, 1]
+ // LINALG: %[[sv:.*]] = memref.subview %[[A]][%[[i]], %[[j]]] [%[[sv0]], %[[sv1]]] [1, 1]
// LINALG-SAME: memref<?x8xf32> to memref<?x?xf32, #[[$map_2d_stride_8x1]]>
// LINALG: linalg.copy(%[[sv]], %[[alloc]]) : memref<?x?xf32, #[[$map_2d_stride_8x1]]>, memref<4x8xf32>
- // LINALG: %[[yielded:.*]] = memref_cast %[[alloc]] :
+ // LINALG: %[[yielded:.*]] = memref.cast %[[alloc]] :
// LINALG-SAME: memref<4x8xf32> to memref<?x8xf32>
// LINALG: scf.yield %[[yielded]], %[[c0]], %[[c0]] :
// LINALG-SAME: memref<?x8xf32>, index, index
@@ -117,7 +117,7 @@ func @split_vector_transfer_read_strided_2d(
// CHECK-DAG: %[[c8:.*]] = constant 8 : index
// CHECK-DAG: %[[cst:.*]] = constant 0.000000e+00 : f32
// alloca for boundary full tile
- // CHECK: %[[alloc:.*]] = alloca() {alignment = 32 : i64} : memref<4x8xf32>
+ // CHECK: %[[alloc:.*]] = memref.alloca() {alignment = 32 : i64} : memref<4x8xf32>
// %i + 4 <= dim(%A, 0)
// CHECK: %[[idx0:.*]] = affine.apply #[[$map_p4]]()[%[[i]]]
// CHECK: %[[cmp0:.*]] = cmpi sle, %[[idx0]], %[[c7]] : index
@@ -128,7 +128,7 @@ func @split_vector_transfer_read_strided_2d(
// CHECK: %[[cond:.*]] = and %[[cmp0]], %[[cmp1]] : i1
// CHECK: %[[ifres:.*]]:3 = scf.if %[[cond]] -> (memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index) {
// inBounds but not cast-compatible: yield a memref_casted form of %A
- // CHECK: %[[casted:.*]] = memref_cast %arg0 :
+ // CHECK: %[[casted:.*]] = memref.cast %arg0 :
// CHECK-SAME: memref<7x8xf32, #[[$map_2d_stride_1]]> to memref<?x8xf32, #[[$map_2d_stride_1]]>
// CHECK: scf.yield %[[casted]], %[[i]], %[[j]] :
// CHECK-SAME: memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index
@@ -140,7 +140,7 @@ func @split_vector_transfer_read_strided_2d(
// CHECK-SAME: memref<4x8xf32> to memref<vector<4x8xf32>>
// CHECK: store %[[slow]], %[[cast_alloc]][] :
// CHECK-SAME: memref<vector<4x8xf32>>
- // CHECK: %[[yielded:.*]] = memref_cast %[[alloc]] :
+ // CHECK: %[[yielded:.*]] = memref.cast %[[alloc]] :
// CHECK-SAME: memref<4x8xf32> to memref<?x8xf32, #[[$map_2d_stride_1]]>
// CHECK: scf.yield %[[yielded]], %[[c0]], %[[c0]] :
// CHECK-SAME: memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index
@@ -154,7 +154,7 @@ func @split_vector_transfer_read_strided_2d(
// LINALG-DAG: %[[c8:.*]] = constant 8 : index
// LINALG-DAG: %[[cst:.*]] = constant 0.000000e+00 : f32
// alloca for boundary full tile
- // LINALG: %[[alloc:.*]] = alloca() {alignment = 32 : i64} : memref<4x8xf32>
+ // LINALG: %[[alloc:.*]] = memref.alloca() {alignment = 32 : i64} : memref<4x8xf32>
// %i + 4 <= dim(%A, 0)
// LINALG: %[[idx0:.*]] = affine.apply #[[$map_p4]]()[%[[i]]]
// LINALG: %[[cmp0:.*]] = cmpi sle, %[[idx0]], %[[c7]] : index
@@ -165,7 +165,7 @@ func @split_vector_transfer_read_strided_2d(
// LINALG: %[[cond:.*]] = and %[[cmp0]], %[[cmp1]] : i1
// LINALG: %[[ifres:.*]]:3 = scf.if %[[cond]] -> (memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index) {
// inBounds but not cast-compatible: yield a memref_casted form of %A
- // LINALG: %[[casted:.*]] = memref_cast %arg0 :
+ // LINALG: %[[casted:.*]] = memref.cast %arg0 :
// LINALG-SAME: memref<7x8xf32, #[[$map_2d_stride_1]]> to memref<?x8xf32, #[[$map_2d_stride_1]]>
// LINALG: scf.yield %[[casted]], %[[i]], %[[j]] :
// LINALG-SAME: memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index
@@ -174,10 +174,10 @@ func @split_vector_transfer_read_strided_2d(
// LINALG: linalg.fill(%[[alloc]], %[[cst]]) : memref<4x8xf32>, f32
// LINALG: %[[sv0:.*]] = affine.min #[[$bounds_map_4]](%[[c7]], %[[i]], %[[c4]])
// LINALG: %[[sv1:.*]] = affine.min #[[$bounds_map_8]](%[[c8]], %[[j]], %[[c8]])
- // LINALG: %[[sv:.*]] = subview %[[A]][%[[i]], %[[j]]] [%[[sv0]], %[[sv1]]] [1, 1]
+ // LINALG: %[[sv:.*]] = memref.subview %[[A]][%[[i]], %[[j]]] [%[[sv0]], %[[sv1]]] [1, 1]
// LINALG-SAME: memref<7x8xf32, #[[$map_2d_stride_1]]> to memref<?x?xf32, #[[$map_2d_stride_1]]>
// LINALG: linalg.copy(%[[sv]], %[[alloc]]) : memref<?x?xf32, #[[$map_2d_stride_1]]>, memref<4x8xf32>
- // LINALG: %[[yielded:.*]] = memref_cast %[[alloc]] :
+ // LINALG: %[[yielded:.*]] = memref.cast %[[alloc]] :
// LINALG-SAME: memref<4x8xf32> to memref<?x8xf32, #[[$map_2d_stride_1]]>
// LINALG: scf.yield %[[yielded]], %[[c0]], %[[c0]] :
// LINALG-SAME: memref<?x8xf32, #[[$map_2d_stride_1]]>, index, index
diff --git a/mlir/test/Dialect/Vector/vector-transfer-lowering.mlir b/mlir/test/Dialect/Vector/vector-transfer-lowering.mlir
index bc23821b856f..ff32f7d5c823 100644
--- a/mlir/test/Dialect/Vector/vector-transfer-lowering.mlir
+++ b/mlir/test/Dialect/Vector/vector-transfer-lowering.mlir
@@ -159,7 +159,7 @@ func @transfer_perm_map(%mem : memref<8x8xf32>, %i : index) -> vector<4xf32> {
// CHECK-LABEL: func @transfer_broadcasting(
// CHECK-SAME: %[[MEM:.*]]: memref<8x8xf32>,
// CHECK-SAME: %[[IDX:.*]]: index) -> vector<4xf32> {
-// CHECK-NEXT: %[[LOAD:.*]] = load %[[MEM]][%[[IDX]], %[[IDX]]] : memref<8x8xf32>
+// CHECK-NEXT: %[[LOAD:.*]] = memref.load %[[MEM]][%[[IDX]], %[[IDX]]] : memref<8x8xf32>
// CHECK-NEXT: %[[RES:.*]] = vector.broadcast %[[LOAD]] : f32 to vector<4xf32>
// CHECK-NEXT: return %[[RES]] : vector<4xf32>
// CHECK-NEXT: }
@@ -177,7 +177,7 @@ func @transfer_broadcasting(%mem : memref<8x8xf32>, %i : index) -> vector<4xf32>
// CHECK-LABEL: func @transfer_broadcasting_2D(
// CHECK-SAME: %[[MEM:.*]]: memref<8x8xf32>,
// CHECK-SAME: %[[IDX:.*]]: index) -> vector<4x4xf32> {
-// CHECK-NEXT: %[[LOAD:.*]] = load %[[MEM]][%[[IDX]], %[[IDX]]] : memref<8x8xf32>
+// CHECK-NEXT: %[[LOAD:.*]] = memref.load %[[MEM]][%[[IDX]], %[[IDX]]] : memref<8x8xf32>
// CHECK-NEXT: %[[RES:.*]] = vector.broadcast %[[LOAD]] : f32 to vector<4x4xf32>
// CHECK-NEXT: return %[[RES]] : vector<4x4xf32>
// CHECK-NEXT: }
diff --git a/mlir/test/Dialect/Vector/vector-transforms.mlir b/mlir/test/Dialect/Vector/vector-transforms.mlir
index f57e5f6e7ea9..da899389de74 100644
--- a/mlir/test/Dialect/Vector/vector-transforms.mlir
+++ b/mlir/test/Dialect/Vector/vector-transforms.mlir
@@ -287,9 +287,9 @@ func @contraction4x4_ikj_xfer_read(%arg0 : memref<4x2xf32>,
func @vector_transfers(%arg0: index, %arg1: index) {
%cst = constant 0.000000e+00 : f32
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
- %1 = alloc(%arg0, %arg1) : memref<?x?xf32>
- %2 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %2 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
%cst_0 = constant 1.000000e+00 : f32
%cst_1 = constant 2.000000e+00 : f32
affine.for %arg2 = 0 to %arg0 step 4 {
@@ -434,7 +434,7 @@ func @vector_transfers_vector_element_type() {
%cf0 = constant 0.000000e+00 : f32
%vf0 = splat %cf0 : vector<2x4xf32>
- %0 = alloc() : memref<6x2x1xvector<2x4xf32>>
+ %0 = memref.alloc() : memref<6x2x1xvector<2x4xf32>>
%1 = vector.transfer_read %0[%c0, %c0, %c0], %vf0
{permutation_map = affine_map<(d0, d1, d2) -> (d1, d2)>}
diff --git a/mlir/test/EDSC/CMakeLists.txt b/mlir/test/EDSC/CMakeLists.txt
index 5a73c250c774..d4380561854b 100644
--- a/mlir/test/EDSC/CMakeLists.txt
+++ b/mlir/test/EDSC/CMakeLists.txt
@@ -16,6 +16,7 @@ target_link_libraries(mlir-edsc-builder-api-test
MLIRIR
MLIRLinalg
MLIRLinalgEDSC
+ MLIRMemRef
MLIRSCF
MLIRStandard
MLIRTransforms
diff --git a/mlir/test/EDSC/builder-api-test.cpp b/mlir/test/EDSC/builder-api-test.cpp
index 352e709ebcce..5cb04e6aba52 100644
--- a/mlir/test/EDSC/builder-api-test.cpp
+++ b/mlir/test/EDSC/builder-api-test.cpp
@@ -12,6 +12,7 @@
#include "mlir/Dialect/Linalg/EDSC/Builders.h"
#include "mlir/Dialect/Linalg/EDSC/Intrinsics.h"
#include "mlir/Dialect/Math/IR/Math.h"
+#include "mlir/Dialect/MemRef/EDSC/Intrinsics.h"
#include "mlir/Dialect/SCF/EDSC/Intrinsics.h"
#include "mlir/Dialect/StandardOps/EDSC/Intrinsics.h"
#include "mlir/Dialect/Vector/EDSC/Intrinsics.h"
@@ -44,6 +45,7 @@ static MLIRContext &globalContext() {
scf::SCFDialect,
linalg::LinalgDialect,
math::MathDialect,
+ memref::MemRefDialect,
StandardOpsDialect,
vector::VectorDialect>();
// clang-format on
@@ -762,9 +764,9 @@ TEST_FUNC(tile_2d) {
// clang-format off
// CHECK-LABEL: func @tile_2d
// CHECK: %[[ZERO:.*]] = constant 0 : index
- // CHECK: %[[M:[0-9]+]] = dim %arg2, %c0{{[_0-9]*}} : memref<?x?x?xf32>
- // CHECK: %[[N:[0-9]+]] = dim %arg2, %c1{{[_0-9]*}} : memref<?x?x?xf32>
- // CHECK: %[[P:[0-9]+]] = dim %arg2, %c2{{[_0-9]*}} : memref<?x?x?xf32>
+ // CHECK: %[[M:[0-9]+]] = memref.dim %arg2, %c0{{[_0-9]*}} : memref<?x?x?xf32>
+ // CHECK: %[[N:[0-9]+]] = memref.dim %arg2, %c1{{[_0-9]*}} : memref<?x?x?xf32>
+ // CHECK: %[[P:[0-9]+]] = memref.dim %arg2, %c2{{[_0-9]*}} : memref<?x?x?xf32>
// CHECK: affine.for %{{.*}} = affine_map<(d0) -> (d0)>(%[[ZERO]]) to affine_map<(d0) -> (d0)>(%[[M]]) step 512 {
// CHECK-NEXT: affine.for %{{.*}} = affine_map<(d0) -> (d0)>(%[[ZERO]]) to affine_map<(d0) -> (d0)>(%[[N]]) step 1024 {
// CHECK-NEXT: affine.for %{{.*}} = affine_map<(d0) -> (d0)>(%[[ZERO]]) to affine_map<(d0) -> (d0)>(%[[P]]) {
@@ -806,7 +808,7 @@ TEST_FUNC(indirect_access) {
Value zero = std_constant_index(0);
MemRefBoundsCapture vC(f.getArgument(2));
AffineIndexedValue B(f.getArgument(1)), D(f.getArgument(3));
- StdIndexedValue A(f.getArgument(0)), C(f.getArgument(2));
+ MemRefIndexedValue A(f.getArgument(0)), C(f.getArgument(2));
Value N(vC.ub(0));
// clang-format off
diff --git a/mlir/test/Examples/Toy/Ch5/affine-lowering.mlir b/mlir/test/Examples/Toy/Ch5/affine-lowering.mlir
index 62fcc880e6d6..a51d49770501 100644
--- a/mlir/test/Examples/Toy/Ch5/affine-lowering.mlir
+++ b/mlir/test/Examples/Toy/Ch5/affine-lowering.mlir
@@ -16,9 +16,9 @@ func @main() {
// CHECK: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// CHECK: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// CHECK: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// CHECK: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_7:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_8:%.*]] = alloc() : memref<2x3xf64>
+// CHECK: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_7:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_8:%.*]] = memref.alloc() : memref<2x3xf64>
// CHECK: affine.store [[VAL_0]], [[VAL_8]][0, 0] : memref<2x3xf64>
// CHECK: affine.store [[VAL_1]], [[VAL_8]][0, 1] : memref<2x3xf64>
// CHECK: affine.store [[VAL_2]], [[VAL_8]][0, 2] : memref<2x3xf64>
@@ -36,9 +36,9 @@ func @main() {
// CHECK: [[VAL_16:%.*]] = mulf [[VAL_14]], [[VAL_15]] : f64
// CHECK: affine.store [[VAL_16]], [[VAL_6]]{{\[}}[[VAL_12]], [[VAL_13]]] : memref<3x2xf64>
// CHECK: toy.print [[VAL_6]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_8]] : memref<2x3xf64>
-// CHECK: dealloc [[VAL_7]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_6]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_8]] : memref<2x3xf64>
+// CHECK: memref.dealloc [[VAL_7]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_6]] : memref<3x2xf64>
// OPT-LABEL: func @main()
// OPT: [[VAL_0:%.*]] = constant 1.000000e+00 : f64
@@ -47,8 +47,8 @@ func @main() {
// OPT: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// OPT: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// OPT: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// OPT: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// OPT: [[VAL_7:%.*]] = alloc() : memref<2x3xf64>
+// OPT: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// OPT: [[VAL_7:%.*]] = memref.alloc() : memref<2x3xf64>
// OPT: affine.store [[VAL_0]], [[VAL_7]][0, 0] : memref<2x3xf64>
// OPT: affine.store [[VAL_1]], [[VAL_7]][0, 1] : memref<2x3xf64>
// OPT: affine.store [[VAL_2]], [[VAL_7]][0, 2] : memref<2x3xf64>
@@ -61,5 +61,5 @@ func @main() {
// OPT: [[VAL_11:%.*]] = mulf [[VAL_10]], [[VAL_10]] : f64
// OPT: affine.store [[VAL_11]], [[VAL_6]]{{\[}}[[VAL_8]], [[VAL_9]]] : memref<3x2xf64>
// OPT: toy.print [[VAL_6]] : memref<3x2xf64>
-// OPT: dealloc [[VAL_7]] : memref<2x3xf64>
-// OPT: dealloc [[VAL_6]] : memref<3x2xf64>
+// OPT: memref.dealloc [[VAL_7]] : memref<2x3xf64>
+// OPT: memref.dealloc [[VAL_6]] : memref<3x2xf64>
diff --git a/mlir/test/Examples/Toy/Ch6/affine-lowering.mlir b/mlir/test/Examples/Toy/Ch6/affine-lowering.mlir
index 79bdd38de6f5..7e0e5ead736c 100644
--- a/mlir/test/Examples/Toy/Ch6/affine-lowering.mlir
+++ b/mlir/test/Examples/Toy/Ch6/affine-lowering.mlir
@@ -16,9 +16,9 @@ func @main() {
// CHECK: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// CHECK: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// CHECK: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// CHECK: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_7:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_8:%.*]] = alloc() : memref<2x3xf64>
+// CHECK: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_7:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_8:%.*]] = memref.alloc() : memref<2x3xf64>
// CHECK: affine.store [[VAL_0]], [[VAL_8]][0, 0] : memref<2x3xf64>
// CHECK: affine.store [[VAL_1]], [[VAL_8]][0, 1] : memref<2x3xf64>
// CHECK: affine.store [[VAL_2]], [[VAL_8]][0, 2] : memref<2x3xf64>
@@ -36,9 +36,9 @@ func @main() {
// CHECK: [[VAL_16:%.*]] = mulf [[VAL_14]], [[VAL_15]] : f64
// CHECK: affine.store [[VAL_16]], [[VAL_6]]{{\[}}[[VAL_12]], [[VAL_13]]] : memref<3x2xf64>
// CHECK: toy.print [[VAL_6]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_8]] : memref<2x3xf64>
-// CHECK: dealloc [[VAL_7]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_6]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_8]] : memref<2x3xf64>
+// CHECK: memref.dealloc [[VAL_7]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_6]] : memref<3x2xf64>
// OPT-LABEL: func @main()
// OPT: [[VAL_0:%.*]] = constant 1.000000e+00 : f64
@@ -47,8 +47,8 @@ func @main() {
// OPT: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// OPT: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// OPT: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// OPT: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// OPT: [[VAL_7:%.*]] = alloc() : memref<2x3xf64>
+// OPT: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// OPT: [[VAL_7:%.*]] = memref.alloc() : memref<2x3xf64>
// OPT: affine.store [[VAL_0]], [[VAL_7]][0, 0] : memref<2x3xf64>
// OPT: affine.store [[VAL_1]], [[VAL_7]][0, 1] : memref<2x3xf64>
// OPT: affine.store [[VAL_2]], [[VAL_7]][0, 2] : memref<2x3xf64>
@@ -61,5 +61,5 @@ func @main() {
// OPT: [[VAL_11:%.*]] = mulf [[VAL_10]], [[VAL_10]] : f64
// OPT: affine.store [[VAL_11]], [[VAL_6]]{{\[}}[[VAL_8]], [[VAL_9]]] : memref<3x2xf64>
// OPT: toy.print [[VAL_6]] : memref<3x2xf64>
-// OPT: dealloc [[VAL_7]] : memref<2x3xf64>
-// OPT: dealloc [[VAL_6]] : memref<3x2xf64>
+// OPT: memref.dealloc [[VAL_7]] : memref<2x3xf64>
+// OPT: memref.dealloc [[VAL_6]] : memref<3x2xf64>
diff --git a/mlir/test/Examples/Toy/Ch7/affine-lowering.mlir b/mlir/test/Examples/Toy/Ch7/affine-lowering.mlir
index 4054eb0aba8e..e19d8f87ab07 100644
--- a/mlir/test/Examples/Toy/Ch7/affine-lowering.mlir
+++ b/mlir/test/Examples/Toy/Ch7/affine-lowering.mlir
@@ -16,9 +16,9 @@ func @main() {
// CHECK: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// CHECK: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// CHECK: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// CHECK: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_7:%.*]] = alloc() : memref<3x2xf64>
-// CHECK: [[VAL_8:%.*]] = alloc() : memref<2x3xf64>
+// CHECK: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_7:%.*]] = memref.alloc() : memref<3x2xf64>
+// CHECK: [[VAL_8:%.*]] = memref.alloc() : memref<2x3xf64>
// CHECK: affine.store [[VAL_0]], [[VAL_8]][0, 0] : memref<2x3xf64>
// CHECK: affine.store [[VAL_1]], [[VAL_8]][0, 1] : memref<2x3xf64>
// CHECK: affine.store [[VAL_2]], [[VAL_8]][0, 2] : memref<2x3xf64>
@@ -36,9 +36,9 @@ func @main() {
// CHECK: [[VAL_16:%.*]] = mulf [[VAL_14]], [[VAL_15]] : f64
// CHECK: affine.store [[VAL_16]], [[VAL_6]]{{\[}}[[VAL_12]], [[VAL_13]]] : memref<3x2xf64>
// CHECK: toy.print [[VAL_6]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_8]] : memref<2x3xf64>
-// CHECK: dealloc [[VAL_7]] : memref<3x2xf64>
-// CHECK: dealloc [[VAL_6]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_8]] : memref<2x3xf64>
+// CHECK: memref.dealloc [[VAL_7]] : memref<3x2xf64>
+// CHECK: memref.dealloc [[VAL_6]] : memref<3x2xf64>
// OPT-LABEL: func @main()
// OPT: [[VAL_0:%.*]] = constant 1.000000e+00 : f64
@@ -47,8 +47,8 @@ func @main() {
// OPT: [[VAL_3:%.*]] = constant 4.000000e+00 : f64
// OPT: [[VAL_4:%.*]] = constant 5.000000e+00 : f64
// OPT: [[VAL_5:%.*]] = constant 6.000000e+00 : f64
-// OPT: [[VAL_6:%.*]] = alloc() : memref<3x2xf64>
-// OPT: [[VAL_7:%.*]] = alloc() : memref<2x3xf64>
+// OPT: [[VAL_6:%.*]] = memref.alloc() : memref<3x2xf64>
+// OPT: [[VAL_7:%.*]] = memref.alloc() : memref<2x3xf64>
// OPT: affine.store [[VAL_0]], [[VAL_7]][0, 0] : memref<2x3xf64>
// OPT: affine.store [[VAL_1]], [[VAL_7]][0, 1] : memref<2x3xf64>
// OPT: affine.store [[VAL_2]], [[VAL_7]][0, 2] : memref<2x3xf64>
@@ -61,5 +61,5 @@ func @main() {
// OPT: [[VAL_11:%.*]] = mulf [[VAL_10]], [[VAL_10]] : f64
// OPT: affine.store [[VAL_11]], [[VAL_6]]{{\[}}[[VAL_8]], [[VAL_9]]] : memref<3x2xf64>
// OPT: toy.print [[VAL_6]] : memref<3x2xf64>
-// OPT: dealloc [[VAL_7]] : memref<2x3xf64>
-// OPT: dealloc [[VAL_6]] : memref<3x2xf64>
+// OPT: memref.dealloc [[VAL_7]] : memref<2x3xf64>
+// OPT: memref.dealloc [[VAL_6]] : memref<3x2xf64>
diff --git a/mlir/test/IR/core-ops.mlir b/mlir/test/IR/core-ops.mlir
index b760157bb828..43108465df1a 100644
--- a/mlir/test/IR/core-ops.mlir
+++ b/mlir/test/IR/core-ops.mlir
@@ -35,9 +35,9 @@ func @func_with_ops(f32) {
%t = "getTensor"() : () -> tensor<4x4x?xf32>
// CHECK: %[[C2:.*]] = constant 2 : index
- // CHECK-NEXT: %{{.*}} = dim %[[T]], %[[C2]] : tensor<4x4x?xf32>
+ // CHECK-NEXT: %{{.*}} = memref.dim %[[T]], %[[C2]] : tensor<4x4x?xf32>
%c2 = constant 2 : index
- %t2 = "std.dim"(%t, %c2) : (tensor<4x4x?xf32>, index) -> index
+ %t2 = "memref.dim"(%t, %c2) : (tensor<4x4x?xf32>, index) -> index
// CHECK: %{{.*}} = addf %[[ARG]], %[[ARG]] : f32
%x = "std.addf"(%a, %a) : (f32,f32) -> (f32)
@@ -50,9 +50,9 @@ func @func_with_ops(f32) {
func @standard_instrs(tensor<4x4x?xf32>, f32, i32, index, i64, f16) {
^bb42(%t: tensor<4x4x?xf32>, %f: f32, %i: i32, %idx : index, %j: i64, %half: f16):
// CHECK: %[[C2:.*]] = constant 2 : index
- // CHECK: %[[A2:.*]] = dim %arg0, %[[C2]] : tensor<4x4x?xf32>
+ // CHECK: %[[A2:.*]] = memref.dim %arg0, %[[C2]] : tensor<4x4x?xf32>
%c2 = constant 2 : index
- %a2 = dim %t, %c2 : tensor<4x4x?xf32>
+ %a2 = memref.dim %t, %c2 : tensor<4x4x?xf32>
// CHECK: %[[F2:.*]] = addf %arg1, %arg1 : f32
%f2 = "std.addf"(%f, %f) : (f32,f32) -> f32
@@ -540,17 +540,17 @@ func @affine_apply() {
// CHECK-LABEL: func @load_store_prefetch
func @load_store_prefetch(memref<4x4xi32>, index) {
^bb0(%0: memref<4x4xi32>, %1: index):
- // CHECK: %0 = load %arg0[%arg1, %arg1] : memref<4x4xi32>
- %2 = "std.load"(%0, %1, %1) : (memref<4x4xi32>, index, index)->i32
+ // CHECK: %0 = memref.load %arg0[%arg1, %arg1] : memref<4x4xi32>
+ %2 = "memref.load"(%0, %1, %1) : (memref<4x4xi32>, index, index)->i32
- // CHECK: %{{.*}} = load %arg0[%arg1, %arg1] : memref<4x4xi32>
- %3 = load %0[%1, %1] : memref<4x4xi32>
+ // CHECK: %{{.*}} = memref.load %arg0[%arg1, %arg1] : memref<4x4xi32>
+ %3 = memref.load %0[%1, %1] : memref<4x4xi32>
- // CHECK: prefetch %arg0[%arg1, %arg1], write, locality<1>, data : memref<4x4xi32>
- prefetch %0[%1, %1], write, locality<1>, data : memref<4x4xi32>
+ // CHECK: memref.prefetch %arg0[%arg1, %arg1], write, locality<1>, data : memref<4x4xi32>
+ memref.prefetch %0[%1, %1], write, locality<1>, data : memref<4x4xi32>
- // CHECK: prefetch %arg0[%arg1, %arg1], read, locality<3>, instr : memref<4x4xi32>
- prefetch %0[%1, %1], read, locality<3>, instr : memref<4x4xi32>
+ // CHECK: memref.prefetch %arg0[%arg1, %arg1], read, locality<3>, instr : memref<4x4xi32>
+ memref.prefetch %0[%1, %1], read, locality<3>, instr : memref<4x4xi32>
return
}
@@ -558,11 +558,11 @@ func @load_store_prefetch(memref<4x4xi32>, index) {
// Test with zero-dimensional operands using no index in load/store.
// CHECK-LABEL: func @zero_dim_no_idx
func @zero_dim_no_idx(%arg0 : memref<i32>, %arg1 : memref<i32>, %arg2 : memref<i32>) {
- %0 = std.load %arg0[] : memref<i32>
- std.store %0, %arg1[] : memref<i32>
+ %0 = memref.load %arg0[] : memref<i32>
+ memref.store %0, %arg1[] : memref<i32>
return
- // CHECK: %0 = load %{{.*}}[] : memref<i32>
- // CHECK: store %{{.*}}, %{{.*}}[] : memref<i32>
+ // CHECK: %0 = memref.load %{{.*}}[] : memref<i32>
+ // CHECK: memref.store %{{.*}}, %{{.*}}[] : memref<i32>
}
// CHECK-LABEL: func @return_op(%arg0: i32) -> i32 {
@@ -600,23 +600,23 @@ func @calls(%arg0: i32) {
// CHECK-LABEL: func @memref_cast(%arg0
func @memref_cast(%arg0: memref<4xf32>, %arg1 : memref<?xf32>, %arg2 : memref<64x16x4xf32, offset: 0, strides: [64, 4, 1]>) {
- // CHECK: %0 = memref_cast %arg0 : memref<4xf32> to memref<?xf32>
- %0 = memref_cast %arg0 : memref<4xf32> to memref<?xf32>
+ // CHECK: %0 = memref.cast %arg0 : memref<4xf32> to memref<?xf32>
+ %0 = memref.cast %arg0 : memref<4xf32> to memref<?xf32>
- // CHECK: %1 = memref_cast %arg1 : memref<?xf32> to memref<4xf32>
- %1 = memref_cast %arg1 : memref<?xf32> to memref<4xf32>
+ // CHECK: %1 = memref.cast %arg1 : memref<?xf32> to memref<4xf32>
+ %1 = memref.cast %arg1 : memref<?xf32> to memref<4xf32>
- // CHECK: {{%.*}} = memref_cast %arg2 : memref<64x16x4xf32, #[[$BASE_MAP0]]> to memref<64x16x4xf32, #[[$BASE_MAP3]]>
- %2 = memref_cast %arg2 : memref<64x16x4xf32, offset: 0, strides: [64, 4, 1]> to memref<64x16x4xf32, offset: ?, strides: [?, ?, ?]>
+ // CHECK: {{%.*}} = memref.cast %arg2 : memref<64x16x4xf32, #[[$BASE_MAP0]]> to memref<64x16x4xf32, #[[$BASE_MAP3]]>
+ %2 = memref.cast %arg2 : memref<64x16x4xf32, offset: 0, strides: [64, 4, 1]> to memref<64x16x4xf32, offset: ?, strides: [?, ?, ?]>
- // CHECK: {{%.*}} = memref_cast {{%.*}} : memref<64x16x4xf32, #[[$BASE_MAP3]]> to memref<64x16x4xf32, #[[$BASE_MAP0]]>
- %3 = memref_cast %2 : memref<64x16x4xf32, offset: ?, strides: [?, ?, ?]> to memref<64x16x4xf32, offset: 0, strides: [64, 4, 1]>
+ // CHECK: {{%.*}} = memref.cast {{%.*}} : memref<64x16x4xf32, #[[$BASE_MAP3]]> to memref<64x16x4xf32, #[[$BASE_MAP0]]>
+ %3 = memref.cast %2 : memref<64x16x4xf32, offset: ?, strides: [?, ?, ?]> to memref<64x16x4xf32, offset: 0, strides: [64, 4, 1]>
- // CHECK: memref_cast %{{.*}} : memref<4xf32> to memref<*xf32>
- %4 = memref_cast %1 : memref<4xf32> to memref<*xf32>
+ // CHECK: memref.cast %{{.*}} : memref<4xf32> to memref<*xf32>
+ %4 = memref.cast %1 : memref<4xf32> to memref<*xf32>
- // CHECK: memref_cast %{{.*}} : memref<*xf32> to memref<4xf32>
- %5 = memref_cast %4 : memref<*xf32> to memref<4xf32>
+ // CHECK: memref.cast %{{.*}} : memref<*xf32> to memref<4xf32>
+ %5 = memref.cast %4 : memref<*xf32> to memref<4xf32>
return
}
@@ -627,19 +627,19 @@ func private @unranked_memref_roundtrip(memref<*xf32, 4>)
// CHECK-LABEL: func @memref_view(%arg0
func @memref_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
// Test two dynamic sizes and dynamic offset.
- // CHECK: %{{.*}} = std.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
- %1 = view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
+ // CHECK: %{{.*}} = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
+ %1 = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8> to memref<?x?xf32>
// Test one dynamic size and dynamic offset.
- // CHECK: %{{.*}} = std.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
- %3 = view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
+ // CHECK: %{{.*}} = memref.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
+ %3 = memref.view %0[%arg2][%arg1] : memref<2048xi8> to memref<4x?xf32>
// Test static sizes and static offset.
- // CHECK: %{{.*}} = std.view %0[{{.*}}][] : memref<2048xi8> to memref<64x4xf32>
+ // CHECK: %{{.*}} = memref.view %0[{{.*}}][] : memref<2048xi8> to memref<64x4xf32>
%c0 = constant 0: index
- %5 = view %0[%c0][] : memref<2048xi8> to memref<64x4xf32>
+ %5 = memref.view %0[%c0][] : memref<2048xi8> to memref<64x4xf32>
return
}
@@ -648,107 +648,107 @@ func @memref_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = alloc() : memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)>>
+ %0 = memref.alloc() : memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)>>
// CHECK: subview %0[%c0, %c0, %c0] [%arg0, %arg1, %arg2] [%c1, %c1, %c1] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<?x?x?xf32, #[[$BASE_MAP3]]>
- %1 = subview %0[%c0, %c0, %c0][%arg0, %arg1, %arg2][%c1, %c1, %c1]
+ %1 = memref.subview %0[%c0, %c0, %c0][%arg0, %arg1, %arg2][%c1, %c1, %c1]
: memref<8x16x4xf32, offset:0, strides: [64, 4, 1]> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- %2 = alloc()[%arg2] : memref<64xf32, affine_map<(d0)[s0] -> (d0 + s0)>>
- // CHECK: subview %2[%c1] [%arg0] [%c1] :
+ %2 = memref.alloc()[%arg2] : memref<64xf32, affine_map<(d0)[s0] -> (d0 + s0)>>
+ // CHECK: memref.subview %2[%c1] [%arg0] [%c1] :
// CHECK-SAME: memref<64xf32, #[[$BASE_MAP1]]>
// CHECK-SAME: to memref<?xf32, #[[$SUBVIEW_MAP1]]>
- %3 = subview %2[%c1][%arg0][%c1]
+ %3 = memref.subview %2[%c1][%arg0][%c1]
: memref<64xf32, affine_map<(d0)[s0] -> (d0 + s0)>> to
memref<?xf32, affine_map<(d0)[s0, s1] -> (d0 * s1 + s0)>>
- %4 = alloc() : memref<64x22xf32, affine_map<(d0, d1) -> (d0 * 22 + d1)>>
- // CHECK: subview %4[%c0, %c1] [%arg0, %arg1] [%c1, %c0] :
+ %4 = memref.alloc() : memref<64x22xf32, affine_map<(d0, d1) -> (d0 * 22 + d1)>>
+ // CHECK: memref.subview %4[%c0, %c1] [%arg0, %arg1] [%c1, %c0] :
// CHECK-SAME: memref<64x22xf32, #[[$BASE_MAP2]]>
// CHECK-SAME: to memref<?x?xf32, #[[$SUBVIEW_MAP2]]>
- %5 = subview %4[%c0, %c1][%arg0, %arg1][%c1, %c0]
+ %5 = memref.subview %4[%c0, %c1][%arg0, %arg1][%c1, %c0]
: memref<64x22xf32, offset:0, strides: [22, 1]> to
memref<?x?xf32, offset:?, strides: [?, ?]>
- // CHECK: subview %0[0, 2, 0] [4, 4, 4] [1, 1, 1] :
+ // CHECK: memref.subview %0[0, 2, 0] [4, 4, 4] [1, 1, 1] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<4x4x4xf32, #[[$SUBVIEW_MAP3]]>
- %6 = subview %0[0, 2, 0][4, 4, 4][1, 1, 1]
+ %6 = memref.subview %0[0, 2, 0][4, 4, 4][1, 1, 1]
: memref<8x16x4xf32, offset:0, strides: [64, 4, 1]> to
memref<4x4x4xf32, offset:8, strides: [64, 4, 1]>
- %7 = alloc(%arg1, %arg2) : memref<?x?xf32>
- // CHECK: subview {{%.*}}[0, 0] [4, 4] [1, 1] :
+ %7 = memref.alloc(%arg1, %arg2) : memref<?x?xf32>
+ // CHECK: memref.subview {{%.*}}[0, 0] [4, 4] [1, 1] :
// CHECK-SAME: memref<?x?xf32>
// CHECK-SAME: to memref<4x4xf32, #[[$SUBVIEW_MAP4]]>
- %8 = subview %7[0, 0][4, 4][1, 1]
+ %8 = memref.subview %7[0, 0][4, 4][1, 1]
: memref<?x?xf32> to memref<4x4xf32, offset: ?, strides:[?, 1]>
- %9 = alloc() : memref<16x4xf32>
- // CHECK: subview {{%.*}}[{{%.*}}, {{%.*}}] [4, 4] [{{%.*}}, {{%.*}}] :
+ %9 = memref.alloc() : memref<16x4xf32>
+ // CHECK: memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [4, 4] [{{%.*}}, {{%.*}}] :
// CHECK-SAME: memref<16x4xf32>
// CHECK-SAME: to memref<4x4xf32, #[[$SUBVIEW_MAP2]]
- %10 = subview %9[%arg1, %arg1][4, 4][%arg2, %arg2]
+ %10 = memref.subview %9[%arg1, %arg1][4, 4][%arg2, %arg2]
: memref<16x4xf32> to memref<4x4xf32, offset: ?, strides:[?, ?]>
- // CHECK: subview {{%.*}}[{{%.*}}, {{%.*}}] [4, 4] [2, 2] :
+ // CHECK: memref.subview {{%.*}}[{{%.*}}, {{%.*}}] [4, 4] [2, 2] :
// CHECK-SAME: memref<16x4xf32>
// CHECK-SAME: to memref<4x4xf32, #[[$SUBVIEW_MAP5]]
- %11 = subview %9[%arg1, %arg2][4, 4][2, 2]
+ %11 = memref.subview %9[%arg1, %arg2][4, 4][2, 2]
: memref<16x4xf32> to memref<4x4xf32, offset: ?, strides:[8, 2]>
- %12 = alloc() : memref<1x9x1x4x1xf32, affine_map<(d0, d1, d2, d3, d4) -> (36 * d0 + 36 * d1 + 4 * d2 + 4 * d3 + d4)>>
- // CHECK: subview %12[%arg1, %arg1, %arg1, %arg1, %arg1]
+ %12 = memref.alloc() : memref<1x9x1x4x1xf32, affine_map<(d0, d1, d2, d3, d4) -> (36 * d0 + 36 * d1 + 4 * d2 + 4 * d3 + d4)>>
+ // CHECK: memref.subview %12[%arg1, %arg1, %arg1, %arg1, %arg1]
// CHECK-SAME: [1, 9, 1, 4, 1] [%arg2, %arg2, %arg2, %arg2, %arg2] :
// CHECK-SAME: memref<1x9x1x4x1xf32, #[[$SUBVIEW_MAP6]]> to memref<9x4xf32, #[[$SUBVIEW_MAP2]]>
- %13 = subview %12[%arg1, %arg1, %arg1, %arg1, %arg1][1, 9, 1, 4, 1][%arg2, %arg2, %arg2, %arg2, %arg2] : memref<1x9x1x4x1xf32, offset: 0, strides: [36, 36, 4, 4, 1]> to memref<9x4xf32, offset: ?, strides: [?, ?]>
- // CHECK: subview %12[%arg1, %arg1, %arg1, %arg1, %arg1]
+ %13 = memref.subview %12[%arg1, %arg1, %arg1, %arg1, %arg1][1, 9, 1, 4, 1][%arg2, %arg2, %arg2, %arg2, %arg2] : memref<1x9x1x4x1xf32, offset: 0, strides: [36, 36, 4, 4, 1]> to memref<9x4xf32, offset: ?, strides: [?, ?]>
+ // CHECK: memref.subview %12[%arg1, %arg1, %arg1, %arg1, %arg1]
// CHECK-SAME: [1, 9, 1, 4, 1] [%arg2, %arg2, %arg2, %arg2, %arg2] :
// CHECK-SAME: memref<1x9x1x4x1xf32, #[[$SUBVIEW_MAP6]]> to memref<1x9x4xf32, #[[$BASE_MAP3]]>
- %14 = subview %12[%arg1, %arg1, %arg1, %arg1, %arg1][1, 9, 1, 4, 1][%arg2, %arg2, %arg2, %arg2, %arg2] : memref<1x9x1x4x1xf32, offset: 0, strides: [36, 36, 4, 4, 1]> to memref<1x9x4xf32, offset: ?, strides: [?, ?, ?]>
+ %14 = memref.subview %12[%arg1, %arg1, %arg1, %arg1, %arg1][1, 9, 1, 4, 1][%arg2, %arg2, %arg2, %arg2, %arg2] : memref<1x9x1x4x1xf32, offset: 0, strides: [36, 36, 4, 4, 1]> to memref<1x9x4xf32, offset: ?, strides: [?, ?, ?]>
- %15 = alloc(%arg1, %arg2)[%c0, %c1, %arg1, %arg0, %arg0, %arg2, %arg2] : memref<1x?x5x1x?x1xf32, affine_map<(d0, d1, d2, d3, d4, d5)[s0, s1, s2, s3, s4, s5, s6] -> (s0 + s1 * d0 + s2 * d1 + s3 * d2 + s4 * d3 + s5 * d4 + s6 * d5)>>
- // CHECK: subview %15[0, 0, 0, 0, 0, 0] [1, %arg1, 5, 1, %arg2, 1] [1, 1, 1, 1, 1, 1] :
+ %15 = memref.alloc(%arg1, %arg2)[%c0, %c1, %arg1, %arg0, %arg0, %arg2, %arg2] : memref<1x?x5x1x?x1xf32, affine_map<(d0, d1, d2, d3, d4, d5)[s0, s1, s2, s3, s4, s5, s6] -> (s0 + s1 * d0 + s2 * d1 + s3 * d2 + s4 * d3 + s5 * d4 + s6 * d5)>>
+ // CHECK: memref.subview %15[0, 0, 0, 0, 0, 0] [1, %arg1, 5, 1, %arg2, 1] [1, 1, 1, 1, 1, 1] :
// CHECK-SAME: memref<1x?x5x1x?x1xf32, #[[$SUBVIEW_MAP7]]> to memref<?x5x?xf32, #[[$BASE_MAP3]]>
- %16 = subview %15[0, 0, 0, 0, 0, 0][1, %arg1, 5, 1, %arg2, 1][1, 1, 1, 1, 1, 1] : memref<1x?x5x1x?x1xf32, offset: ?, strides: [?, ?, ?, ?, ?, ?]> to memref<?x5x?xf32, offset: ?, strides: [?, ?, ?]>
- // CHECK: subview %15[%arg1, %arg1, %arg1, %arg1, %arg1, %arg1] [1, %arg1, 5, 1, %arg2, 1] [1, 1, 1, 1, 1, 1] :
+ %16 = memref.subview %15[0, 0, 0, 0, 0, 0][1, %arg1, 5, 1, %arg2, 1][1, 1, 1, 1, 1, 1] : memref<1x?x5x1x?x1xf32, offset: ?, strides: [?, ?, ?, ?, ?, ?]> to memref<?x5x?xf32, offset: ?, strides: [?, ?, ?]>
+ // CHECK: memref.subview %15[%arg1, %arg1, %arg1, %arg1, %arg1, %arg1] [1, %arg1, 5, 1, %arg2, 1] [1, 1, 1, 1, 1, 1] :
// CHECK-SAME: memref<1x?x5x1x?x1xf32, #[[$SUBVIEW_MAP7]]> to memref<?x5x?x1xf32, #[[$SUBVIEW_MAP8]]>
- %17 = subview %15[%arg1, %arg1, %arg1, %arg1, %arg1, %arg1][1, %arg1, 5, 1, %arg2, 1][1, 1, 1, 1, 1, 1] : memref<1x?x5x1x?x1xf32, offset: ?, strides: [?, ?, ?, ?, ?, ?]> to memref<?x5x?x1xf32, offset: ?, strides: [?, ?, ?, ?]>
+ %17 = memref.subview %15[%arg1, %arg1, %arg1, %arg1, %arg1, %arg1][1, %arg1, 5, 1, %arg2, 1][1, 1, 1, 1, 1, 1] : memref<1x?x5x1x?x1xf32, offset: ?, strides: [?, ?, ?, ?, ?, ?]> to memref<?x5x?x1xf32, offset: ?, strides: [?, ?, ?, ?]>
- %18 = alloc() : memref<1x8xf32>
- // CHECK: subview %18[0, 0] [1, 8] [1, 1] : memref<1x8xf32> to memref<8xf32>
- %19 = subview %18[0, 0][1, 8][1, 1] : memref<1x8xf32> to memref<8xf32>
+ %18 = memref.alloc() : memref<1x8xf32>
+ // CHECK: memref.subview %18[0, 0] [1, 8] [1, 1] : memref<1x8xf32> to memref<8xf32>
+ %19 = memref.subview %18[0, 0][1, 8][1, 1] : memref<1x8xf32> to memref<8xf32>
- %20 = alloc() : memref<8x16x4xf32>
- // CHECK: subview %20[0, 0, 0] [1, 16, 4] [1, 1, 1] : memref<8x16x4xf32> to memref<16x4xf32>
- %21 = subview %20[0, 0, 0][1, 16, 4][1, 1, 1] : memref<8x16x4xf32> to memref<16x4xf32>
+ %20 = memref.alloc() : memref<8x16x4xf32>
+ // CHECK: memref.subview %20[0, 0, 0] [1, 16, 4] [1, 1, 1] : memref<8x16x4xf32> to memref<16x4xf32>
+ %21 = memref.subview %20[0, 0, 0][1, 16, 4][1, 1, 1] : memref<8x16x4xf32> to memref<16x4xf32>
- %22 = subview %20[3, 4, 2][1, 6, 3][1, 1, 1] : memref<8x16x4xf32> to memref<6x3xf32, offset: 210, strides: [4, 1]>
+ %22 = memref.subview %20[3, 4, 2][1, 6, 3][1, 1, 1] : memref<8x16x4xf32> to memref<6x3xf32, offset: 210, strides: [4, 1]>
- %23 = alloc() : memref<f32>
- %78 = subview %23[] [] [] : memref<f32> to memref<f32>
+ %23 = memref.alloc() : memref<f32>
+ %78 = memref.subview %23[] [] [] : memref<f32> to memref<f32>
/// Subview with only leading operands.
- %24 = alloc() : memref<5x3xf32>
- // CHECK: subview %{{.*}}[2] [3] [1] : memref<5x3xf32> to memref<3x3xf32, #[[$SUBVIEW_MAP9]]>
- %25 = subview %24[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
+ %24 = memref.alloc() : memref<5x3xf32>
+ // CHECK: memref.subview %{{.*}}[2] [3] [1] : memref<5x3xf32> to memref<3x3xf32, #[[$SUBVIEW_MAP9]]>
+ %25 = memref.subview %24[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
/// Rank-reducing subview with only leading operands.
- // CHECK: subview %{{.*}}[1] [1] [1] : memref<5x3xf32> to memref<3xf32, #[[$SUBVIEW_MAP10]]>
- %26 = subview %24[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
+ // CHECK: memref.subview %{{.*}}[1] [1] [1] : memref<5x3xf32> to memref<3xf32, #[[$SUBVIEW_MAP10]]>
+ %26 = memref.subview %24[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
// Corner-case of 0-D rank-reducing subview with an offset.
- // CHECK: subview %{{.*}}[1, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, #[[$SUBVIEW_MAP11]]>
- %27 = subview %24[1, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, affine_map<() -> (4)>>
+ // CHECK: memref.subview %{{.*}}[1, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, #[[$SUBVIEW_MAP11]]>
+ %27 = memref.subview %24[1, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, affine_map<() -> (4)>>
- // CHECK: subview %{{.*}}[%{{.*}}, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, #[[$SUBVIEW_MAP12]]>
- %28 = subview %24[%arg0, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, affine_map<()[s0] -> (s0)>>
+ // CHECK: memref.subview %{{.*}}[%{{.*}}, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, #[[$SUBVIEW_MAP12]]>
+ %28 = memref.subview %24[%arg0, 1] [1, 1] [1, 1] : memref<5x3xf32> to memref<f32, affine_map<()[s0] -> (s0)>>
- // CHECK: subview %{{.*}}[0, %{{.*}}] [%{{.*}}, 1] [1, 1] : memref<?x?xf32> to memref<?xf32, #[[$SUBVIEW_MAP1]]>
- %a30 = alloc(%arg0, %arg0) : memref<?x?xf32>
- %30 = subview %a30[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, affine_map<(d0)[s0, s1] -> (d0 * s1 + s0)>>
+ // CHECK: memref.subview %{{.*}}[0, %{{.*}}] [%{{.*}}, 1] [1, 1] : memref<?x?xf32> to memref<?xf32, #[[$SUBVIEW_MAP1]]>
+ %a30 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
+ %30 = memref.subview %a30[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, affine_map<(d0)[s0, s1] -> (d0 * s1 + s0)>>
return
}
@@ -757,9 +757,9 @@ func @memref_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK-SAME: %[[ARG:.*]]: tensor<4x4x?xf32>
func @test_dimop(%arg0: tensor<4x4x?xf32>) {
// CHECK: %[[C2:.*]] = constant 2 : index
- // CHECK: %{{.*}} = dim %[[ARG]], %[[C2]] : tensor<4x4x?xf32>
+ // CHECK: %{{.*}} = memref.dim %[[ARG]], %[[C2]] : tensor<4x4x?xf32>
%c2 = constant 2 : index
- %0 = dim %arg0, %c2 : tensor<4x4x?xf32>
+ %0 = memref.dim %arg0, %c2 : tensor<4x4x?xf32>
// use dim as an index to ensure type correctness
%1 = affine.apply affine_map<(d0) -> (d0)>(%0)
return
@@ -779,19 +779,19 @@ func @test_splat_op(%s : f32) {
// CHECK-LABEL: func @tensor_load_store
func @tensor_load_store(%0 : memref<4x4xi32>) {
- // CHECK: %[[TENSOR:.*]] = tensor_load %[[MEMREF:.*]] : memref<4x4xi32>
- %1 = tensor_load %0 : memref<4x4xi32>
- // CHECK: tensor_store %[[TENSOR]], %[[MEMREF]] : memref<4x4xi32>
- tensor_store %1, %0 : memref<4x4xi32>
+ // CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[MEMREF:.*]] : memref<4x4xi32>
+ %1 = memref.tensor_load %0 : memref<4x4xi32>
+ // CHECK: memref.tensor_store %[[TENSOR]], %[[MEMREF]] : memref<4x4xi32>
+ memref.tensor_store %1, %0 : memref<4x4xi32>
return
}
// CHECK-LABEL: func @unranked_tensor_load_store
func @unranked_tensor_load_store(%0 : memref<*xi32>) {
- // CHECK: %[[TENSOR:.*]] = tensor_load %[[MEMREF:.*]] : memref<*xi32>
- %1 = tensor_load %0 : memref<*xi32>
- // CHECK: tensor_store %[[TENSOR]], %[[MEMREF]] : memref<*xi32>
- tensor_store %1, %0 : memref<*xi32>
+ // CHECK: %[[TENSOR:.*]] = memref.tensor_load %[[MEMREF:.*]] : memref<*xi32>
+ %1 = memref.tensor_load %0 : memref<*xi32>
+ // CHECK: memref.tensor_store %[[TENSOR]], %[[MEMREF]] : memref<*xi32>
+ memref.tensor_store %1, %0 : memref<*xi32>
return
}
@@ -820,8 +820,8 @@ func @generic_atomic_rmw(%I: memref<1x2xf32>, %i : index, %j : index) {
// CHECK-LABEL: func @assume_alignment
// CHECK-SAME: %[[MEMREF:.*]]: memref<4x4xf16>
func @assume_alignment(%0: memref<4x4xf16>) {
- // CHECK: assume_alignment %[[MEMREF]], 16 : memref<4x4xf16>
- assume_alignment %0, 16 : memref<4x4xf16>
+ // CHECK: memref.assume_alignment %[[MEMREF]], 16 : memref<4x4xf16>
+ memref.assume_alignment %0, 16 : memref<4x4xf16>
return
}
diff --git a/mlir/test/IR/invalid-ops.mlir b/mlir/test/IR/invalid-ops.mlir
index 797c1d478364..3747039944b5 100644
--- a/mlir/test/IR/invalid-ops.mlir
+++ b/mlir/test/IR/invalid-ops.mlir
@@ -2,7 +2,7 @@
func @dim(%arg : tensor<1x?xf32>) {
%c2 = constant 2 : index
- dim %arg, %c2 : tensor<1x?xf32> // expected-error {{'std.dim' op index is out of range}}
+ memref.dim %arg, %c2 : tensor<1x?xf32> // expected-error {{'memref.dim' op index is out of range}}
return
}
@@ -10,7 +10,8 @@ func @dim(%arg : tensor<1x?xf32>) {
func @rank(f32) {
^bb(%0: f32):
- "std.rank"(%0): (f32)->index // expected-error {{'std.rank' op operand #0 must be any tensor or memref type}}
+ "std.rank"(%0): (f32)->index // expected-error {{'std.rank' op operand #0 must be any memref or tensor type}}
+
return
}
@@ -88,7 +89,7 @@ func @bad_alloc_wrong_dynamic_dim_count() {
%0 = constant 7 : index
// Test alloc with wrong number of dynamic dimensions.
// expected-error at +1 {{dimension operand count does not equal memref dynamic dimension count}}
- %1 = alloc(%0)[%0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ %1 = memref.alloc(%0)[%0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
return
}
@@ -99,7 +100,7 @@ func @bad_alloc_wrong_symbol_count() {
%0 = constant 7 : index
// Test alloc with wrong number of symbols
// expected-error at +1 {{symbol operand count does not equal memref symbol count}}
- %1 = alloc(%0) : memref<2x?xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ %1 = memref.alloc(%0) : memref<2x?xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
return
}
@@ -107,19 +108,19 @@ func @bad_alloc_wrong_symbol_count() {
func @test_store_zero_results() {
^bb0:
- %0 = alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ %0 = memref.alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
%1 = constant 0 : index
%2 = constant 1 : index
- %3 = load %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ %3 = memref.load %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
// Test that store returns zero results.
- %4 = store %3, %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1> // expected-error {{cannot name an operation with no results}}
+ %4 = memref.store %3, %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1> // expected-error {{cannot name an operation with no results}}
return
}
// -----
func @test_store_zero_results2(%x: i32, %p: memref<i32>) {
- "std.store"(%x,%p) : (i32, memref<i32>) -> i32 // expected-error {{'std.store' op requires zero results}}
+ "memref.store"(%x,%p) : (i32, memref<i32>) -> i32 // expected-error {{'memref.store' op requires zero results}}
return
}
@@ -127,7 +128,7 @@ func @test_store_zero_results2(%x: i32, %p: memref<i32>) {
func @test_alloc_memref_map_rank_mismatch() {
^bb0:
- %0 = alloc() : memref<1024x64xf32, affine_map<(d0) -> (d0)>, 1> // expected-error {{memref affine map dimension mismatch}}
+ %0 = memref.alloc() : memref<1024x64xf32, affine_map<(d0) -> (d0)>, 1> // expected-error {{memref affine map dimension mismatch}}
return
}
@@ -291,14 +292,14 @@ func @invalid_cmp_shape(%idx : () -> ()) {
func @dma_start_not_enough_operands() {
// expected-error at +1 {{expected at least 4 operands}}
- "std.dma_start"() : () -> ()
+ "memref.dma_start"() : () -> ()
}
// -----
func @dma_no_src_memref(%m : f32, %tag : f32, %c0 : index) {
// expected-error at +1 {{expected source to be of memref type}}
- dma_start %m[%c0], %m[%c0], %c0, %tag[%c0] : f32, f32, f32
+ memref.dma_start %m[%c0], %m[%c0], %c0, %tag[%c0] : f32, f32, f32
}
// -----
@@ -306,7 +307,7 @@ func @dma_no_src_memref(%m : f32, %tag : f32, %c0 : index) {
func @dma_start_not_enough_operands_for_src(
%src: memref<2x2x2xf32>, %idx: index) {
// expected-error at +1 {{expected at least 7 operands}}
- "std.dma_start"(%src, %idx, %idx, %idx) : (memref<2x2x2xf32>, index, index, index) -> ()
+ "memref.dma_start"(%src, %idx, %idx, %idx) : (memref<2x2x2xf32>, index, index, index) -> ()
}
// -----
@@ -315,16 +316,16 @@ func @dma_start_src_index_wrong_type(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>, %flt: f32) {
// expected-error at +1 {{expected source indices to be of index type}}
- "std.dma_start"(%src, %idx, %flt, %dst, %idx, %tag, %idx)
+ "memref.dma_start"(%src, %idx, %flt, %dst, %idx, %tag, %idx)
: (memref<2x2xf32>, index, f32, memref<2xf32,1>, index, memref<i32,2>, index) -> ()
}
// -----
func @dma_no_dst_memref(%m : f32, %tag : f32, %c0 : index) {
- %mref = alloc() : memref<8 x f32>
+ %mref = memref.alloc() : memref<8 x f32>
// expected-error at +1 {{expected destination to be of memref type}}
- dma_start %mref[%c0], %m[%c0], %c0, %tag[%c0] : memref<8 x f32>, f32, f32
+ memref.dma_start %mref[%c0], %m[%c0], %c0, %tag[%c0] : memref<8 x f32>, f32, f32
}
// -----
@@ -333,7 +334,7 @@ func @dma_start_not_enough_operands_for_dst(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>) {
// expected-error at +1 {{expected at least 7 operands}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %idx)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %idx)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, index) -> ()
}
@@ -343,7 +344,7 @@ func @dma_start_dst_index_wrong_type(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>, %flt: f32) {
// expected-error at +1 {{expected destination indices to be of index type}}
- "std.dma_start"(%src, %idx, %idx, %dst, %flt, %tag, %idx)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %flt, %tag, %idx)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, f32, memref<i32,2>, index) -> ()
}
@@ -353,16 +354,16 @@ func @dma_start_dst_index_wrong_type(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>, %flt: f32) {
// expected-error at +1 {{expected num elements to be of index type}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %flt, %tag)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %flt, %tag)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, f32, memref<i32,2>) -> ()
}
// -----
func @dma_no_tag_memref(%tag : f32, %c0 : index) {
- %mref = alloc() : memref<8 x f32>
+ %mref = memref.alloc() : memref<8 x f32>
// expected-error at +1 {{expected tag to be of memref type}}
- dma_start %mref[%c0], %mref[%c0], %c0, %tag[%c0] : memref<8 x f32>, memref<8 x f32>, f32
+ memref.dma_start %mref[%c0], %mref[%c0], %c0, %tag[%c0] : memref<8 x f32>, memref<8 x f32>, f32
}
// -----
@@ -371,7 +372,7 @@ func @dma_start_not_enough_operands_for_tag(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<2xi32,2>) {
// expected-error at +1 {{expected at least 8 operands}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, index, memref<2xi32,2>) -> ()
}
@@ -381,7 +382,7 @@ func @dma_start_dst_index_wrong_type(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<2xi32,2>, %flt: f32) {
// expected-error at +1 {{expected tag indices to be of index type}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %flt)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %flt)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, index, memref<2xi32,2>, f32) -> ()
}
@@ -391,7 +392,7 @@ func @dma_start_same_space(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32>,
%tag: memref<i32,2>) {
// expected-error at +1 {{DMA should be between
diff erent memory spaces}}
- dma_start %src[%idx, %idx], %dst[%idx], %idx, %tag[] : memref<2x2xf32>, memref<2xf32>, memref<i32,2>
+ memref.dma_start %src[%idx, %idx], %dst[%idx], %idx, %tag[] : memref<2x2xf32>, memref<2xf32>, memref<i32,2>
}
// -----
@@ -400,7 +401,7 @@ func @dma_start_too_many_operands(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>) {
// expected-error at +1 {{incorrect number of operands}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %idx, %idx, %idx)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %idx, %idx, %idx)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, index, memref<i32,2>, index, index, index) -> ()
}
@@ -411,7 +412,7 @@ func @dma_start_wrong_stride_type(
%src: memref<2x2xf32>, %idx: index, %dst: memref<2xf32,1>,
%tag: memref<i32,2>, %flt: f32) {
// expected-error at +1 {{expected stride and num elements per stride to be of type index}}
- "std.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %idx, %flt)
+ "memref.dma_start"(%src, %idx, %idx, %dst, %idx, %idx, %tag, %idx, %flt)
: (memref<2x2xf32>, index, index, memref<2xf32,1>, index, index, memref<i32,2>, index, f32) -> ()
}
@@ -419,28 +420,28 @@ func @dma_start_wrong_stride_type(
func @dma_wait_not_enough_operands() {
// expected-error at +1 {{expected at least 2 operands}}
- "std.dma_wait"() : () -> ()
+ "memref.dma_wait"() : () -> ()
}
// -----
func @dma_wait_no_tag_memref(%tag : f32, %c0 : index) {
// expected-error at +1 {{expected tag to be of memref type}}
- "std.dma_wait"(%tag, %c0, %c0) : (f32, index, index) -> ()
+ "memref.dma_wait"(%tag, %c0, %c0) : (f32, index, index) -> ()
}
// -----
func @dma_wait_wrong_index_type(%tag : memref<2xi32>, %idx: index, %flt: f32) {
// expected-error at +1 {{expected tag indices to be of index type}}
- "std.dma_wait"(%tag, %flt, %idx) : (memref<2xi32>, f32, index) -> ()
+ "memref.dma_wait"(%tag, %flt, %idx) : (memref<2xi32>, f32, index) -> ()
}
// -----
func @dma_wait_wrong_num_elements_type(%tag : memref<2xi32>, %idx: index, %flt: f32) {
// expected-error at +1 {{expected the number of elements to be of index type}}
- "std.dma_wait"(%tag, %idx, %flt) : (memref<2xi32>, index, f32) -> ()
+ "memref.dma_wait"(%tag, %idx, %flt) : (memref<2xi32>, index, f32) -> ()
}
// -----
@@ -856,9 +857,9 @@ func @invalid_splat(%v : f32) { // expected-note {{prior use here}}
// -----
func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
// expected-error at +1 {{expects 1 offset operand}}
- %1 = view %0[][%arg0, %arg1]
+ %1 = memref.view %0[][%arg0, %arg1]
: memref<2048xi8> to memref<?x?xf32>
return
}
@@ -866,9 +867,9 @@ func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8, affine_map<(d0) -> (d0 floordiv 8, d0 mod 8)>>
+ %0 = memref.alloc() : memref<2048xi8, affine_map<(d0) -> (d0 floordiv 8, d0 mod 8)>>
// expected-error at +1 {{unsupported map for base memref type}}
- %1 = view %0[%arg2][%arg0, %arg1]
+ %1 = memref.view %0[%arg2][%arg0, %arg1]
: memref<2048xi8, affine_map<(d0) -> (d0 floordiv 8, d0 mod 8)>> to
memref<?x?xf32, affine_map<(d0, d1)[s0] -> (d0 * 4 + d1 + s0)>>
return
@@ -877,9 +878,9 @@ func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
// expected-error at +1 {{unsupported map for result memref type}}
- %1 = view %0[%arg2][%arg0, %arg1]
+ %1 = memref.view %0[%arg2][%arg0, %arg1]
: memref<2048xi8> to memref<?x?xf32, affine_map<(d0, d1)[s0] -> (d0, d1, s0)>>
return
}
@@ -887,18 +888,18 @@ func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8, 2>
+ %0 = memref.alloc() : memref<2048xi8, 2>
// expected-error at +1 {{
diff erent memory spaces}}
- %1 = view %0[%arg2][%arg0, %arg1] : memref<2048xi8, 2> to memref<?x?xf32, 1>
+ %1 = memref.view %0[%arg2][%arg0, %arg1] : memref<2048xi8, 2> to memref<?x?xf32, 1>
return
}
// -----
func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
// expected-error at +1 {{incorrect number of size operands for type}}
- %1 = view %0[%arg2][%arg0]
+ %1 = memref.view %0[%arg2][%arg0]
: memref<2048xi8> to memref<?x?xf32>
return
}
@@ -906,9 +907,9 @@ func @invalid_view(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected mixed offsets rank to match mixed sizes rank (2 vs 3) so the rank of the result type is well-formed}}
- %1 = subview %0[0, 0][2, 2, 2][1, 1, 1]
+ %1 = memref.subview %0[0, 0][2, 2, 2][1, 1, 1]
: memref<8x16x4xf32> to memref<8x16x4xf32>
return
}
@@ -916,9 +917,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected mixed sizes rank to match mixed strides rank (3 vs 2) so the rank of the result type is well-formed}}
- %1 = subview %0[0, 0, 0][2, 2, 2][1, 1]
+ %1 = memref.subview %0[0, 0, 0][2, 2, 2][1, 1]
: memref<8x16x4xf32> to memref<8x16x4xf32>
return
}
@@ -926,9 +927,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected mixed sizes rank to match mixed strides rank (3 vs 2) so the rank of the result type is well-formed}}
- %1 = memref_reinterpret_cast %0 to offset: [0], sizes: [2, 2, 2], strides:[1, 1]
+ %1 = memref.reinterpret_cast %0 to offset: [0], sizes: [2, 2, 2], strides:[1, 1]
: memref<8x16x4xf32> to memref<8x16x4xf32>
return
}
@@ -936,9 +937,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32, offset: 0, strides: [64, 4, 1], 2>
+ %0 = memref.alloc() : memref<8x16x4xf32, offset: 0, strides: [64, 4, 1], 2>
// expected-error at +1 {{
diff erent memory spaces}}
- %1 = subview %0[0, 0, 0][%arg2, %arg2, %arg2][1, 1, 1]
+ %1 = memref.subview %0[0, 0, 0][%arg2, %arg2, %arg2][1, 1, 1]
: memref<8x16x4xf32, offset: 0, strides: [64, 4, 1], 2> to
memref<8x?x4xf32, affine_map<(d0, d1, d2)[s0] -> (d0 * s0 + d1 * 4 + d2)>>
return
@@ -947,9 +948,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 + d1, d1 + d2, d2)>>
+ %0 = memref.alloc() : memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 + d1, d1 + d2, d2)>>
// expected-error at +1 {{is not strided}}
- %1 = subview %0[0, 0, 0][%arg2, %arg2, %arg2][1, 1, 1]
+ %1 = memref.subview %0[0, 0, 0][%arg2, %arg2, %arg2][1, 1, 1]
: memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 + d1, d1 + d2, d2)>> to
memref<8x?x4xf32, offset: 0, strides: [?, 4, 1]>
return
@@ -958,9 +959,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected <= 3 offset values}}
- %1 = subview %0[%arg0, %arg1, 0, 0][%arg2, 0, 0, 0][1, 1, 1, 1]
+ %1 = memref.subview %0[%arg0, %arg1, 0, 0][%arg2, 0, 0, 0][1, 1, 1, 1]
: memref<8x16x4xf32> to
memref<8x?x4xf32, offset: 0, strides:[?, ?, 4]>
return
@@ -969,9 +970,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected result type to be 'memref<?x?x?xf32, affine_map<(d0, d1, d2)[s0, s1, s2, s3] -> (d0 * s1 + s0 + d1 * s2 + d2 * s3)>>' or a rank-reduced version. (mismatch of result affine map)}}
- %1 = subview %0[%arg0, %arg1, %arg2][%arg0, %arg1, %arg2][%arg0, %arg1, %arg2]
+ %1 = memref.subview %0[%arg0, %arg1, %arg2][%arg0, %arg1, %arg2][%arg0, %arg1, %arg2]
: memref<8x16x4xf32> to
memref<?x?x?xf32, offset: ?, strides: [64, 4, 1]>
return
@@ -980,9 +981,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected result element type to be 'f32'}}
- %1 = subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
+ %1 = memref.subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
: memref<8x16x4xf32> to
memref<8x16x4xi32>
return
@@ -991,9 +992,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected result rank to be smaller or equal to the source rank.}}
- %1 = subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
+ %1 = memref.subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
: memref<8x16x4xf32> to
memref<8x16x4x3xi32>
return
@@ -1002,9 +1003,9 @@ func @invalid_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
// -----
func @invalid_rank_reducing_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected result type to be 'memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2)>>' or a rank-reduced version. (mismatch of result sizes)}}
- %1 = subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
+ %1 = memref.subview %0[0, 0, 0][8, 16, 4][1, 1, 1]
: memref<8x16x4xf32> to memref<16x4xf32>
return
}
@@ -1012,9 +1013,9 @@ func @invalid_rank_reducing_subview(%arg0 : index, %arg1 : index, %arg2 : index)
// -----
func @invalid_rank_reducing_subview(%arg0 : index, %arg1 : index, %arg2 : index) {
- %0 = alloc() : memref<8x16x4xf32>
+ %0 = memref.alloc() : memref<8x16x4xf32>
// expected-error at +1 {{expected result type to be 'memref<8x16x4xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 4 + d2 + 8)>>' or a rank-reduced version. (mismatch of result sizes)}}
- %1 = subview %0[0, 2, 0][8, 16, 4][1, 1, 1]
+ %1 = memref.subview %0[0, 2, 0][8, 16, 4][1, 1, 1]
: memref<8x16x4xf32> to memref<16x4xf32>
return
}
@@ -1023,7 +1024,7 @@ func @invalid_rank_reducing_subview(%arg0 : index, %arg1 : index, %arg2 : index)
func @invalid_rank_reducing_subview(%arg0 : memref<?x?xf32>, %arg1 : index, %arg2 : index) {
// expected-error at +1 {{expected result type to be 'memref<?x1xf32, affine_map<(d0, d1)[s0, s1] -> (d0 * s1 + s0 + d1)>>' or a rank-reduced version. (mismatch of result affine map)}}
- %0 = subview %arg0[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32>
+ %0 = memref.subview %arg0[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32>
return
}
@@ -1032,7 +1033,7 @@ func @invalid_rank_reducing_subview(%arg0 : memref<?x?xf32>, %arg1 : index, %arg
// The affine map affine_map<(d0)[s0, s1, s2] -> (d0 * s1 + s0)> has an extra unused symbol.
func @invalid_rank_reducing_subview(%arg0 : memref<?x?xf32>, %arg1 : index, %arg2 : index) {
// expected-error at +1 {{expected result type to be 'memref<?x1xf32, affine_map<(d0, d1)[s0, s1] -> (d0 * s1 + s0 + d1)>>' or a rank-reduced version. (mismatch of result affine map) inferred type: (d0)[s0, s1] -> (d0 * s1 + s0)}}
- %0 = subview %arg0[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, affine_map<(d0)[s0, s1, s2] -> (d0 * s1 + s0)>>
+ %0 = memref.subview %arg0[0, %arg1][%arg2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, affine_map<(d0)[s0, s1, s2] -> (d0 * s1 + s0)>>
return
}
@@ -1040,7 +1041,7 @@ func @invalid_rank_reducing_subview(%arg0 : memref<?x?xf32>, %arg1 : index, %arg
func @invalid_memref_cast(%arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]>) {
// expected-error at +1{{operand type 'memref<12x4x16xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 16 + d2)>>' and result type 'memref<12x4x16xf32, affine_map<(d0, d1, d2) -> (d0 * 128 + d1 * 32 + d2 * 2)>>' are cast incompatible}}
- %0 = memref_cast %arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]> to memref<12x4x16xf32, offset:0, strides:[128, 32, 2]>
+ %0 = memref.cast %arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]> to memref<12x4x16xf32, offset:0, strides:[128, 32, 2]>
return
}
@@ -1048,7 +1049,7 @@ func @invalid_memref_cast(%arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16,
func @invalid_memref_cast(%arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]>) {
// expected-error at +1{{operand type 'memref<12x4x16xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 16 + d2)>>' and result type 'memref<12x4x16xf32, affine_map<(d0, d1, d2) -> (d0 * 64 + d1 * 16 + d2 + 16)>>' are cast incompatible}}
- %0 = memref_cast %arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]> to memref<12x4x16xf32, offset:16, strides:[64, 16, 1]>
+ %0 = memref.cast %arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16, 1]> to memref<12x4x16xf32, offset:16, strides:[64, 16, 1]>
return
}
@@ -1056,36 +1057,36 @@ func @invalid_memref_cast(%arg0 : memref<12x4x16xf32, offset:0, strides:[64, 16,
// incompatible element types
func @invalid_memref_cast() {
- %0 = alloc() : memref<2x5xf32, 0>
+ %0 = memref.alloc() : memref<2x5xf32, 0>
// expected-error at +1 {{operand type 'memref<2x5xf32>' and result type 'memref<*xi32>' are cast incompatible}}
- %1 = memref_cast %0 : memref<2x5xf32, 0> to memref<*xi32>
+ %1 = memref.cast %0 : memref<2x5xf32, 0> to memref<*xi32>
return
}
// -----
func @invalid_prefetch_rw(%i : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
// expected-error at +1 {{rw specifier has to be 'read' or 'write'}}
- prefetch %0[%i], rw, locality<0>, data : memref<10xf32>
+ memref.prefetch %0[%i], rw, locality<0>, data : memref<10xf32>
return
}
// -----
func @invalid_prefetch_cache_type(%i : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
// expected-error at +1 {{cache type has to be 'data' or 'instr'}}
- prefetch %0[%i], read, locality<0>, false : memref<10xf32>
+ memref.prefetch %0[%i], read, locality<0>, false : memref<10xf32>
return
}
// -----
func @invalid_prefetch_locality_hint(%i : index) {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
// expected-error at +1 {{32-bit signless integer attribute whose minimum value is 0 whose maximum value is 3}}
- prefetch %0[%i], read, locality<5>, data : memref<10xf32>
+ memref.prefetch %0[%i], read, locality<5>, data : memref<10xf32>
return
}
@@ -1093,9 +1094,9 @@ func @invalid_prefetch_locality_hint(%i : index) {
// incompatible memory space
func @invalid_memref_cast() {
- %0 = alloc() : memref<2x5xf32, 0>
+ %0 = memref.alloc() : memref<2x5xf32, 0>
// expected-error at +1 {{operand type 'memref<2x5xf32>' and result type 'memref<*xf32, 1>' are cast incompatible}}
- %1 = memref_cast %0 : memref<2x5xf32, 0> to memref<*xf32, 1>
+ %1 = memref.cast %0 : memref<2x5xf32, 0> to memref<*xf32, 1>
return
}
@@ -1103,10 +1104,10 @@ func @invalid_memref_cast() {
// unranked to unranked
func @invalid_memref_cast() {
- %0 = alloc() : memref<2x5xf32, 0>
- %1 = memref_cast %0 : memref<2x5xf32, 0> to memref<*xf32, 0>
+ %0 = memref.alloc() : memref<2x5xf32, 0>
+ %1 = memref.cast %0 : memref<2x5xf32, 0> to memref<*xf32, 0>
// expected-error at +1 {{operand type 'memref<*xf32>' and result type 'memref<*xf32>' are cast incompatible}}
- %2 = memref_cast %1 : memref<*xf32, 0> to memref<*xf32, 0>
+ %2 = memref.cast %1 : memref<*xf32, 0> to memref<*xf32, 0>
return
}
@@ -1177,7 +1178,7 @@ func @generic_atomic_rmw_has_side_effects(%I: memref<10xf32>, %i : index) {
%x = generic_atomic_rmw %I[%i] : memref<10xf32> {
^bb0(%old_value : f32):
%c1 = constant 1.0 : f32
- %buf = alloc() : memref<2048xf32>
+ %buf = memref.alloc() : memref<2048xf32>
atomic_yield %c1 : f32
}
}
@@ -1199,7 +1200,7 @@ func @atomic_yield_type_mismatch(%I: memref<10xf32>, %i : index) {
// alignment is not power of 2.
func @assume_alignment(%0: memref<4x4xf16>) {
// expected-error at +1 {{alignment must be power of 2}}
- std.assume_alignment %0, 12 : memref<4x4xf16>
+ memref.assume_alignment %0, 12 : memref<4x4xf16>
return
}
@@ -1208,14 +1209,14 @@ func @assume_alignment(%0: memref<4x4xf16>) {
// 0 alignment value.
func @assume_alignment(%0: memref<4x4xf16>) {
// expected-error at +1 {{attribute 'alignment' failed to satisfy constraint: 32-bit signless integer attribute whose value is positive}}
- std.assume_alignment %0, 0 : memref<4x4xf16>
+ memref.assume_alignment %0, 0 : memref<4x4xf16>
return
}
// -----
"alloca_without_scoped_alloc_parent"() ( {
- std.alloca() : memref<1xf32>
+ memref.alloca() : memref<1xf32>
// expected-error at -1 {{requires an ancestor op with AutomaticAllocationScope trait}}
return
}) : () -> ()
diff --git a/mlir/test/IR/invalid.mlir b/mlir/test/IR/invalid.mlir
index 5751cf404606..419c98626521 100644
--- a/mlir/test/IR/invalid.mlir
+++ b/mlir/test/IR/invalid.mlir
@@ -821,7 +821,7 @@ func @f(f32) {
func @f(%m : memref<?x?xf32>) {
affine.for %i0 = 0 to 42 {
// expected-note at +1 {{previously referenced here}}
- %x = load %m[%i0, %i1] : memref<?x?xf32>
+ %x = memref.load %m[%i0, %i1] : memref<?x?xf32>
}
// expected-error at +1 {{region entry argument '%i1' is already in use}}
affine.for %i1 = 0 to 42 {
diff --git a/mlir/test/IR/memory-ops.mlir b/mlir/test/IR/memory-ops.mlir
index d1860acb5076..000f9a6cae52 100644
--- a/mlir/test/IR/memory-ops.mlir
+++ b/mlir/test/IR/memory-ops.mlir
@@ -6,28 +6,28 @@
func @alloc() {
^bb0:
// Test simple alloc.
- // CHECK: %0 = alloc() : memref<1024x64xf32, 1>
- %0 = alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %0 = memref.alloc() : memref<1024x64xf32, 1>
+ %0 = memref.alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
%c0 = "std.constant"() {value = 0: index} : () -> index
%c1 = "std.constant"() {value = 1: index} : () -> index
// Test alloc with dynamic dimensions.
- // CHECK: %1 = alloc(%c0, %c1) : memref<?x?xf32, 1>
- %1 = alloc(%c0, %c1) : memref<?x?xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %1 = memref.alloc(%c0, %c1) : memref<?x?xf32, 1>
+ %1 = memref.alloc(%c0, %c1) : memref<?x?xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
// Test alloc with no dynamic dimensions and one symbol.
- // CHECK: %2 = alloc()[%c0] : memref<2x4xf32, #map, 1>
- %2 = alloc()[%c0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ // CHECK: %2 = memref.alloc()[%c0] : memref<2x4xf32, #map, 1>
+ %2 = memref.alloc()[%c0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
// Test alloc with dynamic dimensions and one symbol.
- // CHECK: %3 = alloc(%c1)[%c0] : memref<2x?xf32, #map, 1>
- %3 = alloc(%c1)[%c0] : memref<2x?xf32, affine_map<(d0, d1)[s0] -> (d0 + s0, d1)>, 1>
+ // CHECK: %3 = memref.alloc(%c1)[%c0] : memref<2x?xf32, #map, 1>
+ %3 = memref.alloc(%c1)[%c0] : memref<2x?xf32, affine_map<(d0, d1)[s0] -> (d0 + s0, d1)>, 1>
// Alloc with no mappings.
// b/116054838 Parser crash while parsing ill-formed AllocOp
- // CHECK: %4 = alloc() : memref<2xi32>
- %4 = alloc() : memref<2 x i32>
+ // CHECK: %4 = memref.alloc() : memref<2xi32>
+ %4 = memref.alloc() : memref<2 x i32>
// CHECK: return
return
@@ -37,27 +37,27 @@ func @alloc() {
func @alloca() {
^bb0:
// Test simple alloc.
- // CHECK: %0 = alloca() : memref<1024x64xf32, 1>
- %0 = alloca() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %0 = memref.alloca() : memref<1024x64xf32, 1>
+ %0 = memref.alloca() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
%c0 = "std.constant"() {value = 0: index} : () -> index
%c1 = "std.constant"() {value = 1: index} : () -> index
// Test alloca with dynamic dimensions.
- // CHECK: %1 = alloca(%c0, %c1) : memref<?x?xf32, 1>
- %1 = alloca(%c0, %c1) : memref<?x?xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %1 = memref.alloca(%c0, %c1) : memref<?x?xf32, 1>
+ %1 = memref.alloca(%c0, %c1) : memref<?x?xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
// Test alloca with no dynamic dimensions and one symbol.
- // CHECK: %2 = alloca()[%c0] : memref<2x4xf32, #map, 1>
- %2 = alloca()[%c0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
+ // CHECK: %2 = memref.alloca()[%c0] : memref<2x4xf32, #map, 1>
+ %2 = memref.alloca()[%c0] : memref<2x4xf32, affine_map<(d0, d1)[s0] -> ((d0 + s0), d1)>, 1>
// Test alloca with dynamic dimensions and one symbol.
- // CHECK: %3 = alloca(%c1)[%c0] : memref<2x?xf32, #map, 1>
- %3 = alloca(%c1)[%c0] : memref<2x?xf32, affine_map<(d0, d1)[s0] -> (d0 + s0, d1)>, 1>
+ // CHECK: %3 = memref.alloca(%c1)[%c0] : memref<2x?xf32, #map, 1>
+ %3 = memref.alloca(%c1)[%c0] : memref<2x?xf32, affine_map<(d0, d1)[s0] -> (d0 + s0, d1)>, 1>
// Alloca with no mappings, but with alignment.
- // CHECK: %4 = alloca() {alignment = 64 : i64} : memref<2xi32>
- %4 = alloca() {alignment = 64} : memref<2 x i32>
+ // CHECK: %4 = memref.alloca() {alignment = 64 : i64} : memref<2xi32>
+ %4 = memref.alloca() {alignment = 64} : memref<2 x i32>
return
}
@@ -65,28 +65,28 @@ func @alloca() {
// CHECK-LABEL: func @dealloc() {
func @dealloc() {
^bb0:
- // CHECK: %0 = alloc() : memref<1024x64xf32>
- %0 = alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 0>
+ // CHECK: %0 = memref.alloc() : memref<1024x64xf32>
+ %0 = memref.alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 0>
- // CHECK: dealloc %0 : memref<1024x64xf32>
- dealloc %0 : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 0>
+ // CHECK: memref.dealloc %0 : memref<1024x64xf32>
+ memref.dealloc %0 : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 0>
return
}
// CHECK-LABEL: func @load_store
func @load_store() {
^bb0:
- // CHECK: %0 = alloc() : memref<1024x64xf32, 1>
- %0 = alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %0 = memref.alloc() : memref<1024x64xf32, 1>
+ %0 = memref.alloc() : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
%1 = constant 0 : index
%2 = constant 1 : index
- // CHECK: %1 = load %0[%c0, %c1] : memref<1024x64xf32, 1>
- %3 = load %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: %1 = memref.load %0[%c0, %c1] : memref<1024x64xf32, 1>
+ %3 = memref.load %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
- // CHECK: store %1, %0[%c0, %c1] : memref<1024x64xf32, 1>
- store %3, %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
+ // CHECK: memref.store %1, %0[%c0, %c1] : memref<1024x64xf32, 1>
+ memref.store %3, %0[%1, %2] : memref<1024x64xf32, affine_map<(d0, d1) -> (d0, d1)>, 1>
return
}
@@ -97,20 +97,20 @@ func @dma_ops() {
%stride = constant 32 : index
%elt_per_stride = constant 16 : index
- %A = alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
- %Ah = alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 1>
- %tag = alloc() : memref<1 x f32>
+ %A = memref.alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
+ %Ah = memref.alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 1>
+ %tag = memref.alloc() : memref<1 x f32>
%num_elements = constant 256 : index
- dma_start %A[%c0], %Ah[%c0], %num_elements, %tag[%c0] : memref<256 x f32>, memref<256 x f32, 1>, memref<1 x f32>
- dma_wait %tag[%c0], %num_elements : memref<1 x f32>
+ memref.dma_start %A[%c0], %Ah[%c0], %num_elements, %tag[%c0] : memref<256 x f32>, memref<256 x f32, 1>, memref<1 x f32>
+ memref.dma_wait %tag[%c0], %num_elements : memref<1 x f32>
// CHECK: dma_start %0[%c0], %1[%c0], %c256, %2[%c0] : memref<256xf32>, memref<256xf32, 1>, memref<1xf32>
// CHECK-NEXT: dma_wait %2[%c0], %c256 : memref<1xf32>
// DMA with strides
- dma_start %A[%c0], %Ah[%c0], %num_elements, %tag[%c0], %stride, %elt_per_stride : memref<256 x f32>, memref<256 x f32, 1>, memref<1 x f32>
- dma_wait %tag[%c0], %num_elements : memref<1 x f32>
+ memref.dma_start %A[%c0], %Ah[%c0], %num_elements, %tag[%c0], %stride, %elt_per_stride : memref<256 x f32>, memref<256 x f32, 1>, memref<1 x f32>
+ memref.dma_wait %tag[%c0], %num_elements : memref<1 x f32>
// CHECK-NEXT: dma_start %0[%c0], %1[%c0], %c256, %2[%c0], %c32, %c16 : memref<256xf32>, memref<256xf32, 1>, memref<1xf32>
// CHECK-NEXT: dma_wait %2[%c0], %c256 : memref<1xf32>
diff --git a/mlir/test/IR/parser.mlir b/mlir/test/IR/parser.mlir
index eba9a71c1a95..d24f5f07637e 100644
--- a/mlir/test/IR/parser.mlir
+++ b/mlir/test/IR/parser.mlir
@@ -311,7 +311,7 @@ func @triang_loop(%arg0: index, %arg1: memref<?x?xi32>) {
%c = constant 0 : i32 // CHECK: %{{.*}} = constant 0 : i32
affine.for %i0 = 1 to %arg0 { // CHECK: affine.for %{{.*}} = 1 to %{{.*}} {
affine.for %i1 = affine_map<(d0)[]->(d0)>(%i0)[] to %arg0 { // CHECK: affine.for %{{.*}} = #map{{[0-9]+}}(%{{.*}}) to %{{.*}} {
- store %c, %arg1[%i0, %i1] : memref<?x?xi32> // CHECK: store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}]
+ memref.store %c, %arg1[%i0, %i1] : memref<?x?xi32> // CHECK: memref.store %{{.*}}, %{{.*}}[%{{.*}}, %{{.*}}]
} // CHECK: }
} // CHECK: }
return // CHECK: return
diff --git a/mlir/test/IR/slice.mlir b/mlir/test/IR/slice.mlir
index 68ddeb6ad9d0..9ce43fdb10a3 100644
--- a/mlir/test/IR/slice.mlir
+++ b/mlir/test/IR/slice.mlir
@@ -1,18 +1,18 @@
// RUN: mlir-opt -slice-analysis-test %s | FileCheck %s
func @slicing_linalg_op(%arg0 : index, %arg1 : index, %arg2 : index) {
- %a = alloc(%arg0, %arg2) : memref<?x?xf32>
- %b = alloc(%arg2, %arg1) : memref<?x?xf32>
- %c = alloc(%arg0, %arg1) : memref<?x?xf32>
- %d = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %a = memref.alloc(%arg0, %arg2) : memref<?x?xf32>
+ %b = memref.alloc(%arg2, %arg1) : memref<?x?xf32>
+ %c = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %d = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
linalg.matmul ins(%a, %b : memref<?x?xf32>, memref<?x?xf32>)
outs(%c : memref<?x?xf32>)
linalg.matmul ins(%a, %b : memref<?x?xf32>, memref<?x?xf32>)
outs(%d : memref<?x?xf32>)
- dealloc %c : memref<?x?xf32>
- dealloc %b : memref<?x?xf32>
- dealloc %a : memref<?x?xf32>
- dealloc %d : memref<?x?xf32>
+ memref.dealloc %c : memref<?x?xf32>
+ memref.dealloc %b : memref<?x?xf32>
+ memref.dealloc %a : memref<?x?xf32>
+ memref.dealloc %d : memref<?x?xf32>
return
}
@@ -20,16 +20,16 @@ func @slicing_linalg_op(%arg0 : index, %arg1 : index, %arg2 : index) {
// CHECK-SAME: %[[ARG0:[a-zA-Z0-9_]+]]: index
// CHECK-SAME: %[[ARG1:[a-zA-Z0-9_]+]]: index
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: index
-// CHECK-DAG: %[[A:.+]] = alloc(%[[ARG0]], %[[ARG2]]) : memref<?x?xf32>
-// CHECK-DAG: %[[B:.+]] = alloc(%[[ARG2]], %[[ARG1]]) : memref<?x?xf32>
-// CHECK-DAG: %[[C:.+]] = alloc(%[[ARG0]], %[[ARG1]]) : memref<?x?xf32>
+// CHECK-DAG: %[[A:.+]] = memref.alloc(%[[ARG0]], %[[ARG2]]) : memref<?x?xf32>
+// CHECK-DAG: %[[B:.+]] = memref.alloc(%[[ARG2]], %[[ARG1]]) : memref<?x?xf32>
+// CHECK-DAG: %[[C:.+]] = memref.alloc(%[[ARG0]], %[[ARG1]]) : memref<?x?xf32>
// CHECK: return
// CHECK-LABEL: func @slicing_linalg_op__backward_slice__1
// CHECK-SAME: %[[ARG0:[a-zA-Z0-9_]+]]: index
// CHECK-SAME: %[[ARG1:[a-zA-Z0-9_]+]]: index
// CHECK-SAME: %[[ARG2:[a-zA-Z0-9_]+]]: index
-// CHECK-DAG: %[[A:.+]] = alloc(%[[ARG0]], %[[ARG2]]) : memref<?x?xf32>
-// CHECK-DAG: %[[B:.+]] = alloc(%[[ARG2]], %[[ARG1]]) : memref<?x?xf32>
-// CHECK-DAG: %[[C:.+]] = alloc(%[[ARG0]], %[[ARG1]]) : memref<?x?xf32>
+// CHECK-DAG: %[[A:.+]] = memref.alloc(%[[ARG0]], %[[ARG2]]) : memref<?x?xf32>
+// CHECK-DAG: %[[B:.+]] = memref.alloc(%[[ARG2]], %[[ARG1]]) : memref<?x?xf32>
+// CHECK-DAG: %[[C:.+]] = memref.alloc(%[[ARG0]], %[[ARG1]]) : memref<?x?xf32>
// CHECK: return
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
index 753e74541961..0e29209c4fab 100644
--- 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
@@ -60,39 +60,39 @@ func @entry() {
// Sanity check for the function under test.
//
- %LHS10 = alloc() {alignment = 64} : memref<1x10xf32>
- %RHS10 = alloc() {alignment = 64} : memref<1x10xf32>
- %DST10 = alloc() {alignment = 64} : memref<1x10xf32>
+ %LHS10 = memref.alloc() {alignment = 64} : memref<1x10xf32>
+ %RHS10 = memref.alloc() {alignment = 64} : memref<1x10xf32>
+ %DST10 = memref.alloc() {alignment = 64} : memref<1x10xf32>
linalg.fill(%LHS10, %f1) : memref<1x10xf32>, f32
linalg.fill(%RHS10, %f1) : memref<1x10xf32>, f32
- %LHS = memref_cast %LHS10 : memref<1x10xf32> to memref<?x?xf32>
- %RHS = memref_cast %RHS10 : memref<1x10xf32> to memref<?x?xf32>
- %DST = memref_cast %DST10 : memref<1x10xf32> to memref<?x?xf32>
+ %LHS = memref.cast %LHS10 : memref<1x10xf32> to memref<?x?xf32>
+ %RHS = memref.cast %RHS10 : memref<1x10xf32> to memref<?x?xf32>
+ %DST = memref.cast %DST10 : memref<1x10xf32> to memref<?x?xf32>
call @linalg_generic(%LHS, %RHS, %DST)
: (memref<?x?xf32>, memref<?x?xf32>, memref<?x?xf32>) -> ()
// CHECK: [2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
- %U = memref_cast %DST10 : memref<1x10xf32> to memref<*xf32>
+ %U = memref.cast %DST10 : memref<1x10xf32> to memref<*xf32>
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %LHS10: memref<1x10xf32>
- dealloc %RHS10: memref<1x10xf32>
- dealloc %DST10: memref<1x10xf32>
+ memref.dealloc %LHS10: memref<1x10xf32>
+ memref.dealloc %RHS10: memref<1x10xf32>
+ memref.dealloc %DST10: memref<1x10xf32>
//
// Allocate data for microbenchmarks.
//
- %LHS1024 = alloc() {alignment = 64} : memref<1024x1024xf32>
- %RHS1024 = alloc() {alignment = 64} : memref<1024x1024xf32>
- %DST1024 = alloc() {alignment = 64} : memref<1024x1024xf32>
+ %LHS1024 = memref.alloc() {alignment = 64} : memref<1024x1024xf32>
+ %RHS1024 = memref.alloc() {alignment = 64} : memref<1024x1024xf32>
+ %DST1024 = memref.alloc() {alignment = 64} : memref<1024x1024xf32>
- %LHS0 = memref_cast %LHS1024 : memref<1024x1024xf32> to memref<?x?xf32>
- %RHS0 = memref_cast %RHS1024 : memref<1024x1024xf32> to memref<?x?xf32>
- %DST0 = memref_cast %DST1024 : memref<1024x1024xf32> to memref<?x?xf32>
+ %LHS0 = memref.cast %LHS1024 : memref<1024x1024xf32> to memref<?x?xf32>
+ %RHS0 = memref.cast %RHS1024 : memref<1024x1024xf32> to memref<?x?xf32>
+ %DST0 = memref.cast %DST1024 : memref<1024x1024xf32> to memref<?x?xf32>
//
// Warm up.
@@ -117,9 +117,9 @@ func @entry() {
vector.print %t1024 : f64
// Free.
- dealloc %LHS1024: memref<1024x1024xf32>
- dealloc %RHS1024: memref<1024x1024xf32>
- dealloc %DST1024: memref<1024x1024xf32>
+ memref.dealloc %LHS1024: memref<1024x1024xf32>
+ memref.dealloc %RHS1024: memref<1024x1024xf32>
+ memref.dealloc %DST1024: memref<1024x1024xf32>
return
}
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
index adb061a01cce..e4d19bf5c2f9 100644
--- 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
@@ -19,33 +19,33 @@ func @entry() {
%lb = constant 0 : index
%ub = constant 9 : index
- %A = alloc() : memref<9xf32>
- %U = memref_cast %A : memref<9xf32> to memref<*xf32>
+ %A = memref.alloc() : memref<9xf32>
+ %U = memref.cast %A : memref<9xf32> to memref<*xf32>
// 1. %i = (0) to (9) step (1)
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
%0 = index_cast %i : index to i32
%1 = sitofp %0 : i32 to f32
- store %1, %A[%i] : memref<9xf32>
+ memref.store %1, %A[%i] : memref<9xf32>
}
// CHECK: [0, 1, 2, 3, 4, 5, 6, 7, 8]
call @print_memref_f32(%U): (memref<*xf32>) -> ()
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
- store %c0, %A[%i] : memref<9xf32>
+ memref.store %c0, %A[%i] : memref<9xf32>
}
// 2. %i = (0) to (9) step (2)
scf.parallel (%i) = (%lb) to (%ub) step (%c2) {
%0 = index_cast %i : index to i32
%1 = sitofp %0 : i32 to f32
- store %1, %A[%i] : memref<9xf32>
+ memref.store %1, %A[%i] : memref<9xf32>
}
// CHECK: [0, 0, 2, 0, 4, 0, 6, 0, 8]
call @print_memref_f32(%U): (memref<*xf32>) -> ()
scf.parallel (%i) = (%lb) to (%ub) step (%c1) {
- store %c0, %A[%i] : memref<9xf32>
+ memref.store %c0, %A[%i] : memref<9xf32>
}
// 3. %i = (-20) to (-11) step (3)
@@ -56,12 +56,12 @@ func @entry() {
%1 = sitofp %0 : i32 to f32
%2 = constant 20 : index
%3 = addi %i, %2 : index
- store %1, %A[%3] : memref<9xf32>
+ memref.store %1, %A[%3] : memref<9xf32>
}
// CHECK: [-20, 0, 0, -17, 0, 0, -14, 0, 0]
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<9xf32>
+ memref.dealloc %A : memref<9xf32>
return
}
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
index fe5fa8c2c0a8..d0f688c61226 100644
--- 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
@@ -19,8 +19,8 @@ func @entry() {
%lb = constant 0 : index
%ub = constant 8 : index
- %A = alloc() : memref<8x8xf32>
- %U = memref_cast %A : memref<8x8xf32> to memref<*xf32>
+ %A = memref.alloc() : memref<8x8xf32>
+ %U = memref.cast %A : memref<8x8xf32> to memref<*xf32>
// 1. (%i, %i) = (0, 8) to (8, 8) step (1, 1)
scf.parallel (%i, %j) = (%lb, %lb) to (%ub, %ub) step (%c1, %c1) {
@@ -28,7 +28,7 @@ func @entry() {
%1 = addi %j, %0 : index
%2 = index_cast %1 : index to i32
%3 = sitofp %2 : i32 to f32
- store %3, %A[%i, %j] : memref<8x8xf32>
+ memref.store %3, %A[%i, %j] : memref<8x8xf32>
}
// CHECK: [0, 1, 2, 3, 4, 5, 6, 7]
@@ -42,7 +42,7 @@ func @entry() {
call @print_memref_f32(%U): (memref<*xf32>) -> ()
scf.parallel (%i, %j) = (%lb, %lb) to (%ub, %ub) step (%c1, %c1) {
- store %c0, %A[%i, %j] : memref<8x8xf32>
+ memref.store %c0, %A[%i, %j] : memref<8x8xf32>
}
// 2. (%i, %i) = (0, 8) to (8, 8) step (2, 1)
@@ -51,7 +51,7 @@ func @entry() {
%1 = addi %j, %0 : index
%2 = index_cast %1 : index to i32
%3 = sitofp %2 : i32 to f32
- store %3, %A[%i, %j] : memref<8x8xf32>
+ memref.store %3, %A[%i, %j] : memref<8x8xf32>
}
// CHECK: [0, 1, 2, 3, 4, 5, 6, 7]
@@ -65,7 +65,7 @@ func @entry() {
call @print_memref_f32(%U): (memref<*xf32>) -> ()
scf.parallel (%i, %j) = (%lb, %lb) to (%ub, %ub) step (%c1, %c1) {
- store %c0, %A[%i, %j] : memref<8x8xf32>
+ memref.store %c0, %A[%i, %j] : memref<8x8xf32>
}
// 3. (%i, %i) = (0, 8) to (8, 8) step (1, 2)
@@ -74,7 +74,7 @@ func @entry() {
%1 = addi %j, %0 : index
%2 = index_cast %1 : index to i32
%3 = sitofp %2 : i32 to f32
- store %3, %A[%i, %j] : memref<8x8xf32>
+ memref.store %3, %A[%i, %j] : memref<8x8xf32>
}
// CHECK: [0, 0, 2, 0, 4, 0, 6, 0]
@@ -87,7 +87,7 @@ func @entry() {
// CHECK-NEXT: [56, 0, 58, 0, 60, 0, 62, 0]
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<8x8xf32>
+ memref.dealloc %A : memref<8x8xf32>
return
}
diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir
index b935fc7122a0..3894cdbd41e4 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul.mlir
@@ -55,9 +55,9 @@ func @main() {
%v0 = constant 0.0 : !elem_type_a
%v1 = constant 1.0 : !elem_type_a
- %A = alloc() : !row_major_A
- %B = alloc() : !row_major_B
- %C = alloc() : !row_major_C
+ %A = memref.alloc() : !row_major_A
+ %B = memref.alloc() : !row_major_B
+ %C = memref.alloc() : !row_major_C
linalg.fill(%A, %v1) : !row_major_A, !elem_type_a
linalg.fill(%B, %v1) : !row_major_B, !elem_type_b
@@ -89,19 +89,19 @@ func @main() {
call @print_perf(%iters, %tmatmul) : (index, f64) -> ()
// CHECK: {{^0$}}
- %C_ref = alloc() : !row_major_C
+ %C_ref = memref.alloc() : !row_major_C
linalg.fill(%C_ref, %v0) : !row_major_C, !elem_type_c
linalg.matmul ins(%A, %B : !row_major_A, !row_major_B)
outs(%C_ref: !row_major_C)
- %act = memref_cast %C : !row_major_C to memref<*xf32>
- %exp = memref_cast %C_ref : !row_major_C to memref<*xf32>
+ %act = memref.cast %C : !row_major_C to memref<*xf32>
+ %exp = memref.cast %C_ref : !row_major_C to memref<*xf32>
%errors = call @verifyMemRefF32(%act, %exp) : (memref<*xf32>, memref<*xf32>) -> i64
vector.print %errors : i64
- dealloc %C_ref : !row_major_C
+ memref.dealloc %C_ref : !row_major_C
- dealloc %A : !row_major_A
- dealloc %B : !row_major_B
- dealloc %C : !row_major_C
+ memref.dealloc %A : !row_major_A
+ memref.dealloc %B : !row_major_B
+ memref.dealloc %C : !row_major_C
return
}
diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major.mlir
index 0672252b2a43..87340d0fff2f 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_column_major.mlir
@@ -59,9 +59,9 @@ func @main() {
%f0 = constant 0.0 : !elem_type_c
%f1 = constant 1.0 : !elem_type_a
- %cA = alloc() : !column_major_A
- %cB = alloc() : !column_major_B
- %cC = alloc() : !column_major_C
+ %cA = memref.alloc() : !column_major_A
+ %cB = memref.alloc() : !column_major_B
+ %cC = memref.alloc() : !column_major_C
linalg.fill(%cA, %f1) : !column_major_A, !elem_type_a
linalg.fill(%cB, %f1) : !column_major_B, !elem_type_b
@@ -86,19 +86,19 @@ func @main() {
call @print_perf(%iters, %tmatmul_column_major) : (index, f64) -> ()
// CHECK: {{^0$}}
- %cC_ref = alloc() : !column_major_C
+ %cC_ref = memref.alloc() : !column_major_C
linalg.fill(%cC_ref, %f0) : !column_major_C, !elem_type_c
linalg.matmul_column_major ins(%cA, %cB : !column_major_A, !column_major_B)
outs(%cC_ref: !column_major_C)
- %act = memref_cast %cC : !column_major_C to memref<*xf32>
- %exp = memref_cast %cC_ref : !column_major_C to memref<*xf32>
+ %act = memref.cast %cC : !column_major_C to memref<*xf32>
+ %exp = memref.cast %cC_ref : !column_major_C to memref<*xf32>
%errors = call @verifyMemRefF32(%act, %exp) : (memref<*xf32>, memref<*xf32>) -> i64
vector.print %errors : i64
- dealloc %cC_ref : !column_major_C
+ memref.dealloc %cC_ref : !column_major_C
- dealloc %cA : !column_major_A
- dealloc %cB : !column_major_B
- dealloc %cC : !column_major_C
+ memref.dealloc %cA : !column_major_A
+ memref.dealloc %cB : !column_major_B
+ memref.dealloc %cC : !column_major_C
return
}
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
index 89c18cc94796..24b8e656342e 100644
--- 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
@@ -65,9 +65,9 @@ func @main() {
%f0 = constant 0.0 : !elem_type_c
%f1 = constant 1.0 : !elem_type_a
- %cA = alloc() : !column_major_A
- %cB = alloc() : !column_major_B
- %cC = alloc() : !column_major_C
+ %cA = memref.alloc() : !column_major_A
+ %cB = memref.alloc() : !column_major_B
+ %cC = memref.alloc() : !column_major_C
linalg.fill(%cA, %f1) : !column_major_A, !elem_type_a
linalg.fill(%cB, %f1) : !column_major_B, !elem_type_b
@@ -78,9 +78,9 @@ func @main() {
%iters = constant ${ITERS}: index
/// Run and dump performance for matmul_column_major as a row-major
- %A = alloc() : !row_major_A
- %B = alloc() : !row_major_B
- %C = alloc() : !row_major_C
+ %A = memref.alloc() : !row_major_A
+ %B = memref.alloc() : !row_major_B
+ %C = memref.alloc() : !row_major_C
%t_start_matmul_column_major_as_row_major = call @rtclock() : () -> f64
scf.for %arg0 = %c0 to %iters step %c1 {
// linalg.matmul writes %C in place, need to reset it to zero every time.
@@ -97,34 +97,34 @@ func @main() {
call @print_perf(%iters, %tmatmul_column_major_as_row_major) : (index, f64) -> ()
// CHECK: {{^0$}}
- %cC_ref = alloc() : !column_major_C
+ %cC_ref = memref.alloc() : !column_major_C
linalg.fill(%cC_ref, %f0) : !column_major_C, !elem_type_c
linalg.matmul_column_major ins(%cA, %cB : !column_major_A, !column_major_B)
outs(%cC_ref: !column_major_C)
- %act1 = memref_cast %cC : !column_major_C to memref<*xf32>
- %exp1 = memref_cast %cC_ref : !column_major_C to memref<*xf32>
+ %act1 = memref.cast %cC : !column_major_C to memref<*xf32>
+ %exp1 = memref.cast %cC_ref : !column_major_C to memref<*xf32>
%errors1 = call @verifyMemRefF32(%act1, %exp1) : (memref<*xf32>, memref<*xf32>) -> i64
vector.print %errors1 : i64
- dealloc %cC_ref : !column_major_C
+ memref.dealloc %cC_ref : !column_major_C
// CHECK: {{^0$}}
- %C_ref = alloc() : !row_major_C
+ %C_ref = memref.alloc() : !row_major_C
linalg.fill(%C_ref, %f0) : !row_major_C, !elem_type_c
linalg.matmul ins(%A, %B : !row_major_A, !row_major_B)
outs(%C_ref: !row_major_C)
- %act2 = memref_cast %C : !row_major_C to memref<*xf32>
- %exp2 = memref_cast %C_ref : !row_major_C to memref<*xf32>
+ %act2 = memref.cast %C : !row_major_C to memref<*xf32>
+ %exp2 = memref.cast %C_ref : !row_major_C to memref<*xf32>
%errors2 = call @verifyMemRefF32(%act2, %exp2) : (memref<*xf32>, memref<*xf32>) -> i64
vector.print %errors2 : i64
- dealloc %C_ref : !row_major_C
+ memref.dealloc %C_ref : !row_major_C
- dealloc %A : !row_major_A
- dealloc %B : !row_major_B
- dealloc %C : !row_major_C
+ memref.dealloc %A : !row_major_A
+ memref.dealloc %B : !row_major_B
+ memref.dealloc %C : !row_major_C
- dealloc %cA : !column_major_A
- dealloc %cB : !column_major_B
- dealloc %cC : !column_major_C
+ memref.dealloc %cA : !column_major_A
+ memref.dealloc %cB : !column_major_B
+ memref.dealloc %cC : !column_major_C
return
}
diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir
index d4a84345b86b..5f72ccb6442a 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/benchmark_matmul_i8_i8_i32.mlir
@@ -55,9 +55,9 @@ func @main() {
%v0 = constant 0 : !elem_type_c
%v1 = constant 1 : !elem_type_a
- %A = alloc() : !row_major_A
- %B = alloc() : !row_major_B
- %C = alloc() : !row_major_C
+ %A = memref.alloc() : !row_major_A
+ %B = memref.alloc() : !row_major_B
+ %C = memref.alloc() : !row_major_C
linalg.fill(%A, %v1) : !row_major_A, !elem_type_a
linalg.fill(%B, %v1) : !row_major_B, !elem_type_b
@@ -87,19 +87,19 @@ func @main() {
call @print_perf(%iters, %tmatmul) : (index, f64) -> ()
// CHECK: {{^0$}}
- %C_ref = alloc() : !row_major_C
+ %C_ref = memref.alloc() : !row_major_C
linalg.fill(%C_ref, %v0) : !row_major_C, !elem_type_c
linalg.matmul_i8_i8_i32 ins(%A, %B : !row_major_A, !row_major_B)
outs(%C_ref: !row_major_C)
- %res = memref_cast %C : !row_major_C to memref<*xi32>
- %exp = memref_cast %C_ref : !row_major_C to memref<*xi32>
+ %res = memref.cast %C : !row_major_C to memref<*xi32>
+ %exp = memref.cast %C_ref : !row_major_C to memref<*xi32>
%errors = call @verifyMemRefI32(%res, %exp) : (memref<*xi32>, memref<*xi32>) -> i64
vector.print %errors : i64
- dealloc %C_ref : !row_major_C
+ memref.dealloc %C_ref : !row_major_C
- dealloc %A : !row_major_A
- dealloc %B : !row_major_B
- dealloc %C : !row_major_C
+ memref.dealloc %A : !row_major_A
+ memref.dealloc %B : !row_major_B
+ memref.dealloc %C : !row_major_C
return
}
diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir
index f3b682d9a9f7..4e7ab8d62eb0 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/matmul-vs-matvec.mlir
@@ -9,9 +9,9 @@ func @matmul(%A: memref<?x?xf32>, %B: memref<?x?xf32>) -> (memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%f0 = constant 0.0 : f32
- %x = dim %A, %c0 : memref<?x?xf32>
- %y = dim %B, %c1 : memref<?x?xf32>
- %C = alloc(%x, %y) : memref<?x?xf32>
+ %x = memref.dim %A, %c0 : memref<?x?xf32>
+ %y = memref.dim %B, %c1 : memref<?x?xf32>
+ %C = memref.alloc(%x, %y) : memref<?x?xf32>
linalg.fill(%C, %f0) : memref<?x?xf32>, f32
linalg.matmul ins(%A, %B: memref<?x?xf32>, memref<?x?xf32>)
outs(%C: memref<?x?xf32>)
@@ -22,14 +22,14 @@ func @matvec(%A: memref<?x?xf32>, %B: memref<?x?xf32>) -> (memref<?x?xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%f0 = constant 0.0 : f32
- %m = dim %A, %c0 : memref<?x?xf32>
- %x = dim %A, %c1 : memref<?x?xf32>
- %n = dim %B, %c1 : memref<?x?xf32>
- %C = alloc(%m, %n) : memref<?x?xf32>
+ %m = memref.dim %A, %c0 : memref<?x?xf32>
+ %x = memref.dim %A, %c1 : memref<?x?xf32>
+ %n = memref.dim %B, %c1 : memref<?x?xf32>
+ %C = memref.alloc(%m, %n) : memref<?x?xf32>
linalg.fill(%C, %f0) : memref<?x?xf32>, f32
scf.for %i = %c0 to %n step %c1 {
- %b = subview %B[0, %i][%x, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
- %c = subview %C[0, %i][%m, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
+ %b = memref.subview %B[0, %i][%x, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
+ %c = memref.subview %C[0, %i][%m, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
linalg.matvec ins(%A, %b: memref<?x?xf32>, memref<?xf32, offset: ?, strides: [?]>)
outs(%c: memref<?xf32, offset: ?, strides: [?]>)
}
@@ -44,22 +44,22 @@ func @main() {
%n = constant 2 : index
%val1 = constant 13.0 : f32
%val2 = constant 17.0 : f32
- %A = alloc(%m, %x) : memref<?x?xf32>
- %B = alloc(%x, %n) : memref<?x?xf32>
+ %A = memref.alloc(%m, %x) : memref<?x?xf32>
+ %B = memref.alloc(%x, %n) : memref<?x?xf32>
linalg.fill(%A, %val1) : memref<?x?xf32>, f32
linalg.fill(%B, %val2) : memref<?x?xf32>, f32
- store %val1, %B[%c0, %c0] : memref<?x?xf32>
+ memref.store %val1, %B[%c0, %c0] : memref<?x?xf32>
%C1 = call @matmul(%A, %B) : (memref<?x?xf32>, memref<?x?xf32>) -> memref<?x?xf32>
%C2 = call @matvec(%A, %B) : (memref<?x?xf32>, memref<?x?xf32>) -> memref<?x?xf32>
scf.for %i = %c0 to %m step %c1 {
scf.for %j = %c0 to %n step %c1 {
- %e1 = load %C1[%i, %j] : memref<?x?xf32>
- %e2 = load %C2[%i, %j] : memref<?x?xf32>
+ %e1 = memref.load %C1[%i, %j] : memref<?x?xf32>
+ %e2 = memref.load %C2[%i, %j] : memref<?x?xf32>
%c = cmpf oeq, %e1, %e2 : f32
assert %c, "Matmul does not produce same output as matvec"
}
}
- %C2_ = memref_cast %C2 : memref<?x?xf32> to memref<*xf32>
+ %C2_ = memref.cast %C2 : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%C2_) : (memref<*xf32>) -> ()
return
}
diff --git a/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir b/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir
index ff840c598c4a..568acfad529e 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/rank-reducing-subview.mlir
@@ -13,18 +13,18 @@ func @main() {
%f1 = constant 1.0 : f32
%f2 = constant 2.0 : f32
%f3 = constant 3.0 : f32
- %A = alloc(%c2, %c2) : memref<?x?xf32>
- store %f0, %A[%c0, %c0] : memref<?x?xf32>
- store %f1, %A[%c0, %c1] : memref<?x?xf32>
- store %f2, %A[%c1, %c0] : memref<?x?xf32>
- store %f3, %A[%c1, %c1] : memref<?x?xf32>
- %B = subview %A[%c1, 0][1, %c2][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [1]>
- %C = subview %A[0, %c1][%c2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
- %A_ = memref_cast %A : memref<?x?xf32> to memref<*xf32>
+ %A = memref.alloc(%c2, %c2) : memref<?x?xf32>
+ memref.store %f0, %A[%c0, %c0] : memref<?x?xf32>
+ memref.store %f1, %A[%c0, %c1] : memref<?x?xf32>
+ memref.store %f2, %A[%c1, %c0] : memref<?x?xf32>
+ memref.store %f3, %A[%c1, %c1] : memref<?x?xf32>
+ %B = memref.subview %A[%c1, 0][1, %c2][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [1]>
+ %C = memref.subview %A[0, %c1][%c2, 1][1, 1] : memref<?x?xf32> to memref<?xf32, offset: ?, strides: [?]>
+ %A_ = memref.cast %A : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%A_) : (memref<*xf32>) -> ()
- %B_ = memref_cast %B : memref<?xf32, offset: ?, strides: [1]> to memref<*xf32>
+ %B_ = memref.cast %B : memref<?xf32, offset: ?, strides: [1]> to memref<*xf32>
call @print_memref_f32(%B_) : (memref<*xf32>) -> ()
- %C_ = memref_cast %C : memref<?xf32, offset: ?, strides: [?]> to memref<*xf32>
+ %C_ = memref.cast %C : memref<?xf32, offset: ?, strides: [?]> to memref<*xf32>
call @print_memref_f32(%C_) : (memref<*xf32>) -> ()
return
}
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
index 2e6136a25806..d4dbdc2aa332 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-1d-call.mlir
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns a 1-D buffer of size %s1 filled with the value %f
func @alloc_1d_filled_f32(%s1 : index, %f : f32) -> memref<?xf32> {
- %buf = alloc(%s1) : memref<?xf32>
+ %buf = memref.alloc(%s1) : memref<?xf32>
linalg.fill(%buf, %f) : memref<?xf32>, f32
return %buf : memref<?xf32>
}
@@ -47,14 +47,14 @@ func @main() {
%in1D = call @alloc_1d_filled_f32(%c8, %val) : (index, f32) -> (memref<?xf32>)
%out1D = call @alloc_1d_filled_f32(%c6, %zero) : (index, f32) -> (memref<?xf32>)
- store %f10, %in1D[%c3] : memref<?xf32>
+ memref.store %f10, %in1D[%c3] : memref<?xf32>
call @conv_1d(%in1D, %filter1D, %out1D) : (memref<?xf32>, memref<?xf32>, memref<?xf32>) -> ()
- %out1D_ = memref_cast %out1D : memref<?xf32> to memref<*xf32>
+ %out1D_ = memref.cast %out1D : memref<?xf32> to memref<*xf32>
call @print_memref_f32(%out1D_): (memref<*xf32>) -> ()
- dealloc %filter1D : memref<?xf32>
- dealloc %in1D : memref<?xf32>
- dealloc %out1D : memref<?xf32>
+ memref.dealloc %filter1D : memref<?xf32>
+ memref.dealloc %in1D : memref<?xf32>
+ memref.dealloc %out1D : memref<?xf32>
return
}
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
index 3b99a112f8f2..799d014f879b 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 3-D buffer of size (%s1, %s2, %s3) filled with the value %f
func @alloc_3d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %f : f32) -> memref<?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?xf32>, f32
return %buf : memref<?x?x?xf32>
}
@@ -51,14 +51,14 @@ func @main() {
%in1D_ncw = call @alloc_3d_filled_f32(%c1, %c1, %c8, %val) : (index, index, index, f32) -> (memref<?x?x?xf32>)
%out1D_ncw = call @alloc_3d_filled_f32(%c1, %c1, %c6, %zero) : (index, index, index, f32) -> (memref<?x?x?xf32>)
- store %f10, %in1D_ncw[%c0, %c0, %c3] : memref<?x?x?xf32>
+ memref.store %f10, %in1D_ncw[%c0, %c0, %c3] : memref<?x?x?xf32>
call @conv_1d_input_ncw_filter_wcf(%in1D_ncw, %filter1D_ncw, %out1D_ncw) : (memref<?x?x?xf32>, memref<?x?x?xf32>, memref<?x?x?xf32>) -> ()
- %out1D_ncw_ = memref_cast %out1D_ncw : memref<?x?x?xf32> to memref<*xf32>
+ %out1D_ncw_ = memref.cast %out1D_ncw : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out1D_ncw_): (memref<*xf32>) -> ()
- dealloc %filter1D_ncw : memref<?x?x?xf32>
- dealloc %in1D_ncw : memref<?x?x?xf32>
- dealloc %out1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %filter1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %in1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %out1D_ncw : memref<?x?x?xf32>
return
}
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
index 4936e025202b..5dca06fbe298 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 3-D buffer of size (%s1, %s2, %s3) filled with the value %f
func @alloc_3d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %f : f32) -> memref<?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?xf32>, f32
return %buf : memref<?x?x?xf32>
}
@@ -51,14 +51,14 @@ func @main() {
%in1D_nwc = call @alloc_3d_filled_f32(%c3, %c8, %c1, %val) : (index, index, index, f32) -> (memref<?x?x?xf32>)
%out1D_nwc = call @alloc_3d_filled_f32(%c3, %c6, %c1, %zero) : (index, index, index, f32) -> (memref<?x?x?xf32>)
- store %f10, %in1D_nwc[%c0, %c3, %c0] : memref<?x?x?xf32>
+ memref.store %f10, %in1D_nwc[%c0, %c3, %c0] : memref<?x?x?xf32>
call @conv_1d_input_nwc_filter_wcf(%in1D_nwc, %filter1D_nwc, %out1D_nwc) : (memref<?x?x?xf32>, memref<?x?x?xf32>, memref<?x?x?xf32>) -> ()
- %out1D_nwc_ = memref_cast %out1D_nwc : memref<?x?x?xf32> to memref<*xf32>
+ %out1D_nwc_ = memref.cast %out1D_nwc : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out1D_nwc_): (memref<*xf32>) -> ()
- dealloc %filter1D_nwc : memref<?x?x?xf32>
- dealloc %in1D_nwc : memref<?x?x?xf32>
- dealloc %out1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %filter1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %in1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %out1D_nwc : memref<?x?x?xf32>
return
}
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
index 413573922f0d..82a8e691f349 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 3-D buffer of size (%s1, %s2, %s3) filled with the value %f
func @alloc_3d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %f : f32) -> memref<?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?xf32>, f32
return %buf : memref<?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in1D_ncw = call @alloc_3d_filled_f32(%c1, %c1, %c8, %val) : (index, index, index, f32) -> (memref<?x?x?xf32>)
%out1D_ncw = call @alloc_3d_filled_f32(%c1, %c1, %c6, %zero) : (index, index, index, f32) -> (memref<?x?x?xf32>)
- store %f10, %in1D_ncw[%c0, %c0, %c3] : memref<?x?x?xf32>
+ memref.store %f10, %in1D_ncw[%c0, %c0, %c3] : memref<?x?x?xf32>
call @conv_1d_ncw(%in1D_ncw, %filter1D_ncw, %out1D_ncw) : (memref<?x?x?xf32>, memref<?x?x?xf32>, memref<?x?x?xf32>) -> ()
- %out1D_ncw_ = memref_cast %out1D_ncw : memref<?x?x?xf32> to memref<*xf32>
+ %out1D_ncw_ = memref.cast %out1D_ncw : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out1D_ncw_): (memref<*xf32>) -> ()
- dealloc %filter1D_ncw : memref<?x?x?xf32>
- dealloc %in1D_ncw : memref<?x?x?xf32>
- dealloc %out1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %filter1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %in1D_ncw : memref<?x?x?xf32>
+ memref.dealloc %out1D_ncw : memref<?x?x?xf32>
return
}
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
index c8df60d6ae07..ee727f96c72c 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 3-D buffer of size (%s1, %s2, %s3) filled with the value %f
func @alloc_3d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %f : f32) -> memref<?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?xf32>, f32
return %buf : memref<?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in1D_nwc = call @alloc_3d_filled_f32(%c3, %c8, %c1, %val) : (index, index, index, f32) -> (memref<?x?x?xf32>)
%out1D_nwc = call @alloc_3d_filled_f32(%c3, %c6, %c1, %zero) : (index, index, index, f32) -> (memref<?x?x?xf32>)
- store %f10, %in1D_nwc[%c0, %c3, %c0] : memref<?x?x?xf32>
+ memref.store %f10, %in1D_nwc[%c0, %c3, %c0] : memref<?x?x?xf32>
call @conv_1d_nwc(%in1D_nwc, %filter1D_nwc, %out1D_nwc) : (memref<?x?x?xf32>, memref<?x?x?xf32>, memref<?x?x?xf32>) -> ()
- %out1D_nwc_ = memref_cast %out1D_nwc : memref<?x?x?xf32> to memref<*xf32>
+ %out1D_nwc_ = memref.cast %out1D_nwc : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out1D_nwc_): (memref<*xf32>) -> ()
- dealloc %filter1D_nwc : memref<?x?x?xf32>
- dealloc %in1D_nwc : memref<?x?x?xf32>
- dealloc %out1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %filter1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %in1D_nwc : memref<?x?x?xf32>
+ memref.dealloc %out1D_nwc : memref<?x?x?xf32>
return
}
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
index 3b20315f40d3..dfbb5ec54f39 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-2d-call.mlir
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns a 2-D buffer of size (%s1, %s2) filled with the value %f
func @alloc_2d_filled_f32(%s1 : index, %s2 : index, %f : f32) -> memref<?x?xf32> {
- %buf = alloc(%s1, %s2) : memref<?x?xf32>
+ %buf = memref.alloc(%s1, %s2) : memref<?x?xf32>
linalg.fill(%buf, %f) : memref<?x?xf32>, f32
return %buf : memref<?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in2D = call @alloc_2d_filled_f32(%c8, %c8, %val) : (index, index, f32) -> (memref<?x?xf32>)
%out2D = call @alloc_2d_filled_f32(%c6, %c6, %zero) : (index, index, f32) -> (memref<?x?xf32>)
- store %f10, %in2D[%c0, %c3] : memref<?x?xf32>
+ memref.store %f10, %in2D[%c0, %c3] : memref<?x?xf32>
call @conv_2d(%in2D, %filter2D, %out2D) : (memref<?x?xf32>, memref<?x?xf32>, memref<?x?xf32>) -> ()
- %out2D_ = memref_cast %out2D : memref<?x?xf32> to memref<*xf32>
+ %out2D_ = memref.cast %out2D : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%out2D_): (memref<*xf32>) -> ()
- dealloc %filter2D : memref<?x?xf32>
- dealloc %in2D : memref<?x?xf32>
- dealloc %out2D : memref<?x?xf32>
+ memref.dealloc %filter2D : memref<?x?xf32>
+ memref.dealloc %in2D : memref<?x?xf32>
+ memref.dealloc %out2D : memref<?x?xf32>
return
}
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
index 58bc537a0194..3fc07ed00097 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 4-D buffer of size (%s1, %s2, %s3, %s4) filled with the value %f
func @alloc_4d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %f : f32) -> memref<?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?xf32>
}
@@ -51,14 +51,14 @@ func @main() {
%in2D_nchw = call @alloc_4d_filled_f32(%c3, %c1, %c8, %c8, %val) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
%out2D_nchw = call @alloc_4d_filled_f32(%c3, %c1, %c6, %c6, %zero) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
- store %f10, %in2D_nchw[%c0, %c0, %c0, %c3] : memref<?x?x?x?xf32>
+ memref.store %f10, %in2D_nchw[%c0, %c0, %c0, %c3] : memref<?x?x?x?xf32>
call @conv_2d_input_nchw_filter_hwcf(%in2D_nchw, %filter2D_nchw, %out2D_nchw) : (memref<?x?x?x?xf32>, memref<?x?x?x?xf32>, memref<?x?x?x?xf32>) -> ()
- %out2D_nchw_ = memref_cast %out2D_nchw : memref<?x?x?x?xf32> to memref<*xf32>
+ %out2D_nchw_ = memref.cast %out2D_nchw : memref<?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out2D_nchw_): (memref<*xf32>) -> ()
- dealloc %filter2D_nchw : memref<?x?x?x?xf32>
- dealloc %in2D_nchw : memref<?x?x?x?xf32>
- dealloc %out2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %filter2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %in2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %out2D_nchw : memref<?x?x?x?xf32>
return
}
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
index c50c84c63a01..519a15578ced 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 4-D buffer of size (%s1, %s2, %s3, %s4) filled with the value %f
func @alloc_4d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %f : f32) -> memref<?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?xf32>
}
@@ -51,14 +51,14 @@ func @main() {
%in2D_nhwc = call @alloc_4d_filled_f32(%c3, %c8, %c8, %c3, %val) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
%out2D_nhwc = call @alloc_4d_filled_f32(%c3, %c6, %c6, %c1, %zero) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
- store %f10, %in2D_nhwc[%c0, %c0, %c3, %c0] : memref<?x?x?x?xf32>
+ memref.store %f10, %in2D_nhwc[%c0, %c0, %c3, %c0] : memref<?x?x?x?xf32>
call @conv_2d_input_nhwc_filter_hwcf(%in2D_nhwc, %filter2D_nhwc, %out2D_nhwc) : (memref<?x?x?x?xf32>, memref<?x?x?x?xf32>, memref<?x?x?x?xf32>) -> ()
- %out2D_nhwc_ = memref_cast %out2D_nhwc : memref<?x?x?x?xf32> to memref<*xf32>
+ %out2D_nhwc_ = memref.cast %out2D_nhwc : memref<?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out2D_nhwc_): (memref<*xf32>) -> ()
- dealloc %filter2D_nhwc : memref<?x?x?x?xf32>
- dealloc %in2D_nhwc : memref<?x?x?x?xf32>
- dealloc %out2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %filter2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %in2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %out2D_nhwc : memref<?x?x?x?xf32>
return
}
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
index 8fd541238781..feb587a0be5c 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 4-D buffer of size (%s1, %s2, %s3, %s4) filled with the value %f
func @alloc_4d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %f : f32) -> memref<?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in2D_nchw = call @alloc_4d_filled_f32(%c3, %c1, %c8, %c8, %val) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
%out2D_nchw = call @alloc_4d_filled_f32(%c3, %c1, %c6, %c6, %zero) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
- store %f10, %in2D_nchw[%c0, %c0, %c0, %c3] : memref<?x?x?x?xf32>
+ memref.store %f10, %in2D_nchw[%c0, %c0, %c0, %c3] : memref<?x?x?x?xf32>
call @conv_2d_nchw(%in2D_nchw, %filter2D_nchw, %out2D_nchw) : (memref<?x?x?x?xf32>, memref<?x?x?x?xf32>, memref<?x?x?x?xf32>) -> ()
- %out2D_nchw_ = memref_cast %out2D_nchw : memref<?x?x?x?xf32> to memref<*xf32>
+ %out2D_nchw_ = memref.cast %out2D_nchw : memref<?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out2D_nchw_): (memref<*xf32>) -> ()
- dealloc %filter2D_nchw : memref<?x?x?x?xf32>
- dealloc %in2D_nchw : memref<?x?x?x?xf32>
- dealloc %out2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %filter2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %in2D_nchw : memref<?x?x?x?xf32>
+ memref.dealloc %out2D_nchw : memref<?x?x?x?xf32>
return
}
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
index 87dbf2cec820..729ab9f76d8e 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 4-D buffer of size (%s1, %s2, %s3, %s4) filled with the value %f
func @alloc_4d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %f : f32) -> memref<?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4) : memref<?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in2D_nhwc = call @alloc_4d_filled_f32(%c3, %c8, %c8, %c3, %val) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
%out2D_nhwc = call @alloc_4d_filled_f32(%c3, %c6, %c6, %c1, %zero) : (index, index, index, index, f32) -> (memref<?x?x?x?xf32>)
- store %f10, %in2D_nhwc[%c0, %c0, %c3, %c0] : memref<?x?x?x?xf32>
+ memref.store %f10, %in2D_nhwc[%c0, %c0, %c3, %c0] : memref<?x?x?x?xf32>
call @conv_2d_nhwc(%in2D_nhwc, %filter2D_nhwc, %out2D_nhwc) : (memref<?x?x?x?xf32>, memref<?x?x?x?xf32>, memref<?x?x?x?xf32>) -> ()
- %out2D_nhwc_ = memref_cast %out2D_nhwc : memref<?x?x?x?xf32> to memref<*xf32>
+ %out2D_nhwc_ = memref.cast %out2D_nhwc : memref<?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out2D_nhwc_): (memref<*xf32>) -> ()
- dealloc %filter2D_nhwc : memref<?x?x?x?xf32>
- dealloc %in2D_nhwc : memref<?x?x?x?xf32>
- dealloc %out2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %filter2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %in2D_nhwc : memref<?x?x?x?xf32>
+ memref.dealloc %out2D_nhwc : memref<?x?x?x?xf32>
return
}
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
index 38e72fef0017..8f4ef13b94aa 100644
--- a/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir
+++ b/mlir/test/Integration/Dialect/Linalg/CPU/test-conv-3d-call.mlir
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 3-D buffer of size (%s1, %s2, %s3) filled with the value %f
func @alloc_3d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %f : f32) -> memref<?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3) : memref<?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?xf32>, f32
return %buf : memref<?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in3D = call @alloc_3d_filled_f32(%c8, %c8, %c8, %val) : (index, index, index, f32) -> (memref<?x?x?xf32>)
%out3D = call @alloc_3d_filled_f32(%c6, %c6, %c6, %zero) : (index, index, index, f32) -> (memref<?x?x?xf32>)
- store %f10, %in3D[%c0, %c0, %c3] : memref<?x?x?xf32>
+ memref.store %f10, %in3D[%c0, %c0, %c3] : memref<?x?x?xf32>
call @conv_3d(%in3D, %filter3D, %out3D) : (memref<?x?x?xf32>, memref<?x?x?xf32>, memref<?x?x?xf32>) -> ()
- %out3D_ = memref_cast %out3D : memref<?x?x?xf32> to memref<*xf32>
+ %out3D_ = memref.cast %out3D : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out3D_): (memref<*xf32>) -> ()
- dealloc %filter3D : memref<?x?x?xf32>
- dealloc %in3D : memref<?x?x?xf32>
- dealloc %out3D : memref<?x?x?xf32>
+ memref.dealloc %filter3D : memref<?x?x?xf32>
+ memref.dealloc %in3D : memref<?x?x?xf32>
+ memref.dealloc %out3D : memref<?x?x?xf32>
return
}
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
index 9cf45f1bd18c..7c64bd1406a7 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 5-D buffer of size (%s1, %s2, %s3, %s4, %s5) filled with the value %f
func @alloc_5d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %s5 : index, %f : f32) -> memref<?x?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?x?xf32>
}
@@ -51,14 +51,14 @@ func @main() {
%in3D_ncdhw = call @alloc_5d_filled_f32(%c1, %c1, %c8, %c8, %c8, %val) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
%out3D_ncdhw = call @alloc_5d_filled_f32(%c1, %c1, %c6, %c6, %c6, %zero) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
- store %f10, %in3D_ncdhw[%c0, %c0, %c0, %c0, %c3] : memref<?x?x?x?x?xf32>
+ memref.store %f10, %in3D_ncdhw[%c0, %c0, %c0, %c0, %c3] : memref<?x?x?x?x?xf32>
call @conv_3d_input_ncdhw_filter_dhwcf(%in3D_ncdhw, %filter3D_ncdhw, %out3D_ncdhw) : (memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>) -> ()
- %out3D_ncdhw_ = memref_cast %out3D_ncdhw : memref<?x?x?x?x?xf32> to memref<*xf32>
+ %out3D_ncdhw_ = memref.cast %out3D_ncdhw : memref<?x?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out3D_ncdhw_): (memref<*xf32>) -> ()
- dealloc %filter3D_ncdhw : memref<?x?x?x?x?xf32>
- dealloc %in3D_ncdhw : memref<?x?x?x?x?xf32>
- dealloc %out3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %filter3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %in3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %out3D_ncdhw : memref<?x?x?x?x?xf32>
return
}
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
index e425b9a132f4..3497a5cf99d4 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 5-D buffer of size (%s1, %s2, %s3, %s4, %s5) filled with the value %f
func @alloc_5d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %s5 : index, %f : f32) -> memref<?x?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?x?xf32>
}
@@ -52,14 +52,14 @@ func @main() {
%in3D_ndhwc = call @alloc_5d_filled_f32(%c1, %c8, %c8, %c8, %c1, %val) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
%out3D_ndhwc = call @alloc_5d_filled_f32(%c1, %c6, %c6, %c6, %c1, %zero) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
- store %f10, %in3D_ndhwc[%c0, %c0, %c0, %c3, %c0] : memref<?x?x?x?x?xf32>
+ memref.store %f10, %in3D_ndhwc[%c0, %c0, %c0, %c3, %c0] : memref<?x?x?x?x?xf32>
call @conv_3d_input_ndhwc_filter_dhwcf(%in3D_ndhwc, %filter3D_ndhwc, %out3D_ndhwc) : (memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>) -> ()
- %out3D_ndhwc_ = memref_cast %out3D_ndhwc : memref<?x?x?x?x?xf32> to memref<*xf32>
+ %out3D_ndhwc_ = memref.cast %out3D_ndhwc : memref<?x?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out3D_ndhwc_): (memref<*xf32>) -> ()
- dealloc %filter3D_ndhwc : memref<?x?x?x?x?xf32>
- dealloc %in3D_ndhwc : memref<?x?x?x?x?xf32>
- dealloc %out3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %filter3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %in3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %out3D_ndhwc : memref<?x?x?x?x?xf32>
return
}
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
index b6707539d461..f17d38391a08 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 5-D buffer of size (%s1, %s2, %s3, %s4, %s5) filled with the value %f
func @alloc_5d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %s5 : index, %f : f32) -> memref<?x?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?x?xf32>
}
@@ -49,14 +49,14 @@ func @main() {
%in3D_ncdhw = call @alloc_5d_filled_f32(%c1, %c1, %c8, %c8, %c8, %val) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
%out3D_ncdhw = call @alloc_5d_filled_f32(%c1, %c1, %c6, %c6, %c6, %zero) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
- store %f10, %in3D_ncdhw[%c0, %c0, %c0, %c0, %c3] : memref<?x?x?x?x?xf32>
+ memref.store %f10, %in3D_ncdhw[%c0, %c0, %c0, %c0, %c3] : memref<?x?x?x?x?xf32>
call @conv_3d_ncdhw(%in3D_ncdhw, %filter3D_ncdhw, %out3D_ncdhw) : (memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>) -> ()
- %out3D_ncdhw_ = memref_cast %out3D_ncdhw : memref<?x?x?x?x?xf32> to memref<*xf32>
+ %out3D_ncdhw_ = memref.cast %out3D_ncdhw : memref<?x?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out3D_ncdhw_): (memref<*xf32>) -> ()
- dealloc %filter3D_ncdhw : memref<?x?x?x?x?xf32>
- dealloc %in3D_ncdhw : memref<?x?x?x?x?xf32>
- dealloc %out3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %filter3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %in3D_ncdhw : memref<?x?x?x?x?xf32>
+ memref.dealloc %out3D_ncdhw : memref<?x?x?x?x?xf32>
return
}
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
index bc4713810ed4..b364d85a252f 100644
--- 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
@@ -24,7 +24,7 @@ func private @print_memref_f32(memref<*xf32>)
// Creates and returns 5-D buffer of size (%s1, %s2, %s3, %s4, %s5) filled with the value %f
func @alloc_5d_filled_f32(%s1 : index, %s2 : index, %s3 : index, %s4 : index, %s5 : index, %f : f32) -> memref<?x?x?x?x?xf32> {
- %buf = alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
+ %buf = memref.alloc(%s1, %s2, %s3, %s4, %s5) : memref<?x?x?x?x?xf32>
linalg.fill(%buf, %f) : memref<?x?x?x?x?xf32>, f32
return %buf : memref<?x?x?x?x?xf32>
}
@@ -50,14 +50,14 @@ func @main() {
%in3D_ndhwc = call @alloc_5d_filled_f32(%c1, %c8, %c8, %c8, %c1, %val) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
%out3D_ndhwc = call @alloc_5d_filled_f32(%c1, %c6, %c6, %c6, %c1, %zero) : (index, index, index, index, index, f32) -> (memref<?x?x?x?x?xf32>)
- store %f10, %in3D_ndhwc[%c0, %c0, %c0, %c3, %c0] : memref<?x?x?x?x?xf32>
+ memref.store %f10, %in3D_ndhwc[%c0, %c0, %c0, %c3, %c0] : memref<?x?x?x?x?xf32>
call @conv_3d_ndhwc(%in3D_ndhwc, %filter3D_ndhwc, %out3D_ndhwc) : (memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>, memref<?x?x?x?x?xf32>) -> ()
- %out3D_ndhwc_ = memref_cast %out3D_ndhwc : memref<?x?x?x?x?xf32> to memref<*xf32>
+ %out3D_ndhwc_ = memref.cast %out3D_ndhwc : memref<?x?x?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%out3D_ndhwc_): (memref<*xf32>) -> ()
- dealloc %filter3D_ndhwc : memref<?x?x?x?x?xf32>
- dealloc %in3D_ndhwc : memref<?x?x?x?x?xf32>
- dealloc %out3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %filter3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %in3D_ndhwc : memref<?x?x?x?x?xf32>
+ memref.dealloc %out3D_ndhwc : memref<?x?x?x?x?xf32>
return
}
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
index a7c06c7c8391..9d625f3daead 100644
--- 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
@@ -17,13 +17,13 @@ func @entry() {
%c20 = constant 20: i32
%c10 = constant 10: i32
%cmin10 = constant -10: i32
- %A = alloc() : memref<40xi32>
+ %A = memref.alloc() : memref<40xi32>
// print numerator
affine.for %i = 0 to 40 {
%ii = index_cast %i: index to i32
%ii30 = subi %ii, %c20 : i32
- store %ii30, %A[%i] : memref<40xi32>
+ memref.store %ii30, %A[%i] : memref<40xi32>
}
call @transfer_read_2d(%A, %c0) : (memref<40xi32>, index) -> ()
@@ -32,7 +32,7 @@ func @entry() {
%ii = index_cast %i: index to i32
%ii30 = subi %ii, %c20 : i32
%val = ceildivi_signed %ii30, %c10 : i32
- store %val, %A[%i] : memref<40xi32>
+ memref.store %val, %A[%i] : memref<40xi32>
}
call @transfer_read_2d(%A, %c0) : (memref<40xi32>, index) -> ()
@@ -41,7 +41,7 @@ func @entry() {
%ii = index_cast %i: index to i32
%ii30 = subi %ii, %c20 : i32
%val = floordivi_signed %ii30, %c10 : i32
- store %val, %A[%i] : memref<40xi32>
+ memref.store %val, %A[%i] : memref<40xi32>
}
call @transfer_read_2d(%A, %c0) : (memref<40xi32>, index) -> ()
@@ -51,7 +51,7 @@ func @entry() {
%ii = index_cast %i: index to i32
%ii30 = subi %ii, %c20 : i32
%val = ceildivi_signed %ii30, %cmin10 : i32
- store %val, %A[%i] : memref<40xi32>
+ memref.store %val, %A[%i] : memref<40xi32>
}
call @transfer_read_2d(%A, %c0) : (memref<40xi32>, index) -> ()
@@ -60,7 +60,7 @@ func @entry() {
%ii = index_cast %i: index to i32
%ii30 = subi %ii, %c20 : i32
%val = floordivi_signed %ii30, %cmin10 : i32
- store %val, %A[%i] : memref<40xi32>
+ memref.store %val, %A[%i] : memref<40xi32>
}
call @transfer_read_2d(%A, %c0) : (memref<40xi32>, index) -> ()
diff --git a/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir b/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir
index 9fdaa3f21853..dace65911db8 100644
--- a/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir
+++ b/mlir/test/Integration/Dialect/Standard/CPU/test_subview.mlir
@@ -2,7 +2,7 @@
// RUN: mlir-cpu-runner -e main -entry-point-result=void \
// RUN: -shared-libs=%mlir_integration_test_dir/libmlir_runner_utils%shlibext | FileCheck %s
-global_memref "private" constant @__constant_5x3xf32 : memref<5x3xf32> =
+memref.global "private" constant @__constant_5x3xf32 : memref<5x3xf32> =
dense<[[0.0, 1.0, 2.0],
[3.0, 4.0, 5.0],
[6.0, 7.0, 8.0],
@@ -10,11 +10,11 @@ dense<[[0.0, 1.0, 2.0],
[12.0, 13.0, 14.0]]>
func @main() {
- %0 = get_global_memref @__constant_5x3xf32 : memref<5x3xf32>
+ %0 = memref.get_global @__constant_5x3xf32 : memref<5x3xf32>
/// Subview with only leading operands.
- %1 = subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
- %unranked = memref_cast %1 : memref<3x3xf32, offset: 6, strides: [3, 1]> to memref<*xf32>
+ %1 = memref.subview %0[2][3][1]: memref<5x3xf32> to memref<3x3xf32, offset: 6, strides: [3, 1]>
+ %unranked = memref.cast %1 : memref<3x3xf32, offset: 6, strides: [3, 1]> to memref<*xf32>
call @print_memref_f32(%unranked) : (memref<*xf32>) -> ()
// CHECK: Unranked Memref base@ = {{0x[-9a-f]*}}
@@ -26,8 +26,8 @@ func @main() {
// CHECK-SAME: ]
/// Regular subview.
- %2 = subview %0[0, 2][5, 1][1, 1]: memref<5x3xf32> to memref<5x1xf32, offset: 2, strides: [3, 1]>
- %unranked2 = memref_cast %2 : memref<5x1xf32, offset: 2, strides: [3, 1]> to memref<*xf32>
+ %2 = memref.subview %0[0, 2][5, 1][1, 1]: memref<5x3xf32> to memref<5x1xf32, offset: 2, strides: [3, 1]>
+ %unranked2 = memref.cast %2 : memref<5x1xf32, offset: 2, strides: [3, 1]> to memref<*xf32>
call @print_memref_f32(%unranked2) : (memref<*xf32>) -> ()
// CHECK: Unranked Memref base@ = {{0x[-9a-f]*}}
@@ -41,8 +41,8 @@ func @main() {
// CHECK-SAME: ]
/// Rank-reducing subview.
- %3 = subview %0[0, 2][5, 1][1, 1]: memref<5x3xf32> to memref<5xf32, offset: 2, strides: [3]>
- %unranked3 = memref_cast %3 : memref<5xf32, offset: 2, strides: [3]> to memref<*xf32>
+ %3 = memref.subview %0[0, 2][5, 1][1, 1]: memref<5x3xf32> to memref<5xf32, offset: 2, strides: [3]>
+ %unranked3 = memref.cast %3 : memref<5xf32, offset: 2, strides: [3]> to memref<*xf32>
call @print_memref_f32(%unranked3) : (memref<*xf32>) -> ()
// CHECK: Unranked Memref base@ = {{0x[-9a-f]*}}
@@ -50,8 +50,8 @@ func @main() {
// CHECK-NEXT: [2, 5, 8, 11, 14]
/// Rank-reducing subview with only leading operands.
- %4 = subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
- %unranked4 = memref_cast %4 : memref<3xf32, offset: 3, strides: [1]> to memref<*xf32>
+ %4 = memref.subview %0[1][1][1]: memref<5x3xf32> to memref<3xf32, offset: 3, strides: [1]>
+ %unranked4 = memref.cast %4 : memref<3xf32, offset: 3, strides: [1]> to memref<*xf32>
call @print_memref_f32(%unranked4) : (memref<*xf32>) -> ()
// CHECK: Unranked Memref base@ = {{0x[-9a-f]*}}
// CHECK-SAME: rank = 1 offset = 3 sizes = [3] strides = [1] data =
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir
index 76022203e113..6178d3e9a6ae 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-compress.mlir
@@ -27,7 +27,7 @@ func @printmem16(%A: memref<?xf32>) {
%m = vector.broadcast %z : f32 to vector<16xf32>
%mem = scf.for %i = %c0 to %c16 step %c1
iter_args(%m_iter = %m) -> (vector<16xf32>) {
- %c = load %A[%i] : memref<?xf32>
+ %c = memref.load %A[%i] : memref<?xf32>
%i32 = index_cast %i : index to i32
%m_new = vector.insertelement %c, %m_iter[%i32 : i32] : vector<16xf32>
scf.yield %m_new : vector<16xf32>
@@ -41,12 +41,12 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c16 = constant 16: index
- %A = alloc(%c16) : memref<?xf32>
+ %A = memref.alloc(%c16) : memref<?xf32>
%z = constant 0.0: f32
%v = vector.broadcast %z : f32 to vector<16xf32>
%value = scf.for %i = %c0 to %c16 step %c1
iter_args(%v_iter = %v) -> (vector<16xf32>) {
- store %z, %A[%i] : memref<?xf32>
+ memref.store %z, %A[%i] : memref<?xf32>
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
%v_new = vector.insertelement %fi, %v_iter[%i32 : i32] : vector<16xf32>
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir
index b63294f76e7a..3a66d7f8dcc7 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-expand.mlir
@@ -26,11 +26,11 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c16 = constant 16: index
- %A = alloc(%c16) : memref<?xf32>
+ %A = memref.alloc(%c16) : memref<?xf32>
scf.for %i = %c0 to %c16 step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
- store %fi, %A[%i] : memref<?xf32>
+ memref.store %fi, %A[%i] : memref<?xf32>
}
// Set up pass thru vector.
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir
index 0e5ea8557fbc..9273d880207b 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-gather.mlir
@@ -16,11 +16,11 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c10 = constant 10: index
- %A = alloc(%c10) : memref<?xf32>
+ %A = memref.alloc(%c10) : memref<?xf32>
scf.for %i = %c0 to %c10 step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
- store %fi, %A[%i] : memref<?xf32>
+ memref.store %fi, %A[%i] : memref<?xf32>
}
// Set up idx vector.
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir
index d5353affae45..da4dd5a45881 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedload.mlir
@@ -24,11 +24,11 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c16 = constant 16: index
- %A = alloc(%c16) : memref<?xf32>
+ %A = memref.alloc(%c16) : memref<?xf32>
scf.for %i = %c0 to %c16 step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
- store %fi, %A[%i] : memref<?xf32>
+ memref.store %fi, %A[%i] : memref<?xf32>
}
// Set up pass thru vector.
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir
index a98789847d48..ce895c0070c5 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-maskedstore.mlir
@@ -27,7 +27,7 @@ func @printmem16(%A: memref<?xf32>) {
%m = vector.broadcast %z : f32 to vector<16xf32>
%mem = scf.for %i = %c0 to %c16 step %c1
iter_args(%m_iter = %m) -> (vector<16xf32>) {
- %c = load %A[%i] : memref<?xf32>
+ %c = memref.load %A[%i] : memref<?xf32>
%i32 = index_cast %i : index to i32
%m_new = vector.insertelement %c, %m_iter[%i32 : i32] : vector<16xf32>
scf.yield %m_new : vector<16xf32>
@@ -42,9 +42,9 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c16 = constant 16: index
- %A = alloc(%c16) : memref<?xf32>
+ %A = memref.alloc(%c16) : memref<?xf32>
scf.for %i = %c0 to %c16 step %c1 {
- store %f0, %A[%i] : memref<?xf32>
+ memref.store %f0, %A[%i] : memref<?xf32>
}
// Set up value vector.
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir
index f4316faee5c7..abe67da8bd34 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-scatter.mlir
@@ -20,7 +20,7 @@ func @printmem8(%A: memref<?xf32>) {
%m = vector.broadcast %z : f32 to vector<8xf32>
%mem = scf.for %i = %c0 to %c8 step %c1
iter_args(%m_iter = %m) -> (vector<8xf32>) {
- %c = load %A[%i] : memref<?xf32>
+ %c = memref.load %A[%i] : memref<?xf32>
%i32 = index_cast %i : index to i32
%m_new = vector.insertelement %c, %m_iter[%i32 : i32] : vector<8xf32>
scf.yield %m_new : vector<8xf32>
@@ -34,11 +34,11 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c8 = constant 8: index
- %A = alloc(%c8) : memref<?xf32>
+ %A = memref.alloc(%c8) : memref<?xf32>
scf.for %i = %c0 to %c8 step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
- store %fi, %A[%i] : memref<?xf32>
+ memref.store %fi, %A[%i] : memref<?xf32>
}
// Set up idx vector.
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
index 4f399886dd91..f4aee2627351 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-sparse-dot-matvec.mlir
@@ -62,12 +62,12 @@ func @spmv8x8(%AVAL: memref<8xvector<4xf32>>,
%mask = vector.constant_mask [4] : vector<4xi1>
%pass = vector.broadcast %f0 : f32 to vector<4xf32>
scf.for %i = %c0 to %cn step %c1 {
- %aval = load %AVAL[%i] : memref<8xvector<4xf32>>
- %aidx = load %AIDX[%i] : memref<8xvector<4xi32>>
+ %aval = memref.load %AVAL[%i] : memref<8xvector<4xf32>>
+ %aidx = memref.load %AIDX[%i] : memref<8xvector<4xi32>>
%0 = vector.gather %X[%c0][%aidx], %mask, %pass
: memref<?xf32>, vector<4xi32>, vector<4xi1>, vector<4xf32> into vector<4xf32>
%1 = vector.contract #dot_trait %aval, %0, %f0 : vector<4xf32>, vector<4xf32> into f32
- store %1, %B[%i] : memref<?xf32>
+ memref.store %1, %B[%i] : memref<?xf32>
}
return
}
@@ -106,10 +106,10 @@ func @entry() {
// Allocate.
//
- %AVAL = alloc() {alignment = 64} : memref<8xvector<4xf32>>
- %AIDX = alloc() {alignment = 64} : memref<8xvector<4xi32>>
- %X = alloc(%c8) {alignment = 64} : memref<?xf32>
- %B = alloc(%c8) {alignment = 64} : memref<?xf32>
+ %AVAL = memref.alloc() {alignment = 64} : memref<8xvector<4xf32>>
+ %AIDX = memref.alloc() {alignment = 64} : memref<8xvector<4xi32>>
+ %X = memref.alloc(%c8) {alignment = 64} : memref<?xf32>
+ %B = memref.alloc(%c8) {alignment = 64} : memref<?xf32>
//
// Initialize.
@@ -118,88 +118,88 @@ func @entry() {
%vf1 = vector.broadcast %f1 : f32 to vector<4xf32>
%0 = vector.insert %f2, %vf1[1] : f32 into vector<4xf32>
- store %0, %AVAL[%c0] : memref<8xvector<4xf32>>
+ memref.store %0, %AVAL[%c0] : memref<8xvector<4xf32>>
%1 = vector.insert %f8, %vf1[1] : f32 into vector<4xf32>
%2 = vector.insert %f3, %1[2] : f32 into vector<4xf32>
- store %2, %AVAL[%c1] : memref<8xvector<4xf32>>
+ memref.store %2, %AVAL[%c1] : memref<8xvector<4xf32>>
%3 = vector.insert %f2, %vf1[1] : f32 into vector<4xf32>
%4 = vector.insert %f6, %3[2] : f32 into vector<4xf32>
%5 = vector.insert %f2, %4[3] : f32 into vector<4xf32>
- store %5, %AVAL[%c2] : memref<8xvector<4xf32>>
+ memref.store %5, %AVAL[%c2] : memref<8xvector<4xf32>>
%6 = vector.insert %f3, %vf1[0] : f32 into vector<4xf32>
- store %6, %AVAL[%c3] : memref<8xvector<4xf32>>
+ memref.store %6, %AVAL[%c3] : memref<8xvector<4xf32>>
%7 = vector.insert %f5, %vf1[0] : f32 into vector<4xf32>
- store %7, %AVAL[%c4] : memref<8xvector<4xf32>>
+ memref.store %7, %AVAL[%c4] : memref<8xvector<4xf32>>
%8 = vector.insert %f3, %vf1[0] : f32 into vector<4xf32>
%9 = vector.insert %f2, %8[1] : f32 into vector<4xf32>
%10 = vector.insert %f2, %9[3] : f32 into vector<4xf32>
- store %10, %AVAL[%c5] : memref<8xvector<4xf32>>
+ memref.store %10, %AVAL[%c5] : memref<8xvector<4xf32>>
%11 = vector.insert %f4, %vf1[0] : f32 into vector<4xf32>
%12 = vector.insert %f7, %11[1] : f32 into vector<4xf32>
- store %12, %AVAL[%c6] : memref<8xvector<4xf32>>
+ memref.store %12, %AVAL[%c6] : memref<8xvector<4xf32>>
%13 = vector.insert %f3, %vf1[0] : f32 into vector<4xf32>
%14 = vector.insert %f2, %13[1] : f32 into vector<4xf32>
- store %14, %AVAL[%c7] : memref<8xvector<4xf32>>
+ memref.store %14, %AVAL[%c7] : memref<8xvector<4xf32>>
%vi0 = vector.broadcast %i0 : i32 to vector<4xi32>
%20 = vector.insert %i2, %vi0[1] : i32 into vector<4xi32>
%21 = vector.insert %i5, %20[2] : i32 into vector<4xi32>
%22 = vector.insert %i7, %21[3] : i32 into vector<4xi32>
- store %22, %AIDX[%c0] : memref<8xvector<4xi32>>
+ memref.store %22, %AIDX[%c0] : memref<8xvector<4xi32>>
%23 = vector.insert %i1, %vi0[1] : i32 into vector<4xi32>
%24 = vector.insert %i4, %23[2] : i32 into vector<4xi32>
%25 = vector.insert %i6, %24[3] : i32 into vector<4xi32>
- store %25, %AIDX[%c1] : memref<8xvector<4xi32>>
+ memref.store %25, %AIDX[%c1] : memref<8xvector<4xi32>>
%26 = vector.insert %i2, %vi0[0] : i32 into vector<4xi32>
%27 = vector.insert %i5, %26[1] : i32 into vector<4xi32>
%28 = vector.insert %i6, %27[2] : i32 into vector<4xi32>
%29 = vector.insert %i7, %28[3] : i32 into vector<4xi32>
- store %29, %AIDX[%c2] : memref<8xvector<4xi32>>
+ memref.store %29, %AIDX[%c2] : memref<8xvector<4xi32>>
%30 = vector.insert %i1, %vi0[0] : i32 into vector<4xi32>
%31 = vector.insert %i3, %30[1] : i32 into vector<4xi32>
%32 = vector.insert %i5, %31[2] : i32 into vector<4xi32>
%33 = vector.insert %i7, %32[3] : i32 into vector<4xi32>
- store %33, %AIDX[%c3] : memref<8xvector<4xi32>>
+ memref.store %33, %AIDX[%c3] : memref<8xvector<4xi32>>
%34 = vector.insert %i3, %vi0[1] : i32 into vector<4xi32>
%35 = vector.insert %i4, %34[2] : i32 into vector<4xi32>
%36 = vector.insert %i5, %35[3] : i32 into vector<4xi32>
- store %36, %AIDX[%c4] : memref<8xvector<4xi32>>
+ memref.store %36, %AIDX[%c4] : memref<8xvector<4xi32>>
%37 = vector.insert %i1, %vi0[0] : i32 into vector<4xi32>
%38 = vector.insert %i4, %37[1] : i32 into vector<4xi32>
%39 = vector.insert %i5, %38[2] : i32 into vector<4xi32>
%40 = vector.insert %i6, %39[3] : i32 into vector<4xi32>
- store %40, %AIDX[%c5] : memref<8xvector<4xi32>>
+ memref.store %40, %AIDX[%c5] : memref<8xvector<4xi32>>
%41 = vector.insert %i2, %vi0[1] : i32 into vector<4xi32>
%42 = vector.insert %i4, %41[2] : i32 into vector<4xi32>
%43 = vector.insert %i6, %42[3] : i32 into vector<4xi32>
- store %43, %AIDX[%c6] : memref<8xvector<4xi32>>
+ memref.store %43, %AIDX[%c6] : memref<8xvector<4xi32>>
%44 = vector.insert %i1, %vi0[0] : i32 into vector<4xi32>
%45 = vector.insert %i3, %44[1] : i32 into vector<4xi32>
%46 = vector.insert %i6, %45[2] : i32 into vector<4xi32>
%47 = vector.insert %i7, %46[3] : i32 into vector<4xi32>
- store %47, %AIDX[%c7] : memref<8xvector<4xi32>>
+ memref.store %47, %AIDX[%c7] : memref<8xvector<4xi32>>
scf.for %i = %c0 to %c8 step %c1 {
%ix = addi %i, %c1 : index
%kx = index_cast %ix : index to i32
%fx = sitofp %kx : i32 to f32
- store %fx, %X[%i] : memref<?xf32>
- store %f0, %B[%i] : memref<?xf32>
+ memref.store %fx, %X[%i] : memref<?xf32>
+ memref.store %f0, %B[%i] : memref<?xf32>
}
//
@@ -215,17 +215,17 @@ func @entry() {
//
scf.for %i = %c0 to %c8 step %c1 {
- %aval = load %AVAL[%i] : memref<8xvector<4xf32>>
+ %aval = memref.load %AVAL[%i] : memref<8xvector<4xf32>>
vector.print %aval : vector<4xf32>
}
scf.for %i = %c0 to %c8 step %c1 {
- %aidx = load %AIDX[%i] : memref<8xvector<4xi32>>
+ %aidx = memref.load %AIDX[%i] : memref<8xvector<4xi32>>
vector.print %aidx : vector<4xi32>
}
scf.for %i = %c0 to %c8 step %c1 {
- %ldb = load %B[%i] : memref<?xf32>
+ %ldb = memref.load %B[%i] : memref<?xf32>
vector.print %ldb : f32
}
@@ -262,10 +262,10 @@ func @entry() {
// Free.
//
- dealloc %AVAL : memref<8xvector<4xf32>>
- dealloc %AIDX : memref<8xvector<4xi32>>
- dealloc %X : memref<?xf32>
- dealloc %B : memref<?xf32>
+ memref.dealloc %AVAL : memref<8xvector<4xf32>>
+ memref.dealloc %AIDX : memref<8xvector<4xi32>>
+ memref.dealloc %X : memref<?xf32>
+ memref.dealloc %B : memref<?xf32>
return
}
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
index bda77483806e..dc3cdc0c3026 100644
--- 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
@@ -53,16 +53,16 @@ func @spmv8x8(%AVAL: memref<4xvector<8xf32>>,
%f0 = constant 0.0 : f32
%mask = vector.constant_mask [8] : vector<8xi1>
%pass = vector.broadcast %f0 : f32 to vector<8xf32>
- %b = load %B[%c0] : memref<1xvector<8xf32>>
+ %b = memref.load %B[%c0] : memref<1xvector<8xf32>>
%b_out = scf.for %k = %c0 to %cn step %c1 iter_args(%b_iter = %b) -> (vector<8xf32>) {
- %aval = load %AVAL[%k] : memref<4xvector<8xf32>>
- %aidx = load %AIDX[%k] : memref<4xvector<8xi32>>
+ %aval = memref.load %AVAL[%k] : memref<4xvector<8xf32>>
+ %aidx = memref.load %AIDX[%k] : memref<4xvector<8xi32>>
%0 = vector.gather %X[%c0][%aidx], %mask, %pass
: memref<?xf32>, vector<8xi32>, vector<8xi1>, vector<8xf32> into vector<8xf32>
%b_new = vector.fma %aval, %0, %b_iter : vector<8xf32>
scf.yield %b_new : vector<8xf32>
}
- store %b_out, %B[%c0] : memref<1xvector<8xf32>>
+ memref.store %b_out, %B[%c0] : memref<1xvector<8xf32>>
return
}
@@ -100,10 +100,10 @@ func @entry() {
// Allocate.
//
- %AVAL = alloc() {alignment = 64} : memref<4xvector<8xf32>>
- %AIDX = alloc() {alignment = 64} : memref<4xvector<8xi32>>
- %X = alloc(%c8) {alignment = 64} : memref<?xf32>
- %B = alloc() {alignment = 64} : memref<1xvector<8xf32>>
+ %AVAL = memref.alloc() {alignment = 64} : memref<4xvector<8xf32>>
+ %AIDX = memref.alloc() {alignment = 64} : memref<4xvector<8xi32>>
+ %X = memref.alloc(%c8) {alignment = 64} : memref<?xf32>
+ %B = memref.alloc() {alignment = 64} : memref<1xvector<8xf32>>
//
// Initialize.
@@ -116,7 +116,7 @@ func @entry() {
%2 = vector.insert %f3, %1[5] : f32 into vector<8xf32>
%3 = vector.insert %f4, %2[6] : f32 into vector<8xf32>
%4 = vector.insert %f3, %3[7] : f32 into vector<8xf32>
- store %4, %AVAL[%c0] : memref<4xvector<8xf32>>
+ memref.store %4, %AVAL[%c0] : memref<4xvector<8xf32>>
%5 = vector.insert %f2, %vf1[0] : f32 into vector<8xf32>
%6 = vector.insert %f8, %5[1] : f32 into vector<8xf32>
@@ -124,15 +124,15 @@ func @entry() {
%8 = vector.insert %f2, %7[5] : f32 into vector<8xf32>
%9 = vector.insert %f7, %8[6] : f32 into vector<8xf32>
%10 = vector.insert %f2, %9[7] : f32 into vector<8xf32>
- store %10, %AVAL[%c1] : memref<4xvector<8xf32>>
+ memref.store %10, %AVAL[%c1] : memref<4xvector<8xf32>>
%11 = vector.insert %f3, %vf1[1] : f32 into vector<8xf32>
%12 = vector.insert %f6, %11[2] : f32 into vector<8xf32>
- store %12, %AVAL[%c2] : memref<4xvector<8xf32>>
+ memref.store %12, %AVAL[%c2] : memref<4xvector<8xf32>>
%13 = vector.insert %f2, %vf1[2] : f32 into vector<8xf32>
%14 = vector.insert %f2, %13[5] : f32 into vector<8xf32>
- store %14, %AVAL[%c3] : memref<4xvector<8xf32>>
+ memref.store %14, %AVAL[%c3] : memref<4xvector<8xf32>>
%vi0 = vector.broadcast %i0 : i32 to vector<8xi32>
@@ -140,7 +140,7 @@ func @entry() {
%21 = vector.insert %i1, %20[3] : i32 into vector<8xi32>
%22 = vector.insert %i1, %21[5] : i32 into vector<8xi32>
%23 = vector.insert %i1, %22[7] : i32 into vector<8xi32>
- store %23, %AIDX[%c0] : memref<4xvector<8xi32>>
+ memref.store %23, %AIDX[%c0] : memref<4xvector<8xi32>>
%24 = vector.insert %i2, %vi0[0] : i32 into vector<8xi32>
%25 = vector.insert %i1, %24[1] : i32 into vector<8xi32>
@@ -150,7 +150,7 @@ func @entry() {
%29 = vector.insert %i4, %28[5] : i32 into vector<8xi32>
%30 = vector.insert %i2, %29[6] : i32 into vector<8xi32>
%31 = vector.insert %i3, %30[7] : i32 into vector<8xi32>
- store %31, %AIDX[%c1] : memref<4xvector<8xi32>>
+ memref.store %31, %AIDX[%c1] : memref<4xvector<8xi32>>
%32 = vector.insert %i5, %vi0[0] : i32 into vector<8xi32>
%33 = vector.insert %i4, %32[1] : i32 into vector<8xi32>
@@ -160,7 +160,7 @@ func @entry() {
%37 = vector.insert %i5, %36[5] : i32 into vector<8xi32>
%38 = vector.insert %i4, %37[6] : i32 into vector<8xi32>
%39 = vector.insert %i6, %38[7] : i32 into vector<8xi32>
- store %39, %AIDX[%c2] : memref<4xvector<8xi32>>
+ memref.store %39, %AIDX[%c2] : memref<4xvector<8xi32>>
%40 = vector.insert %i7, %vi0[0] : i32 into vector<8xi32>
%41 = vector.insert %i6, %40[1] : i32 into vector<8xi32>
@@ -170,16 +170,16 @@ func @entry() {
%45 = vector.insert %i6, %44[5] : i32 into vector<8xi32>
%46 = vector.insert %i6, %45[6] : i32 into vector<8xi32>
%47 = vector.insert %i7, %46[7] : i32 into vector<8xi32>
- store %47, %AIDX[%c3] : memref<4xvector<8xi32>>
+ memref.store %47, %AIDX[%c3] : memref<4xvector<8xi32>>
%vf0 = vector.broadcast %f0 : f32 to vector<8xf32>
- store %vf0, %B[%c0] : memref<1xvector<8xf32>>
+ memref.store %vf0, %B[%c0] : memref<1xvector<8xf32>>
scf.for %i = %c0 to %c8 step %c1 {
%ix = addi %i, %c1 : index
%kx = index_cast %ix : index to i32
%fx = sitofp %kx : i32 to f32
- store %fx, %X[%i] : memref<?xf32>
+ memref.store %fx, %X[%i] : memref<?xf32>
}
//
@@ -196,16 +196,16 @@ func @entry() {
//
scf.for %i = %c0 to %c4 step %c1 {
- %aval = load %AVAL[%i] : memref<4xvector<8xf32>>
+ %aval = memref.load %AVAL[%i] : memref<4xvector<8xf32>>
vector.print %aval : vector<8xf32>
}
scf.for %i = %c0 to %c4 step %c1 {
- %aidx = load %AIDX[%i] : memref<4xvector<8xi32>>
+ %aidx = memref.load %AIDX[%i] : memref<4xvector<8xi32>>
vector.print %aidx : vector<8xi32>
}
- %ldb = load %B[%c0] : memref<1xvector<8xf32>>
+ %ldb = memref.load %B[%c0] : memref<1xvector<8xf32>>
vector.print %ldb : vector<8xf32>
//
@@ -226,10 +226,10 @@ func @entry() {
// Free.
//
- dealloc %AVAL : memref<4xvector<8xf32>>
- dealloc %AIDX : memref<4xvector<8xi32>>
- dealloc %X : memref<?xf32>
- dealloc %B : memref<1xvector<8xf32>>
+ memref.dealloc %AVAL : memref<4xvector<8xf32>>
+ memref.dealloc %AIDX : memref<4xvector<8xi32>>
+ memref.dealloc %X : memref<?xf32>
+ memref.dealloc %B : memref<1xvector<8xf32>>
return
}
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
index 3d259661d081..a731b929aaa3 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read-2d.mlir
@@ -33,7 +33,7 @@ func @entry() {
// work with dims of 4, not of 3
%first = constant 3: index
%second = constant 4: index
- %A = alloc(%first, %second) : memref<?x?xf32>
+ %A = memref.alloc(%first, %second) : memref<?x?xf32>
scf.for %i = %c0 to %first step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
@@ -42,7 +42,7 @@ func @entry() {
%j32 = index_cast %j : index to i32
%fj = sitofp %j32 : i32 to f32
%fres = addf %fi10, %fj : f32
- store %fres, %A[%i, %j] : memref<?x?xf32>
+ memref.store %fres, %A[%i, %j] : memref<?x?xf32>
}
}
// On input, memory contains [[ 0, 1, 2, ...], [10, 11, 12, ...], ...]
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir
index e6fa0df1ed7e..2bf57ecf2f7a 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-read.mlir
@@ -37,11 +37,11 @@ func @entry() {
%c3 = constant 3: index
%c4 = constant 4: index
%c5 = constant 5: index
- %A = alloc(%c5) : memref<?xf32>
+ %A = memref.alloc(%c5) : memref<?xf32>
scf.for %i = %c0 to %c5 step %c1 {
%i32 = index_cast %i : index to i32
%fi = sitofp %i32 : i32 to f32
- store %fi, %A[%i] : memref<?xf32>
+ memref.store %fi, %A[%i] : memref<?xf32>
}
// On input, memory contains [[ 0, 1, 2, 3, 4, xxx garbage xxx ]]
// Read shifted by 2 and pad with -42:
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
index 15d043f11e77..19bf28b57c3f 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-to-loops.mlir
@@ -13,14 +13,14 @@ func @alloc_2d_filled_f32(%arg0: index, %arg1: index) -> memref<?x?xf32> {
%c1 = constant 1 : index
%c10 = constant 10 : index
%c100 = constant 100 : index
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.for %arg5 = %c0 to %arg0 step %c1 {
scf.for %arg6 = %c0 to %arg1 step %c1 {
%arg66 = muli %arg6, %c100 : index
%tmp1 = addi %arg5, %arg66 : index
%tmp2 = index_cast %tmp1 : index to i32
%tmp3 = sitofp %tmp2 : i32 to f32
- store %tmp3, %0[%arg5, %arg6] : memref<?x?xf32>
+ memref.store %tmp3, %0[%arg5, %arg6] : memref<?x?xf32>
}
}
return %0 : memref<?x?xf32>
@@ -34,7 +34,7 @@ func @main() {
%c6 = constant 6 : index
%cst = constant -4.2e+01 : f32
%0 = call @alloc_2d_filled_f32(%c6, %c6) : (index, index) -> memref<?x?xf32>
- %converted = memref_cast %0 : memref<?x?xf32> to memref<*xf32>
+ %converted = memref.cast %0 : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%converted): (memref<*xf32>) -> ()
// CHECK: Unranked{{.*}}data =
// CHECK: [
@@ -100,6 +100,6 @@ func @main() {
vector.print %5 : vector<5xf32>
// CHECK-NEXT: ( 403, 503, 502, -42, -42 )
- dealloc %0 : memref<?x?xf32>
+ memref.dealloc %0 : memref<?x?xf32>
return
}
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir
index c61a1629dcfb..bb818e2ba591 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-transfer-write.mlir
@@ -43,10 +43,10 @@ func @entry() {
%c0 = constant 0: index
%c1 = constant 1: index
%c32 = constant 32: index
- %A = alloc(%c32) {alignment=64} : memref<?xf32>
+ %A = memref.alloc(%c32) {alignment=64} : memref<?xf32>
scf.for %i = %c0 to %c32 step %c1 {
%f = constant 0.0: f32
- store %f, %A[%i] : memref<?xf32>
+ memref.store %f, %A[%i] : memref<?xf32>
}
// On input, memory contains all zeros.
diff --git a/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir b/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir
index a6c053a5b014..a43cf3c4bfa6 100644
--- a/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir
+++ b/mlir/test/Integration/Dialect/Vector/CPU/test-vector-distribute.mlir
@@ -18,12 +18,12 @@ func private @print_memref_f32(memref<*xf32>)
func @alloc_1d_filled_inc_f32(%arg0: index, %arg1: f32) -> memref<?xf32> {
%c0 = constant 0 : index
%c1 = constant 1 : index
- %0 = alloc(%arg0) : memref<?xf32>
+ %0 = memref.alloc(%arg0) : memref<?xf32>
scf.for %arg2 = %c0 to %arg0 step %c1 {
%tmp = index_cast %arg2 : index to i32
%tmp1 = sitofp %tmp : i32 to f32
%tmp2 = addf %tmp1, %arg1 : f32
- store %tmp2, %0[%arg2] : memref<?xf32>
+ memref.store %tmp2, %0[%arg2] : memref<?xf32>
}
return %0 : memref<?xf32>
}
@@ -37,7 +37,7 @@ func @main() {
%c1 = constant 1 : index
%c32 = constant 32 : index
%c64 = constant 64 : index
- %out = alloc(%c64) : memref<?xf32>
+ %out = memref.alloc(%c64) : memref<?xf32>
%in1 = call @alloc_1d_filled_inc_f32(%c64, %cf1) : (index, f32) -> memref<?xf32>
%in2 = call @alloc_1d_filled_inc_f32(%c64, %cf2) : (index, f32) -> memref<?xf32>
// Check that the tansformatio correctly happened.
@@ -51,7 +51,7 @@ func @main() {
%b = vector.transfer_read %in2[%c0], %cf0: memref<?xf32>, vector<64xf32>
%acc = addf %a, %b: vector<64xf32>
vector.transfer_write %acc, %out[%c0]: vector<64xf32>, memref<?xf32>
- %converted = memref_cast %out : memref<?xf32> to memref<*xf32>
+ %converted = memref.cast %out : memref<?xf32> to memref<*xf32>
call @print_memref_f32(%converted): (memref<*xf32>) -> ()
// CHECK: Unranked{{.*}}data =
// CHECK: [
@@ -61,8 +61,8 @@ func @main() {
// CHECK-SAME: 77, 79, 81, 83, 85, 87, 89, 91, 93, 95, 97, 99,
// CHECK-SAME: 101, 103, 105, 107, 109, 111, 113, 115, 117, 119,
// CHECK-SAME: 121, 123, 125, 127, 129]
- dealloc %out : memref<?xf32>
- dealloc %in1 : memref<?xf32>
- dealloc %in2 : memref<?xf32>
+ memref.dealloc %out : memref<?xf32>
+ memref.dealloc %in1 : memref<?xf32>
+ memref.dealloc %in2 : memref<?xf32>
return
}
diff --git a/mlir/test/Integration/Sparse/CPU/frostt-example.mlir b/mlir/test/Integration/Sparse/CPU/frostt-example.mlir
index bfda9d62159d..ed204d2d2649 100644
--- a/mlir/test/Integration/Sparse/CPU/frostt-example.mlir
+++ b/mlir/test/Integration/Sparse/CPU/frostt-example.mlir
@@ -35,8 +35,8 @@ module {
// Setup memrefs to get meta data, indices and values.
// The index array should provide sufficient space.
//
- %idata = alloc(%c10) : memref<?xindex>
- %ddata = alloc(%c1) : memref<?xf64>
+ %idata = memref.alloc(%c10) : memref<?xindex>
+ %ddata = memref.alloc(%c1) : memref<?xf64>
//
// Obtain the sparse tensor filename through this test helper.
@@ -55,12 +55,12 @@ module {
//
// Print some meta data.
//
- %rank = load %idata[%c0] : memref<?xindex>
- %nnz = load %idata[%c1] : memref<?xindex>
+ %rank = memref.load %idata[%c0] : memref<?xindex>
+ %nnz = memref.load %idata[%c1] : memref<?xindex>
vector.print %rank : index
vector.print %nnz : index
scf.for %r = %c2 to %c10 step %c1 {
- %d = load %idata[%r] : memref<?xindex>
+ %d = memref.load %idata[%r] : memref<?xindex>
vector.print %d : index
}
@@ -77,13 +77,13 @@ module {
//
%0 = vector.broadcast %i0 : i64 to vector<8xi64>
%1 = scf.for %r = %c0 to %rank step %c1 iter_args(%in = %0) -> vector<8xi64> {
- %i = load %idata[%r] : memref<?xindex>
+ %i = memref.load %idata[%r] : memref<?xindex>
%ii = index_cast %i : index to i64
%ri = index_cast %r : index to i32
%out = vector.insertelement %ii, %in[%ri : i32] : vector<8xi64>
scf.yield %out : vector<8xi64>
}
- %2 = load %ddata[%c0] : memref<?xf64>
+ %2 = memref.load %ddata[%c0] : memref<?xf64>
vector.print %1 : vector<8xi64>
vector.print %2 : f64
}
@@ -145,8 +145,8 @@ module {
//
// Free.
//
- dealloc %idata : memref<?xindex>
- dealloc %ddata : memref<?xf64>
+ memref.dealloc %idata : memref<?xindex>
+ memref.dealloc %ddata : memref<?xf64>
return
}
diff --git a/mlir/test/Integration/Sparse/CPU/matrix-market-example.mlir b/mlir/test/Integration/Sparse/CPU/matrix-market-example.mlir
index 8e90797703a3..719340354e2f 100644
--- a/mlir/test/Integration/Sparse/CPU/matrix-market-example.mlir
+++ b/mlir/test/Integration/Sparse/CPU/matrix-market-example.mlir
@@ -35,8 +35,8 @@ module {
//
// Setup memrefs to get meta data, indices, and values.
//
- %idata = alloc(%c4) : memref<?xindex>
- %ddata = alloc(%c1) : memref<?xf64>
+ %idata = memref.alloc(%c4) : memref<?xindex>
+ %ddata = memref.alloc(%c1) : memref<?xf64>
//
// Obtain the sparse matrix filename through this test helper.
@@ -51,20 +51,20 @@ module {
// nonzero elements (nnz), and the size (m x n) through a memref array.
//
%tensor = call @openTensor(%fileName, %idata) : (!Filename, memref<?xindex>) -> (!Tensor)
- %rank = load %idata[%c0] : memref<?xindex>
- %nnz = load %idata[%c1] : memref<?xindex>
- %m = load %idata[%c2] : memref<?xindex>
- %n = load %idata[%c3] : memref<?xindex>
+ %rank = memref.load %idata[%c0] : memref<?xindex>
+ %nnz = memref.load %idata[%c1] : memref<?xindex>
+ %m = memref.load %idata[%c2] : memref<?xindex>
+ %n = memref.load %idata[%c3] : memref<?xindex>
//
// At this point, code should prepare a proper sparse storage scheme for
// an m x n matrix with nnz nonzero elements. For simplicity, here we
// simply intialize a dense m x n matrix to all zeroes.
//
- %a = alloc(%m, %n) : memref<?x?xf64>
+ %a = memref.alloc(%m, %n) : memref<?x?xf64>
scf.for %ii = %c0 to %m step %c1 {
scf.for %jj = %c0 to %n step %c1 {
- store %d0, %a[%ii, %jj] : memref<?x?xf64>
+ memref.store %d0, %a[%ii, %jj] : memref<?x?xf64>
}
}
@@ -75,10 +75,10 @@ module {
//
scf.for %k = %c0 to %nnz step %c1 {
call @readTensorItem(%tensor, %idata, %ddata) : (!Tensor, memref<?xindex>, memref<?xf64>) -> ()
- %i = load %idata[%c0] : memref<?xindex>
- %j = load %idata[%c1] : memref<?xindex>
- %d = load %ddata[%c0] : memref<?xf64>
- store %d, %a[%i, %j] : memref<?x?xf64>
+ %i = memref.load %idata[%c0] : memref<?xindex>
+ %j = memref.load %idata[%c1] : memref<?xindex>
+ %d = memref.load %ddata[%c0] : memref<?xf64>
+ memref.store %d, %a[%i, %j] : memref<?x?xf64>
}
//
@@ -112,9 +112,9 @@ module {
//
// Free.
//
- dealloc %idata : memref<?xindex>
- dealloc %ddata : memref<?xf64>
- dealloc %a : memref<?x?xf64>
+ memref.dealloc %idata : memref<?xindex>
+ memref.dealloc %ddata : memref<?xf64>
+ memref.dealloc %a : memref<?x?xf64>
return
}
diff --git a/mlir/test/Integration/Sparse/CPU/sparse_sampled_matmul.mlir b/mlir/test/Integration/Sparse/CPU/sparse_sampled_matmul.mlir
index 5f2b199fb704..a1070db5d577 100644
--- a/mlir/test/Integration/Sparse/CPU/sparse_sampled_matmul.mlir
+++ b/mlir/test/Integration/Sparse/CPU/sparse_sampled_matmul.mlir
@@ -82,32 +82,32 @@ module {
// Mark both dimensions of the matrix as sparse and encode the
// storage scheme types (this must match the metadata in the
// trait and compiler switches).
- %annotations = alloc(%c2) : memref<?xi1>
+ %annotations = memref.alloc(%c2) : memref<?xi1>
%sparse = constant true
- store %sparse, %annotations[%c0] : memref<?xi1>
- store %sparse, %annotations[%c1] : memref<?xi1>
+ memref.store %sparse, %annotations[%c0] : memref<?xi1>
+ memref.store %sparse, %annotations[%c1] : memref<?xi1>
%i32 = constant 3 : index
%f32 = constant 1 : index
// Setup memory for the dense matrices and initialize.
- %adata = alloc(%c5, %c10) : memref<?x?xf32>
- %bdata = alloc(%c10, %c5) : memref<?x?xf32>
- %xdata = alloc(%c5, %c5) : memref<?x?xf32>
+ %adata = memref.alloc(%c5, %c10) : memref<?x?xf32>
+ %bdata = memref.alloc(%c10, %c5) : memref<?x?xf32>
+ %xdata = memref.alloc(%c5, %c5) : memref<?x?xf32>
scf.for %i = %c0 to %c5 step %c1 {
scf.for %j = %c0 to %c5 step %c1 {
- store %d0, %xdata[%i, %j] : memref<?x?xf32>
+ memref.store %d0, %xdata[%i, %j] : memref<?x?xf32>
}
%p = addi %i, %c1 : index
%q = index_cast %p : index to i32
%d = sitofp %q : i32 to f32
scf.for %j = %c0 to %c10 step %c1 {
- store %d, %adata[%i, %j] : memref<?x?xf32>
- store %d, %bdata[%j, %i] : memref<?x?xf32>
+ memref.store %d, %adata[%i, %j] : memref<?x?xf32>
+ memref.store %d, %bdata[%j, %i] : memref<?x?xf32>
}
}
- %a = tensor_load %adata : memref<?x?xf32>
- %b = tensor_load %bdata : memref<?x?xf32>
- %x = tensor_load %xdata : memref<?x?xf32>
+ %a = memref.tensor_load %adata : memref<?x?xf32>
+ %b = memref.tensor_load %bdata : memref<?x?xf32>
+ %x = memref.tensor_load %xdata : memref<?x?xf32>
// Read the sparse matrix from file, construct sparse storage
// according to <sparse,sparse> in memory, and call the kernel.
@@ -125,7 +125,7 @@ module {
// CHECK: ( 164, 0, 0, 640, 0 )
// CHECK: ( 0, 520, 0, 0, 1250 )
//
- %r = tensor_to_memref %0 : memref<?x?xf32>
+ %r = memref.buffer_cast %0 : memref<?x?xf32>
scf.for %i = %c0 to %c5 step %c1 {
%v = vector.transfer_read %r[%i, %c0], %d0: memref<?x?xf32>, vector<5xf32>
vector.print %v : vector<5xf32>
@@ -133,9 +133,9 @@ module {
// Release the resources.
call @delSparseTensor(%s) : (!SparseTensor) -> ()
- dealloc %adata : memref<?x?xf32>
- dealloc %bdata : memref<?x?xf32>
- dealloc %xdata : memref<?x?xf32>
+ memref.dealloc %adata : memref<?x?xf32>
+ memref.dealloc %bdata : memref<?x?xf32>
+ memref.dealloc %xdata : memref<?x?xf32>
return
}
diff --git a/mlir/test/Integration/Sparse/CPU/sparse_sum.mlir b/mlir/test/Integration/Sparse/CPU/sparse_sum.mlir
index 78e5940eae60..10b0981dc58f 100644
--- a/mlir/test/Integration/Sparse/CPU/sparse_sum.mlir
+++ b/mlir/test/Integration/Sparse/CPU/sparse_sum.mlir
@@ -72,18 +72,18 @@ module {
// Mark both dimensions of the matrix as sparse and encode the
// storage scheme types (this must match the metadata in the
// trait and compiler switches).
- %annotations = alloc(%c2) : memref<?xi1>
+ %annotations = memref.alloc(%c2) : memref<?xi1>
%sparse = constant true
- store %sparse, %annotations[%c0] : memref<?xi1>
- store %sparse, %annotations[%c1] : memref<?xi1>
+ memref.store %sparse, %annotations[%c0] : memref<?xi1>
+ memref.store %sparse, %annotations[%c1] : memref<?xi1>
%i64 = constant 2 : index
%f64 = constant 0 : index
// Setup memory for a single reduction scalar,
// initialized to zero.
- %xdata = alloc() : memref<f64>
- store %d0, %xdata[] : memref<f64>
- %x = tensor_load %xdata : memref<f64>
+ %xdata = memref.alloc() : memref<f64>
+ memref.store %d0, %xdata[] : memref<f64>
+ %x = memref.tensor_load %xdata : memref<f64>
// Read the sparse matrix from file, construct sparse storage
// according to <sparse,sparse> in memory, and call the kernel.
@@ -97,13 +97,13 @@ module {
//
// CHECK: 28.2
//
- %m = tensor_to_memref %0 : memref<f64>
- %v = load %m[] : memref<f64>
+ %m = memref.buffer_cast %0 : memref<f64>
+ %v = memref.load %m[] : memref<f64>
vector.print %v : f64
// Release the resources.
call @delSparseTensor(%a) : (!SparseTensor) -> ()
- dealloc %xdata : memref<f64>
+ memref.dealloc %xdata : memref<f64>
return
}
diff --git a/mlir/test/Transforms/buffer-deallocation.mlir b/mlir/test/Transforms/buffer-deallocation.mlir
index 4965324e2aea..25197d14fba7 100644
--- a/mlir/test/Transforms/buffer-deallocation.mlir
+++ b/mlir/test/Transforms/buffer-deallocation.mlir
@@ -21,7 +21,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -30,17 +30,17 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
}
// CHECK-NEXT: cond_br
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
// CHECK-NEXT: br ^bb3(%[[ALLOC0]]
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK-NEXT: br ^bb3(%[[ALLOC2]]
// CHECK: test.copy
-// CHECK-NEXT: dealloc
+// CHECK-NEXT: memref.dealloc
// CHECK-NEXT: return
// -----
@@ -68,7 +68,7 @@ func @condBranchDynamicType(
^bb1:
br ^bb3(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
br ^bb3(%1 : memref<?xf32>)
^bb3(%2: memref<?xf32>):
@@ -77,21 +77,21 @@ func @condBranchDynamicType(
}
// CHECK-NEXT: cond_br
-// CHECK: %[[DIM0:.*]] = dim
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[DIM0]])
+// CHECK: %[[DIM0:.*]] = memref.dim
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[DIM0]])
// CHECK-NEXT: linalg.copy(%{{.*}}, %[[ALLOC0]])
// CHECK-NEXT: br ^bb3(%[[ALLOC0]]
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[DIM1:.*]] = dim %[[ALLOC1]]
-// CHECK-NEXT: %[[ALLOC2:.*]] = alloc(%[[DIM1]])
+// CHECK: %[[DIM1:.*]] = memref.dim %[[ALLOC1]]
+// CHECK-NEXT: %[[ALLOC2:.*]] = memref.alloc(%[[DIM1]])
// CHECK-NEXT: linalg.copy(%[[ALLOC1]], %[[ALLOC2]])
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK-NEXT: br ^bb3
// CHECK-NEXT: ^bb3(%[[ALLOC3:.*]]:{{.*}})
// CHECK: test.copy(%[[ALLOC3]],
-// CHECK-NEXT: dealloc %[[ALLOC3]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC3]]
// CHECK-NEXT: return
// -----
@@ -126,7 +126,7 @@ func @condBranchDynamicTypeNested(
^bb1:
br ^bb6(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
cond_br %arg0, ^bb3, ^bb4
^bb3:
@@ -144,12 +144,12 @@ func @condBranchDynamicTypeNested(
// CHECK-NEXT: cond_br
// CHECK: ^bb1
-// CHECK: %[[DIM0:.*]] = dim
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[DIM0]])
+// CHECK: %[[DIM0:.*]] = memref.dim
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[DIM0]])
// CHECK-NEXT: linalg.copy(%{{.*}}, %[[ALLOC0]])
// CHECK-NEXT: br ^bb6
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
// CHECK: cond_br
// CHECK: ^bb3:
@@ -157,16 +157,16 @@ func @condBranchDynamicTypeNested(
// CHECK: ^bb4:
// CHECK-NEXT: br ^bb5(%[[ALLOC1]]{{.*}})
// CHECK-NEXT: ^bb5(%[[ALLOC2:.*]]:{{.*}})
-// CHECK: %[[DIM2:.*]] = dim %[[ALLOC2]]
-// CHECK-NEXT: %[[ALLOC3:.*]] = alloc(%[[DIM2]])
+// CHECK: %[[DIM2:.*]] = memref.dim %[[ALLOC2]]
+// CHECK-NEXT: %[[ALLOC3:.*]] = memref.alloc(%[[DIM2]])
// CHECK-NEXT: linalg.copy(%[[ALLOC2]], %[[ALLOC3]])
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK-NEXT: br ^bb6(%[[ALLOC3]]{{.*}})
// CHECK-NEXT: ^bb6(%[[ALLOC4:.*]]:{{.*}})
// CHECK-NEXT: br ^bb7(%[[ALLOC4]]{{.*}})
// CHECK-NEXT: ^bb7(%[[ALLOC5:.*]]:{{.*}})
// CHECK: test.copy(%[[ALLOC5]],
-// CHECK-NEXT: dealloc %[[ALLOC4]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC4]]
// CHECK-NEXT: return
// -----
@@ -177,11 +177,11 @@ func @condBranchDynamicTypeNested(
// CHECK-LABEL: func @emptyUsesValue
func @emptyUsesValue(%arg0: memref<4xf32>) {
- %0 = alloc() : memref<4xf32>
+ %0 = memref.alloc() : memref<4xf32>
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
-// CHECK-NEXT: dealloc %[[ALLOC]]
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
+// CHECK-NEXT: memref.dealloc %[[ALLOC]]
// CHECK-NEXT: return
// -----
@@ -200,7 +200,7 @@ func @emptyUsesValue(%arg0: memref<4xf32>) {
func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
cond_br %arg0, ^bb1, ^bb2(%arg1 : memref<2xf32>)
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb2(%0 : memref<2xf32>)
^bb2(%1: memref<2xf32>):
@@ -208,16 +208,16 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
// CHECK-NEXT: cond_br
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK: test.copy
-// CHECK-NEXT: dealloc
+// CHECK-NEXT: memref.dealloc
// CHECK-NEXT: return
// -----
@@ -233,7 +233,7 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @invCriticalEdge
func @invCriticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0, ^bb1, ^bb2(%arg1 : memref<2xf32>)
^bb1:
@@ -261,7 +261,7 @@ func @invCriticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElse
func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -271,19 +271,19 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb2(%3: memref<2xf32>, %4: memref<2xf32>):
br ^bb3(%3, %4 : memref<2xf32>, memref<2xf32>)
^bb3(%5: memref<2xf32>, %6: memref<2xf32>):
- %7 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
test.buffer_based in(%5: memref<2xf32>) out(%7: memref<2xf32>)
test.copy(%7, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[SECOND_ALLOC:.*]] = alloc()
+// CHECK: %[[SECOND_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: dealloc %[[FIRST_ALLOC]]
+// CHECK: memref.dealloc %[[FIRST_ALLOC]]
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[SECOND_ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[SECOND_ALLOC]]
// CHECK-NEXT: return
// -----
@@ -299,7 +299,7 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNoUsers
func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -313,9 +313,9 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = memref.alloc()
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[FIRST_ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[FIRST_ALLOC]]
// CHECK-NEXT: return
// -----
@@ -334,7 +334,7 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNested
func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -348,19 +348,19 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb4(%6: memref<2xf32>):
br ^bb5(%3, %6 : memref<2xf32>, memref<2xf32>)
^bb5(%7: memref<2xf32>, %8: memref<2xf32>):
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%9: memref<2xf32>)
test.copy(%9, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[SECOND_ALLOC:.*]] = alloc()
+// CHECK: %[[SECOND_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: dealloc %[[FIRST_ALLOC]]
+// CHECK: memref.dealloc %[[FIRST_ALLOC]]
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[SECOND_ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[SECOND_ALLOC]]
// CHECK-NEXT: return
// -----
@@ -371,17 +371,17 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @redundantOperations
func @redundantOperations(%arg0: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%0: memref<2xf32>) out(%1: memref<2xf32>)
return
}
// CHECK: (%[[ARG0:.*]]: {{.*}})
-// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[FIRST_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based in(%[[ARG0]]{{.*}}out(%[[FIRST_ALLOC]]
-// CHECK: %[[SECOND_ALLOC:.*]] = alloc()
+// CHECK: %[[SECOND_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based in(%[[FIRST_ALLOC]]{{.*}}out(%[[SECOND_ALLOC]]
// CHECK: dealloc
// CHECK-NEXT: dealloc
@@ -407,11 +407,11 @@ func @moving_alloc_and_inserting_missing_dealloc(
%arg1: memref<2xf32>) {
cond_br %cond, ^bb1, ^bb2
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
br ^exit(%0 : memref<2xf32>)
^bb2:
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%1: memref<2xf32>)
br ^exit(%1 : memref<2xf32>)
^exit(%arg2: memref<2xf32>):
@@ -422,22 +422,22 @@ func @moving_alloc_and_inserting_missing_dealloc(
// CHECK-NEXT: cond_br
// CHECK: ^bb1
// CHECK: ^bb1
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
-// CHECK-NEXT: dealloc %[[ALLOC0]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
// CHECK-NEXT: br ^bb3(%[[ALLOC1]]
// CHECK-NEXT: ^bb2
-// CHECK-NEXT: %[[ALLOC2:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC3:.*]] = alloc()
+// CHECK: %[[ALLOC3:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
-// CHECK-NEXT: dealloc %[[ALLOC2]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC2]]
// CHECK-NEXT: br ^bb3(%[[ALLOC3]]
// CHECK-NEXT: ^bb3(%[[ALLOC4:.*]]:{{.*}})
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC4]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC4]]
// CHECK-NEXT: return
// -----
@@ -457,23 +457,23 @@ func @moving_invalid_dealloc_op_complex(
%cond: i1,
%arg0: memref<2xf32>,
%arg1: memref<2xf32>) {
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
cond_br %cond, ^bb1, ^bb2
^bb1:
br ^exit(%arg0 : memref<2xf32>)
^bb2:
test.buffer_based in(%arg0: memref<2xf32>) out(%1: memref<2xf32>)
- dealloc %1 : memref<2xf32>
+ memref.dealloc %1 : memref<2xf32>
br ^exit(%1 : memref<2xf32>)
^exit(%arg2: memref<2xf32>):
test.copy(%arg2, %arg1) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: cond_br
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC0]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
// CHECK-NEXT: return
// -----
@@ -484,15 +484,15 @@ func @moving_invalid_dealloc_op_complex(
func @inserting_missing_dealloc_simple(
%arg0 : memref<2xf32>,
%arg1: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
test.copy(%0, %arg1) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC0]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
// -----
@@ -501,16 +501,16 @@ func @inserting_missing_dealloc_simple(
// CHECK-LABEL: func @moving_invalid_dealloc_op
func @moving_invalid_dealloc_op(%arg0 : memref<2xf32>, %arg1: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
- dealloc %0 : memref<2xf32>
+ memref.dealloc %0 : memref<2xf32>
test.copy(%0, %arg1) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC0]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
// -----
@@ -530,10 +530,10 @@ func @nested_regions_and_cond_branch(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -545,18 +545,18 @@ func @nested_regions_and_cond_branch(
}
// CHECK: (%[[cond:.*]]: {{.*}}, %[[ARG1:.*]]: {{.*}}, %{{.*}}: {{.*}})
// CHECK-NEXT: cond_br %[[cond]], ^[[BB1:.*]], ^[[BB2:.*]]
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ARG1]], %[[ALLOC0]])
// CHECK: ^[[BB2]]:
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.region_buffer_based in(%[[ARG1]]{{.*}}out(%[[ALLOC1]]
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based in(%[[ARG1]]{{.*}}out(%[[ALLOC2]]
-// CHECK: dealloc %[[ALLOC2]]
+// CHECK: memref.dealloc %[[ALLOC2]]
// CHECK-NEXT: %{{.*}} = math.exp
-// CHECK: %[[ALLOC3:.*]] = alloc()
+// CHECK: %[[ALLOC3:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC1]], %[[ALLOC3]])
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK: ^[[BB3:.*]]({{.*}}):
// CHECK: test.copy
// CHECK-NEXT: dealloc
@@ -573,18 +573,18 @@ func @memref_in_function_results(
%arg0: memref<5xf32>,
%arg1: memref<10xf32>,
%arg2: memref<5xf32>) -> (memref<10xf32>, memref<15xf32>) {
- %x = alloc() : memref<15xf32>
- %y = alloc() : memref<5xf32>
+ %x = memref.alloc() : memref<15xf32>
+ %y = memref.alloc() : memref<5xf32>
test.buffer_based in(%arg0: memref<5xf32>) out(%y: memref<5xf32>)
test.copy(%y, %arg2) : (memref<5xf32>, memref<5xf32>)
return %arg1, %x : memref<10xf32>, memref<15xf32>
}
// CHECK: (%[[ARG0:.*]]: memref<5xf32>, %[[ARG1:.*]]: memref<10xf32>,
// CHECK-SAME: %[[RESULT:.*]]: memref<5xf32>)
-// CHECK: %[[X:.*]] = alloc()
-// CHECK: %[[Y:.*]] = alloc()
+// CHECK: %[[X:.*]] = memref.alloc()
+// CHECK: %[[Y:.*]] = memref.alloc()
// CHECK: test.copy
-// CHECK: dealloc %[[Y]]
+// CHECK: memref.dealloc %[[Y]]
// CHECK: return %[[ARG1]], %[[X]]
// -----
@@ -599,21 +599,21 @@ func @nested_region_control_flow(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %1 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.if
// CHECK: scf.yield %[[ALLOC0]]
-// CHECK: %[[ALLOC2:.*]] = alloc(%arg0, %arg1)
-// CHECK-NEXT: dealloc %[[ALLOC2]]
+// CHECK: %[[ALLOC2:.*]] = memref.alloc(%arg0, %arg1)
+// CHECK-NEXT: memref.dealloc %[[ALLOC2]]
// CHECK-NEXT: scf.yield %[[ALLOC0]]
// CHECK: return %[[ALLOC1]]
@@ -629,27 +629,27 @@ func @nested_region_control_flow_div(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %3 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.if
-// CHECK: %[[ALLOC2:.*]] = alloc
+// CHECK: %[[ALLOC2:.*]] = memref.alloc
// CHECK-NEXT: linalg.copy(%[[ALLOC0]], %[[ALLOC2]])
// CHECK: scf.yield %[[ALLOC2]]
-// CHECK: %[[ALLOC3:.*]] = alloc(%arg0, %arg1)
-// CHECK: %[[ALLOC4:.*]] = alloc
+// CHECK: %[[ALLOC3:.*]] = memref.alloc(%arg0, %arg1)
+// CHECK: %[[ALLOC4:.*]] = memref.alloc
// CHECK-NEXT: linalg.copy(%[[ALLOC3]], %[[ALLOC4]])
-// CHECK: dealloc %[[ALLOC3]]
+// CHECK: memref.dealloc %[[ALLOC3]]
// CHECK: scf.yield %[[ALLOC4]]
-// CHECK: dealloc %[[ALLOC0]]
+// CHECK: memref.dealloc %[[ALLOC0]]
// CHECK-NEXT: return %[[ALLOC1]]
// -----
@@ -660,7 +660,7 @@ func @nested_region_control_flow_div(
// CHECK-LABEL: func @inner_region_control_flow
func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
- %0 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%1 = test.region_if %0 : memref<?x?xf32> -> (memref<?x?xf32>) then {
^bb0(%arg1 : memref<?x?xf32>):
test.region_if_yield %arg1 : memref<?x?xf32>
@@ -674,7 +674,7 @@ func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
return %1 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %[[ALLOC1:.*]] = test.region_if
// CHECK-NEXT: ^bb0(%[[ALLOC2:.*]]:{{.*}}):
// CHECK-NEXT: test.region_if_yield %[[ALLOC2]]
@@ -688,8 +688,8 @@ func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
// CHECK-LABEL: func @subview
func @subview(%arg0 : index, %arg1 : index, %arg2 : memref<?x?xf32>) {
- %0 = alloc() : memref<64x4xf32, offset: 0, strides: [4, 1]>
- %1 = subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
+ %0 = memref.alloc() : memref<64x4xf32, offset: 0, strides: [4, 1]>
+ %1 = memref.subview %0[%arg0, %arg1][%arg0, %arg1][%arg0, %arg1] :
memref<64x4xf32, offset: 0, strides: [4, 1]>
to memref<?x?xf32, offset: ?, strides: [?, ?]>
test.copy(%1, %arg2) :
@@ -697,10 +697,10 @@ func @subview(%arg0 : index, %arg1 : index, %arg2 : memref<?x?xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
-// CHECK-NEXT: subview
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
+// CHECK-NEXT: memref.subview
// CHECK-NEXT: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC]]
// CHECK-NEXT: return
// -----
@@ -714,7 +714,7 @@ func @condBranchAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloca() : memref<2xf32>
+ %0 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -723,7 +723,7 @@ func @condBranchAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
}
// CHECK-NEXT: cond_br
-// CHECK: %[[ALLOCA:.*]] = alloca()
+// CHECK: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: br ^bb3(%[[ALLOCA:.*]])
// CHECK-NEXT: ^bb3
// CHECK-NEXT: test.copy
@@ -737,7 +737,7 @@ func @condBranchAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseAlloca
func @ifElseAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -747,17 +747,17 @@ func @ifElseAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb2(%3: memref<2xf32>, %4: memref<2xf32>):
br ^bb3(%3, %4 : memref<2xf32>, memref<2xf32>)
^bb3(%5: memref<2xf32>, %6: memref<2xf32>):
- %7 = alloca() : memref<2xf32>
+ %7 = memref.alloca() : memref<2xf32>
test.buffer_based in(%5: memref<2xf32>) out(%7: memref<2xf32>)
test.copy(%7, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOCA:.*]] = alloca()
+// CHECK: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
-// CHECK: dealloc %[[ALLOC]]
+// CHECK: memref.dealloc %[[ALLOC]]
// CHECK: test.copy
// CHECK-NEXT: return
@@ -768,7 +768,7 @@ func @ifElseNestedAlloca(
%arg0: i1,
%arg1: memref<2xf32>,
%arg2: memref<2xf32>) {
- %0 = alloca() : memref<2xf32>
+ %0 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -782,18 +782,18 @@ func @ifElseNestedAlloca(
^bb4(%6: memref<2xf32>):
br ^bb5(%3, %6 : memref<2xf32>, memref<2xf32>)
^bb5(%7: memref<2xf32>, %8: memref<2xf32>):
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%9: memref<2xf32>)
test.copy(%9, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC:.*]] = alloc()
+// CHECK: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// CHECK: test.copy
-// CHECK-NEXT: dealloc %[[ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC]]
// CHECK-NEXT: return
// -----
@@ -807,10 +807,10 @@ func @nestedRegionsAndCondBranchAlloca(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloca() : memref<2xf32>
+ %1 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -823,17 +823,17 @@ func @nestedRegionsAndCondBranchAlloca(
// CHECK: (%[[cond:.*]]: {{.*}}, %[[ARG1:.*]]: {{.*}}, %{{.*}}: {{.*}})
// CHECK-NEXT: cond_br %[[cond]], ^[[BB1:.*]], ^[[BB2:.*]]
// CHECK: ^[[BB1]]:
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
// CHECK: ^[[BB2]]:
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.region_buffer_based in(%[[ARG1]]{{.*}}out(%[[ALLOC1]]
-// CHECK: %[[ALLOCA:.*]] = alloca()
+// CHECK: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based in(%[[ARG1]]{{.*}}out(%[[ALLOCA]]
// CHECK: %{{.*}} = math.exp
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy
-// CHECK-NEXT: dealloc %[[ALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC1]]
// CHECK: ^[[BB3:.*]]({{.*}}):
// CHECK: test.copy
// CHECK-NEXT: dealloc
@@ -845,20 +845,20 @@ func @nestedRegionControlFlowAlloca(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloca(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloca(%arg0, %arg1) : memref<?x?xf32>
scf.yield %1 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.if
// CHECK: scf.yield %[[ALLOC0]]
-// CHECK: %[[ALLOCA:.*]] = alloca(%arg0, %arg1)
+// CHECK: %[[ALLOCA:.*]] = memref.alloca(%arg0, %arg1)
// CHECK-NEXT: scf.yield %[[ALLOC0]]
// CHECK: return %[[ALLOC1]]
@@ -875,33 +875,33 @@ func @loop_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %3 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
-// CHECK-NEXT: dealloc %[[ALLOC0]]
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc()
// CHECK: linalg.copy(%arg3, %[[ALLOC1]])
// CHECK: %[[ALLOC2:.*]] = scf.for {{.*}} iter_args
// CHECK-SAME: (%[[IALLOC:.*]] = %[[ALLOC1]]
// CHECK: cmpi
-// CHECK: dealloc %[[IALLOC]]
-// CHECK: %[[ALLOC3:.*]] = alloc()
-// CHECK: %[[ALLOC4:.*]] = alloc()
+// CHECK: memref.dealloc %[[IALLOC]]
+// CHECK: %[[ALLOC3:.*]] = memref.alloc()
+// CHECK: %[[ALLOC4:.*]] = memref.alloc()
// CHECK: linalg.copy(%[[ALLOC3]], %[[ALLOC4]])
-// CHECK: dealloc %[[ALLOC3]]
+// CHECK: memref.dealloc %[[ALLOC3]]
// CHECK: scf.yield %[[ALLOC4]]
// CHECK: }
// CHECK: test.copy(%[[ALLOC2]], %arg4)
-// CHECK-NEXT: dealloc %[[ALLOC2]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC2]]
// -----
@@ -918,7 +918,7 @@ func @loop_nested_if_no_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
@@ -933,14 +933,14 @@ func @loop_nested_if_no_alloc(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.for {{.*}} iter_args(%[[IALLOC:.*]] =
// CHECK: %[[ALLOC2:.*]] = scf.if
// CHECK: scf.yield %[[ALLOC0]]
// CHECK: scf.yield %[[IALLOC]]
// CHECK: scf.yield %[[ALLOC2]]
// CHECK: test.copy(%[[ALLOC1]], %arg4)
-// CHECK: dealloc %[[ALLOC0]]
+// CHECK: memref.dealloc %[[ALLOC0]]
// -----
@@ -958,12 +958,12 @@ func @loop_nested_if_alloc(
%ub: index,
%step: index,
%buf: memref<2xf32>) -> memref<2xf32> {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
%3 = scf.if %2 -> (memref<2xf32>) {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
scf.yield %4 : memref<2xf32>
} else {
scf.yield %0 : memref<2xf32>
@@ -973,30 +973,30 @@ func @loop_nested_if_alloc(
return %1 : memref<2xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%arg3, %[[ALLOC1]])
// CHECK-NEXT: %[[ALLOC2:.*]] = scf.for {{.*}} iter_args
// CHECK-SAME: (%[[IALLOC:.*]] = %[[ALLOC1]]
-// CHECK: dealloc %[[IALLOC]]
+// CHECK: memref.dealloc %[[IALLOC]]
// CHECK: %[[ALLOC3:.*]] = scf.if
-// CHECK: %[[ALLOC4:.*]] = alloc()
-// CHECK-NEXT: %[[ALLOC5:.*]] = alloc()
+// CHECK: %[[ALLOC4:.*]] = memref.alloc()
+// CHECK-NEXT: %[[ALLOC5:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC4]], %[[ALLOC5]])
-// CHECK-NEXT: dealloc %[[ALLOC4]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC4]]
// CHECK-NEXT: scf.yield %[[ALLOC5]]
-// CHECK: %[[ALLOC6:.*]] = alloc()
+// CHECK: %[[ALLOC6:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC0]], %[[ALLOC6]])
// CHECK-NEXT: scf.yield %[[ALLOC6]]
-// CHECK: %[[ALLOC7:.*]] = alloc()
+// CHECK: %[[ALLOC7:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC3:.*]], %[[ALLOC7]])
-// CHECK-NEXT: dealloc %[[ALLOC3]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC3]]
// CHECK-NEXT: scf.yield %[[ALLOC7]]
-// CHECK: dealloc %[[ALLOC0]]
+// CHECK: memref.dealloc %[[ALLOC0]]
// CHECK-NEXT: return %[[ALLOC2]]
// -----
@@ -1013,17 +1013,17 @@ func @loop_nested_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<2xf32> {
%3 = scf.for %i3 = %lb to %ub step %step
iter_args(%iterBuf3 = %iterBuf2) -> memref<2xf32> {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
%5 = cmpi eq, %i, %ub : index
%6 = scf.if %5 -> (memref<2xf32>) {
- %7 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
scf.yield %7 : memref<2xf32>
} else {
scf.yield %iterBuf3 : memref<2xf32>
@@ -1038,55 +1038,55 @@ func @loop_nested_alloc(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
-// CHECK-NEXT: dealloc %[[ALLOC0]]
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%arg3, %[[ALLOC1]])
// CHECK-NEXT: %[[VAL_7:.*]] = scf.for {{.*}} iter_args
// CHECK-SAME: (%[[IALLOC0:.*]] = %[[ALLOC1]])
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[IALLOC0]], %[[ALLOC2]])
-// CHECK-NEXT: dealloc %[[IALLOC0]]
+// CHECK-NEXT: memref.dealloc %[[IALLOC0]]
// CHECK-NEXT: %[[ALLOC3:.*]] = scf.for {{.*}} iter_args
// CHECK-SAME: (%[[IALLOC1:.*]] = %[[ALLOC2]])
-// CHECK: %[[ALLOC5:.*]] = alloc()
+// CHECK: %[[ALLOC5:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[IALLOC1]], %[[ALLOC5]])
-// CHECK-NEXT: dealloc %[[IALLOC1]]
+// CHECK-NEXT: memref.dealloc %[[IALLOC1]]
// CHECK: %[[ALLOC6:.*]] = scf.for {{.*}} iter_args
// CHECK-SAME: (%[[IALLOC2:.*]] = %[[ALLOC5]])
-// CHECK: %[[ALLOC8:.*]] = alloc()
-// CHECK-NEXT: dealloc %[[ALLOC8]]
+// CHECK: %[[ALLOC8:.*]] = memref.alloc()
+// CHECK-NEXT: memref.dealloc %[[ALLOC8]]
// CHECK: %[[ALLOC9:.*]] = scf.if
-// CHECK: %[[ALLOC11:.*]] = alloc()
-// CHECK-NEXT: %[[ALLOC12:.*]] = alloc()
+// CHECK: %[[ALLOC11:.*]] = memref.alloc()
+// CHECK-NEXT: %[[ALLOC12:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC11]], %[[ALLOC12]])
-// CHECK-NEXT: dealloc %[[ALLOC11]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC11]]
// CHECK-NEXT: scf.yield %[[ALLOC12]]
-// CHECK: %[[ALLOC13:.*]] = alloc()
+// CHECK: %[[ALLOC13:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[IALLOC2]], %[[ALLOC13]])
// CHECK-NEXT: scf.yield %[[ALLOC13]]
-// CHECK: dealloc %[[IALLOC2]]
-// CHECK-NEXT: %[[ALLOC10:.*]] = alloc()
+// CHECK: memref.dealloc %[[IALLOC2]]
+// CHECK-NEXT: %[[ALLOC10:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC9]], %[[ALLOC10]])
-// CHECK-NEXT: dealloc %[[ALLOC9]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC9]]
// CHECK-NEXT: scf.yield %[[ALLOC10]]
-// CHECK: %[[ALLOC7:.*]] = alloc()
+// CHECK: %[[ALLOC7:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC6]], %[[ALLOC7]])
-// CHECK-NEXT: dealloc %[[ALLOC6]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC6]]
// CHECK-NEXT: scf.yield %[[ALLOC7]]
-// CHECK: %[[ALLOC4:.*]] = alloc()
+// CHECK: %[[ALLOC4:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ALLOC3]], %[[ALLOC4]])
-// CHECK-NEXT: dealloc %[[ALLOC3]]
+// CHECK-NEXT: memref.dealloc %[[ALLOC3]]
// CHECK-NEXT: scf.yield %[[ALLOC4]]
// CHECK: test.copy(%[[VAL_7]], %arg4)
-// CHECK-NEXT: dealloc %[[VAL_7]]
+// CHECK-NEXT: memref.dealloc %[[VAL_7]]
// -----
@@ -1113,7 +1113,7 @@ func @loop_dynalloc(
%const1 = constant 1 : i32
%inc = addi %val, %const1 : i32
%size = std.index_cast %inc : i32 to index
- %alloc1 = alloc(%size) : memref<?xf32>
+ %alloc1 = memref.alloc(%size) : memref<?xf32>
br ^loopHeader(%inc, %alloc1 : i32, memref<?xf32>)
^exit(%buff3 : memref<?xf32>):
@@ -1139,7 +1139,7 @@ func @do_loop_alloc(
^loopBody(%val : i32, %buff2: memref<2xf32>):
%const1 = constant 1 : i32
%inc = addi %val, %const1 : i32
- %alloc1 = alloc() : memref<2xf32>
+ %alloc1 = memref.alloc() : memref<2xf32>
br ^loopHeader(%inc, %alloc1 : i32, memref<2xf32>)
^loopHeader(%i : i32, %buff : memref<2xf32>):
@@ -1162,12 +1162,12 @@ func @assumingOp(
%arg3: memref<2xf32>) {
// Confirm the alloc will be dealloc'ed in the block.
%1 = shape.assuming %arg0 -> memref<2xf32> {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
shape.assuming_yield %arg2 : memref<2xf32>
}
// Confirm the alloc will be returned and dealloc'ed after its use.
%3 = shape.assuming %arg0 -> memref<2xf32> {
- %2 = alloc() : memref<2xf32>
+ %2 = memref.alloc() : memref<2xf32>
shape.assuming_yield %2 : memref<2xf32>
}
test.copy(%3, %arg3) : (memref<2xf32>, memref<2xf32>)
@@ -1178,17 +1178,17 @@ func @assumingOp(
// CHECK-SAME: %[[ARG1:.*]]: {{.*}},
// CHECK-SAME: %[[ARG2:.*]]: {{.*}}
// CHECK: %[[UNUSED_RESULT:.*]] = shape.assuming %[[ARG0]]
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
-// CHECK-NEXT: dealloc %[[ALLOC0]]
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
+// CHECK-NEXT: memref.dealloc %[[ALLOC0]]
// CHECK-NEXT: shape.assuming_yield %[[ARG1]]
// CHECK: %[[ASSUMING_RESULT:.*]] = shape.assuming %[[ARG0]]
-// CHECK-NEXT: %[[TMP_ALLOC:.*]] = alloc()
-// CHECK-NEXT: %[[RETURNING_ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[TMP_ALLOC:.*]] = memref.alloc()
+// CHECK-NEXT: %[[RETURNING_ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[TMP_ALLOC]], %[[RETURNING_ALLOC]])
-// CHECK-NEXT: dealloc %[[TMP_ALLOC]]
+// CHECK-NEXT: memref.dealloc %[[TMP_ALLOC]]
// CHECK-NEXT: shape.assuming_yield %[[RETURNING_ALLOC]]
// CHECK: test.copy(%[[ASSUMING_RESULT:.*]], %[[ARG2]])
-// CHECK-NEXT: dealloc %[[ASSUMING_RESULT]]
+// CHECK-NEXT: memref.dealloc %[[ASSUMING_RESULT]]
// -----
diff --git a/mlir/test/Transforms/buffer-hoisting.mlir b/mlir/test/Transforms/buffer-hoisting.mlir
index f45fffde491f..2eed0481c6a9 100644
--- a/mlir/test/Transforms/buffer-hoisting.mlir
+++ b/mlir/test/Transforms/buffer-hoisting.mlir
@@ -18,7 +18,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -26,7 +26,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: cond_br
// -----
@@ -51,7 +51,7 @@ func @condBranchDynamicType(
^bb1:
br ^bb3(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
br ^bb3(%1 : memref<?xf32>)
^bb3(%2: memref<?xf32>):
@@ -62,7 +62,7 @@ func @condBranchDynamicType(
// CHECK-NEXT: cond_br
// CHECK: ^bb2
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
// -----
@@ -93,7 +93,7 @@ func @condBranchDynamicTypeNested(
^bb1:
br ^bb6(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
cond_br %arg0, ^bb3, ^bb4
^bb3:
@@ -112,7 +112,7 @@ func @condBranchDynamicTypeNested(
// CHECK-NEXT: cond_br
// CHECK: ^bb2
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
// -----
@@ -130,7 +130,7 @@ func @condBranchDynamicTypeNested(
func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
cond_br %arg0, ^bb1, ^bb2(%arg1 : memref<2xf32>)
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb2(%0 : memref<2xf32>)
^bb2(%1: memref<2xf32>):
@@ -138,7 +138,7 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: cond_br
// -----
@@ -153,7 +153,7 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElse
func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -163,18 +163,18 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb2(%3: memref<2xf32>, %4: memref<2xf32>):
br ^bb3(%3, %4 : memref<2xf32>, memref<2xf32>)
^bb3(%5: memref<2xf32>, %6: memref<2xf32>):
- %7 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%7: memref<2xf32>)
test.copy(%7, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// CHECK: br ^bb3
// CHECK: br ^bb3
// CHECK-NEXT: ^bb3
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// CHECK: test.copy(%[[ALLOC1]]
// CHECK-NEXT: return
@@ -191,7 +191,7 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNoUsers
func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -205,7 +205,7 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -223,7 +223,7 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNested
func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -237,19 +237,19 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb4(%6: memref<2xf32>):
br ^bb5(%3, %6 : memref<2xf32>, memref<2xf32>)
^bb5(%7: memref<2xf32>, %8: memref<2xf32>):
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%9: memref<2xf32>)
test.copy(%9, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// CHECK: br ^bb5
// CHECK: br ^bb5
// CHECK: br ^bb5
// CHECK-NEXT: ^bb5
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -259,16 +259,16 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @redundantOperations
func @redundantOperations(%arg0: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%0: memref<2xf32>) out(%1: memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -289,11 +289,11 @@ func @moving_alloc_and_inserting_missing_dealloc(
%arg1: memref<2xf32>) {
cond_br %cond, ^bb1, ^bb2
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
br ^exit(%0 : memref<2xf32>)
^bb2:
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%1: memref<2xf32>)
br ^exit(%1 : memref<2xf32>)
^exit(%arg2: memref<2xf32>):
@@ -301,8 +301,8 @@ func @moving_alloc_and_inserting_missing_dealloc(
return
}
-// CHECK-NEXT: %{{.*}} = alloc()
-// CHECK-NEXT: %{{.*}} = alloc()
+// CHECK-NEXT: %{{.*}} = memref.alloc()
+// CHECK-NEXT: %{{.*}} = memref.alloc()
// CHECK-NEXT: cond_br
// -----
@@ -326,16 +326,16 @@ func @moving_invalid_dealloc_op_complex(
^bb1:
br ^exit(%arg0 : memref<2xf32>)
^bb2:
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%1: memref<2xf32>)
- dealloc %1 : memref<2xf32>
+ memref.dealloc %1 : memref<2xf32>
br ^exit(%1 : memref<2xf32>)
^exit(%arg2: memref<2xf32>):
test.copy(%arg2, %arg1) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %{{.*}} = alloc()
+// CHECK-NEXT: %{{.*}} = memref.alloc()
// CHECK-NEXT: cond_br
// -----
@@ -355,10 +355,10 @@ func @nested_regions_and_cond_branch(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -368,10 +368,10 @@ func @nested_regions_and_cond_branch(
test.copy(%1, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: cond_br
// CHECK: test.region_buffer_based
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -385,20 +385,20 @@ func @nested_region_control_flow(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %1 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %{{.*}} = scf.if
// CHECK: else
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%arg0, %arg1)
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%arg0, %arg1)
// -----
@@ -411,18 +411,18 @@ func @nested_region_control_flow_div(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %3 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%arg0, %arg1)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%arg0, %arg1)
// CHECK-NEXT: %{{.*}} = scf.if
// -----
@@ -437,24 +437,24 @@ func @nested_region_control_flow_div_nested(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
%3 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %4 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %4 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %4 : memref<?x?xf32>
}
scf.yield %3 : memref<?x?xf32>
} else {
- %5 = alloc(%arg1, %arg1) : memref<?x?xf32>
+ %5 = memref.alloc(%arg1, %arg1) : memref<?x?xf32>
scf.yield %5 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%arg0, %arg1)
-// CHECK-NEXT: %[[ALLOC2:.*]] = alloc(%arg1, %arg1)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%arg0, %arg1)
+// CHECK-NEXT: %[[ALLOC2:.*]] = memref.alloc(%arg1, %arg1)
// CHECK-NEXT: %{{.*}} = scf.if
// -----
@@ -472,13 +472,13 @@ func @nested_region_control_flow_div_nested_dependencies(
%1 = constant 1 : i32
%2 = addi %arg0, %1 : i32
%3 = index_cast %2 : i32 to index
- %4 = alloc(%arg2, %3) : memref<?x?xf32>
+ %4 = memref.alloc(%arg2, %3) : memref<?x?xf32>
scf.yield %4 : memref<?x?xf32>
} else {
%1 = constant 2 : i32
%2 = addi %arg0, %1 : i32
%3 = index_cast %2 : i32 to index
- %4 = alloc(%arg2, %3) : memref<?x?xf32>
+ %4 = memref.alloc(%arg2, %3) : memref<?x?xf32>
scf.yield %4 : memref<?x?xf32>
}
return %0 : memref<?x?xf32>
@@ -504,7 +504,7 @@ func @nested_region_control_flow_div_nested_dependencies(
// CHECK-LABEL: func @inner_region_control_flow
func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
- %0 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%1 = test.region_if %0 : memref<?x?xf32> -> (memref<?x?xf32>) then {
^bb0(%arg1 : memref<?x?xf32>):
test.region_if_yield %arg1 : memref<?x?xf32>
@@ -518,7 +518,7 @@ func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
return %1 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: {{.*}} test.region_if
// -----
@@ -531,13 +531,13 @@ func @inner_region_control_flow(%arg0 : index) -> memref<?x?xf32> {
func @inner_region_control_flow_div(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
- %0 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %0 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%1 = test.region_if %0 : memref<?x?xf32> -> (memref<?x?xf32>) then {
^bb0(%arg2 : memref<?x?xf32>):
test.region_if_yield %arg2 : memref<?x?xf32>
} else {
^bb0(%arg2 : memref<?x?xf32>):
- %2 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %2 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
test.region_if_yield %2 : memref<?x?xf32>
} join {
^bb0(%arg2 : memref<?x?xf32>):
@@ -546,8 +546,8 @@ func @inner_region_control_flow_div(
return %1 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%arg0, %arg1)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%arg0, %arg1)
// CHECK-NEXT: {{.*}} test.region_if
// -----
@@ -560,7 +560,7 @@ func @condBranchAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloca() : memref<2xf32>
+ %0 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -571,7 +571,7 @@ func @condBranchAlloca(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-NEXT: cond_br
// CHECK: ^bb2
// CHECK: ^bb2
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
// -----
@@ -584,7 +584,7 @@ func @ifElseNestedAlloca(
%arg0: i1,
%arg1: memref<2xf32>,
%arg2: memref<2xf32>) {
- %0 = alloca() : memref<2xf32>
+ %0 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -598,19 +598,19 @@ func @ifElseNestedAlloca(
^bb4(%6: memref<2xf32>):
br ^bb5(%3, %6 : memref<2xf32>, memref<2xf32>)
^bb5(%7: memref<2xf32>, %8: memref<2xf32>):
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%9: memref<2xf32>)
test.copy(%9, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
// CHECK: ^bb5
// CHECK: ^bb5
// CHECK: ^bb5
// CHECK-NEXT: ^bb5
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -627,10 +627,10 @@ func @nestedRegionsAndCondBranchAlloca(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloca() : memref<2xf32>
+ %1 = memref.alloca() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -640,10 +640,10 @@ func @nestedRegionsAndCondBranchAlloca(
test.copy(%1, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: cond_br
// CHECK: test.region_buffer_based
-// CHECK: %[[ALLOCA:.*]] = alloca()
+// CHECK: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
// -----
@@ -658,20 +658,20 @@ func @loop_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %3 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// -----
@@ -685,12 +685,12 @@ func @loop_nested_if_alloc(
%ub: index,
%step: index,
%buf: memref<2xf32>) -> memref<2xf32> {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
%3 = scf.if %2 -> (memref<2xf32>) {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
scf.yield %4 : memref<2xf32>
} else {
scf.yield %0 : memref<2xf32>
@@ -700,9 +700,9 @@ func @loop_nested_if_alloc(
return %1 : memref<2xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// -----
@@ -717,17 +717,17 @@ func @loop_nested_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<2xf32> {
%3 = scf.for %i3 = %lb to %ub step %step
iter_args(%iterBuf3 = %iterBuf2) -> memref<2xf32> {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
%5 = cmpi eq, %i, %ub : index
%6 = scf.if %5 -> (memref<2xf32>) {
- %7 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
scf.yield %7 : memref<2xf32>
} else {
scf.yield %iterBuf3 : memref<2xf32>
@@ -742,12 +742,12 @@ func @loop_nested_alloc(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc()
-// CHECK: %[[ALLOC2:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc()
+// CHECK: %[[ALLOC2:.*]] = memref.alloc()
// -----
@@ -759,7 +759,7 @@ func @loop_nested_alloc_dyn_dependency(
%arg0: index,
%buf: memref<?xf32>,
%res: memref<?xf32>) {
- %0 = alloc(%arg0) : memref<?xf32>
+ %0 = memref.alloc(%arg0) : memref<?xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<?xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
@@ -768,7 +768,7 @@ func @loop_nested_alloc_dyn_dependency(
iter_args(%iterBuf3 = %iterBuf2) -> memref<?xf32> {
%5 = cmpi eq, %i, %ub : index
%6 = scf.if %5 -> (memref<?xf32>) {
- %7 = alloc(%i3) : memref<?xf32>
+ %7 = memref.alloc(%i3) : memref<?xf32>
scf.yield %7 : memref<?xf32>
} else {
scf.yield %iterBuf3 : memref<?xf32>
@@ -784,8 +784,8 @@ func @loop_nested_alloc_dyn_dependency(
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC1:.*]] = memref.alloc({{.*}})
diff --git a/mlir/test/Transforms/buffer-loop-hoisting.mlir b/mlir/test/Transforms/buffer-loop-hoisting.mlir
index 085a4d63b762..c5e727b89449 100644
--- a/mlir/test/Transforms/buffer-loop-hoisting.mlir
+++ b/mlir/test/Transforms/buffer-loop-hoisting.mlir
@@ -17,7 +17,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -26,7 +26,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
}
// CHECK-NEXT: cond_br
-// CHECK: %[[ALLOC:.*]] = alloc()
+// CHECK: %[[ALLOC:.*]] = memref.alloc()
// -----
@@ -50,7 +50,7 @@ func @condBranchDynamicType(
^bb1:
br ^bb3(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
br ^bb3(%1 : memref<?xf32>)
^bb3(%2: memref<?xf32>):
@@ -61,7 +61,7 @@ func @condBranchDynamicType(
// CHECK-NEXT: cond_br
// CHECK: ^bb2
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
// -----
@@ -81,10 +81,10 @@ func @nested_regions_and_cond_branch(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -95,9 +95,9 @@ func @nested_regions_and_cond_branch(
return
}
// CHECK-NEXT: cond_br
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK: test.region_buffer_based
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// CHECK-NEXT: test.buffer_based
// -----
@@ -111,20 +111,20 @@ func @nested_region_control_flow(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %1 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %{{.*}} = scf.if
// CHECK: else
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc(%arg0, %arg1)
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc(%arg0, %arg1)
// -----
@@ -138,20 +138,20 @@ func @loop_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %3 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// -----
@@ -165,12 +165,12 @@ func @loop_nested_if_alloc(
%ub: index,
%step: index,
%buf: memref<2xf32>) -> memref<2xf32> {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
%3 = scf.if %2 -> (memref<2xf32>) {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
scf.yield %4 : memref<2xf32>
} else {
scf.yield %0 : memref<2xf32>
@@ -180,9 +180,9 @@ func @loop_nested_if_alloc(
return %1 : memref<2xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc()
+// CHECK: %[[ALLOC1:.*]] = memref.alloc()
// -----
@@ -198,23 +198,23 @@ func @loop_nested_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<2xf32> {
%3 = scf.for %i3 = %lb to %ub step %step
iter_args(%iterBuf3 = %iterBuf2) -> memref<2xf32> {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
%5 = cmpi eq, %i, %ub : index
%6 = scf.if %5 -> (memref<2xf32>) {
- %7 = alloc() : memref<2xf32>
- %8 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
+ %8 = memref.alloc() : memref<2xf32>
scf.yield %8 : memref<2xf32>
} else {
scf.yield %iterBuf3 : memref<2xf32>
}
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
scf.yield %6 : memref<2xf32>
}
scf.yield %3 : memref<2xf32>
@@ -225,15 +225,15 @@ func @loop_nested_alloc(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc()
-// CHECK-NEXT: %[[ALLOC2:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc()
+// CHECK-NEXT: %[[ALLOC2:.*]] = memref.alloc()
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
// CHECK: {{.*}} = scf.if
-// CHECK: %[[ALLOC3:.*]] = alloc()
-// CHECK: %[[ALLOC4:.*]] = alloc()
+// CHECK: %[[ALLOC3:.*]] = memref.alloc()
+// CHECK: %[[ALLOC4:.*]] = memref.alloc()
// -----
@@ -245,22 +245,22 @@ func @loop_nested_alloc_dyn_dependency(
%arg0: index,
%buf: memref<?xf32>,
%res: memref<?xf32>) {
- %0 = alloc(%arg0) : memref<?xf32>
+ %0 = memref.alloc(%arg0) : memref<?xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<?xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<?xf32> {
%3 = scf.for %i3 = %lb to %ub step %step
iter_args(%iterBuf3 = %iterBuf2) -> memref<?xf32> {
- %4 = alloc(%i3) : memref<?xf32>
+ %4 = memref.alloc(%i3) : memref<?xf32>
%5 = cmpi eq, %i, %ub : index
%6 = scf.if %5 -> (memref<?xf32>) {
- %7 = alloc(%i3) : memref<?xf32>
+ %7 = memref.alloc(%i3) : memref<?xf32>
scf.yield %7 : memref<?xf32>
} else {
scf.yield %iterBuf3 : memref<?xf32>
}
- %8 = alloc(%i3) : memref<?xf32>
+ %8 = memref.alloc(%i3) : memref<?xf32>
scf.yield %6 : memref<?xf32>
}
scf.yield %3 : memref<?xf32>
@@ -271,12 +271,12 @@ func @loop_nested_alloc_dyn_dependency(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
// CHECK-NEXT: {{.*}} = scf.for
-// CHECK: %[[ALLOC1:.*]] = alloc({{.*}})
-// CHECK: %[[ALLOC2:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC1:.*]] = memref.alloc({{.*}})
+// CHECK: %[[ALLOC2:.*]] = memref.alloc({{.*}})
// -----
@@ -287,18 +287,18 @@ func @hoist_one_loop(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
- %2 = alloc() : memref<2xf32>
+ %2 = memref.alloc() : memref<2xf32>
scf.yield %0 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
// -----
@@ -312,7 +312,7 @@ func @no_hoist_one_loop(
%res: memref<2xf32>) {
%0 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
scf.yield %1 : memref<2xf32>
}
test.copy(%0, %res) : (memref<2xf32>, memref<2xf32>)
@@ -320,7 +320,7 @@ func @no_hoist_one_loop(
}
// CHECK: {{.*}} = scf.for
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// -----
@@ -331,12 +331,12 @@ func @hoist_multiple_loop(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<2xf32> {
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %0 : memref<2xf32>
}
scf.yield %0 : memref<2xf32>
@@ -345,8 +345,8 @@ func @hoist_multiple_loop(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
// -----
@@ -362,7 +362,7 @@ func @no_hoist_one_loop_conditional(
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%1 = cmpi eq, %i, %ub : index
%2 = scf.if %1 -> (memref<2xf32>) {
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %3 : memref<2xf32>
} else {
scf.yield %iterBuf : memref<2xf32>
@@ -375,7 +375,7 @@ func @no_hoist_one_loop_conditional(
// CHECK: {{.*}} = scf.for
// CHECK: {{.*}} = scf.if
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// -----
@@ -386,12 +386,12 @@ func @hoist_one_loop_conditional(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = cmpi eq, %lb, %ub : index
%2 = scf.if %1 -> (memref<2xf32>) {
%3 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
scf.yield %0 : memref<2xf32>
}
scf.yield %0 : memref<2xf32>
@@ -405,7 +405,7 @@ func @hoist_one_loop_conditional(
}
// CHECK: {{.*}} = scf.if
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// CHECK: {{.*}} = scf.for
// -----
@@ -417,19 +417,19 @@ func @no_hoist_one_loop_dependency(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
- %2 = alloc(%i) : memref<?xf32>
+ %2 = memref.alloc(%i) : memref<?xf32>
scf.yield %0 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc({{.*}})
// -----
@@ -440,12 +440,12 @@ func @partial_hoist_multiple_loop_dependency(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = scf.for %i2 = %lb to %ub step %step
iter_args(%iterBuf2 = %iterBuf) -> memref<2xf32> {
- %3 = alloc(%i) : memref<?xf32>
+ %3 = memref.alloc(%i) : memref<?xf32>
scf.yield %0 : memref<2xf32>
}
scf.yield %0 : memref<2xf32>
@@ -454,7 +454,7 @@ func @partial_hoist_multiple_loop_dependency(
return
}
-// CHECK: %[[ALLOC0:.*]] = alloc({{.*}})
+// CHECK: %[[ALLOC0:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
-// CHECK-NEXT: %[[ALLOC1:.*]] = alloc({{.*}})
+// CHECK-NEXT: %[[ALLOC1:.*]] = memref.alloc({{.*}})
// CHECK-NEXT: {{.*}} = scf.for
diff --git a/mlir/test/Transforms/buffer-results-to-out-params.mlir b/mlir/test/Transforms/buffer-results-to-out-params.mlir
index cde929739475..cac3e7461225 100644
--- a/mlir/test/Transforms/buffer-results-to-out-params.mlir
+++ b/mlir/test/Transforms/buffer-results-to-out-params.mlir
@@ -60,7 +60,7 @@ func private @mixed_result_attrs() -> (memref<1xf32>, memref<2xf32> {test.some_a
func private @callee() -> memref<1xf32>
// CHECK-LABEL: func @call_basic() {
-// CHECK: %[[OUTPARAM:.*]] = alloc() : memref<1xf32>
+// CHECK: %[[OUTPARAM:.*]] = memref.alloc() : memref<1xf32>
// CHECK: call @callee(%[[OUTPARAM]]) : (memref<1xf32>) -> ()
// CHECK: "test.sink"(%[[OUTPARAM]]) : (memref<1xf32>) -> ()
// CHECK: return
@@ -77,8 +77,8 @@ func @call_basic() {
func private @callee() -> (memref<1xf32>, memref<2xf32>)
// CHECK-LABEL: func @call_multiple_result() {
-// CHECK: %[[RESULT0:.*]] = alloc() : memref<1xf32>
-// CHECK: %[[RESULT1:.*]] = alloc() : memref<2xf32>
+// CHECK: %[[RESULT0:.*]] = memref.alloc() : memref<1xf32>
+// CHECK: %[[RESULT1:.*]] = memref.alloc() : memref<2xf32>
// CHECK: call @callee(%[[RESULT0]], %[[RESULT1]]) : (memref<1xf32>, memref<2xf32>) -> ()
// CHECK: "test.sink"(%[[RESULT0]], %[[RESULT1]]) : (memref<1xf32>, memref<2xf32>) -> ()
// CHECK: }
@@ -93,7 +93,7 @@ func @call_multiple_result() {
func private @callee() -> (i1, memref<1xf32>, i32)
// CHECK-LABEL: func @call_non_memref_result() {
-// CHECK: %[[RESULT0:.*]] = alloc() : memref<1xf32>
+// CHECK: %[[RESULT0:.*]] = memref.alloc() : memref<1xf32>
// CHECK: %[[NON_MEMREF_RESULTS:.*]]:2 = call @callee(%[[RESULT0]]) : (memref<1xf32>) -> (i1, i32)
// CHECK: "test.sink"(%[[NON_MEMREF_RESULTS]]#0, %[[RESULT0]], %[[NON_MEMREF_RESULTS]]#1) : (i1, memref<1xf32>, i32) -> ()
// CHECK: }
diff --git a/mlir/test/Transforms/canonicalize-block-merge.mlir b/mlir/test/Transforms/canonicalize-block-merge.mlir
index 98d274a63fc2..2606fb776e7b 100644
--- a/mlir/test/Transforms/canonicalize-block-merge.mlir
+++ b/mlir/test/Transforms/canonicalize-block-merge.mlir
@@ -213,13 +213,13 @@ func @mismatch_operand_types(%arg0 : i1, %arg1 : memref<i32>, %arg2 : memref<i1>
cond_br %arg0, ^bb2, ^bb3
^bb2:
- // CHECK: store %{{.*}}, %{{.*}} : memref<i32>
- store %c0_i32, %arg1[] : memref<i32>
+ // CHECK: memref.store %{{.*}}, %{{.*}} : memref<i32>
+ memref.store %c0_i32, %arg1[] : memref<i32>
br ^bb1
^bb3:
- // CHECK: store %{{.*}}, %{{.*}} : memref<i1>
- store %true, %arg2[] : memref<i1>
+ // CHECK: memref.store %{{.*}}, %{{.*}} : memref<i1>
+ memref.store %true, %arg2[] : memref<i1>
br ^bb1
}
diff --git a/mlir/test/Transforms/canonicalize.mlir b/mlir/test/Transforms/canonicalize.mlir
index 3bc3eeee8354..5009d0f4207d 100644
--- a/mlir/test/Transforms/canonicalize.mlir
+++ b/mlir/test/Transforms/canonicalize.mlir
@@ -29,7 +29,7 @@ func @dim(%arg0: tensor<8x4xf32>) -> index {
// CHECK: %c4 = constant 4 : index
%c1 = constant 1 : index
- %0 = dim %arg0, %c1 : tensor<8x4xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<8x4xf32>
// CHECK-NEXT: return %c4
return %0 : index
@@ -53,7 +53,7 @@ func @test_commutative(%arg0: i32) -> (i32, i32) {
// CHECK-LABEL: func @trivial_dce
func @trivial_dce(%arg0: tensor<8x4xf32>) {
%c1 = constant 1 : index
- %0 = dim %arg0, %c1 : tensor<8x4xf32>
+ %0 = memref.dim %arg0, %c1 : tensor<8x4xf32>
// CHECK-NEXT: return
return
}
@@ -61,9 +61,9 @@ func @trivial_dce(%arg0: tensor<8x4xf32>) {
// CHECK-LABEL: func @load_dce
func @load_dce(%arg0: index) {
%c4 = constant 4 : index
- %a = alloc(%c4) : memref<?xf32>
- %2 = load %a[%arg0] : memref<?xf32>
- dealloc %a: memref<?xf32>
+ %a = memref.alloc(%c4) : memref<?xf32>
+ %2 = memref.load %a[%arg0] : memref<?xf32>
+ memref.dealloc %a: memref<?xf32>
// CHECK-NEXT: return
return
}
@@ -313,22 +313,22 @@ func @xor_self_tensor(%arg0: tensor<4x5xi32>) -> tensor<4x5xi32> {
// CHECK-LABEL: func @memref_cast_folding
func @memref_cast_folding(%arg0: memref<4 x f32>, %arg1: f32) -> (f32, f32) {
- %0 = memref_cast %arg0 : memref<4xf32> to memref<?xf32>
+ %0 = memref.cast %arg0 : memref<4xf32> to memref<?xf32>
// CHECK-NEXT: %c0 = constant 0 : index
%c0 = constant 0 : index
- %dim = dim %0, %c0 : memref<? x f32>
+ %dim = memref.dim %0, %c0 : memref<? x f32>
// CHECK-NEXT: affine.load %arg0[3]
%1 = affine.load %0[%dim - 1] : memref<?xf32>
- // CHECK-NEXT: store %arg1, %arg0[%c0] : memref<4xf32>
- store %arg1, %0[%c0] : memref<?xf32>
+ // CHECK-NEXT: memref.store %arg1, %arg0[%c0] : memref<4xf32>
+ memref.store %arg1, %0[%c0] : memref<?xf32>
- // CHECK-NEXT: %{{.*}} = load %arg0[%c0] : memref<4xf32>
- %2 = load %0[%c0] : memref<?xf32>
+ // CHECK-NEXT: %{{.*}} = memref.load %arg0[%c0] : memref<4xf32>
+ %2 = memref.load %0[%c0] : memref<?xf32>
- // CHECK-NEXT: dealloc %arg0 : memref<4xf32>
- dealloc %0: memref<?xf32>
+ // CHECK-NEXT: memref.dealloc %arg0 : memref<4xf32>
+ memref.dealloc %0: memref<?xf32>
// CHECK-NEXT: return %{{.*}}
return %1, %2 : f32, f32
@@ -337,10 +337,10 @@ func @memref_cast_folding(%arg0: memref<4 x f32>, %arg1: f32) -> (f32, f32) {
// CHECK-LABEL: @fold_memref_cast_in_memref_cast
// CHECK-SAME: (%[[ARG0:.*]]: memref<42x42xf64>)
func @fold_memref_cast_in_memref_cast(%0: memref<42x42xf64>) {
- // CHECK: %[[folded:.*]] = memref_cast %[[ARG0]] : memref<42x42xf64> to memref<?x?xf64>
- %4 = memref_cast %0 : memref<42x42xf64> to memref<?x42xf64>
- // CHECK-NOT: memref_cast
- %5 = memref_cast %4 : memref<?x42xf64> to memref<?x?xf64>
+ // CHECK: %[[folded:.*]] = memref.cast %[[ARG0]] : memref<42x42xf64> to memref<?x?xf64>
+ %4 = memref.cast %0 : memref<42x42xf64> to memref<?x42xf64>
+ // CHECK-NOT: memref.cast
+ %5 = memref.cast %4 : memref<?x42xf64> to memref<?x?xf64>
// CHECK: "test.user"(%[[folded]])
"test.user"(%5) : (memref<?x?xf64>) -> ()
return
@@ -349,9 +349,9 @@ func @fold_memref_cast_in_memref_cast(%0: memref<42x42xf64>) {
// CHECK-LABEL: @fold_memref_cast_chain
// CHECK-SAME: (%[[ARG0:.*]]: memref<42x42xf64>)
func @fold_memref_cast_chain(%0: memref<42x42xf64>) {
- // CHECK-NOT: memref_cast
- %4 = memref_cast %0 : memref<42x42xf64> to memref<?x42xf64>
- %5 = memref_cast %4 : memref<?x42xf64> to memref<42x42xf64>
+ // CHECK-NOT: memref.cast
+ %4 = memref.cast %0 : memref<42x42xf64> to memref<?x42xf64>
+ %5 = memref.cast %4 : memref<?x42xf64> to memref<42x42xf64>
// CHECK: "test.user"(%[[ARG0]])
"test.user"(%5) : (memref<42x42xf64>) -> ()
return
@@ -359,11 +359,11 @@ func @fold_memref_cast_chain(%0: memref<42x42xf64>) {
// CHECK-LABEL: func @alloc_const_fold
func @alloc_const_fold() -> memref<?xf32> {
- // CHECK-NEXT: %0 = alloc() : memref<4xf32>
+ // CHECK-NEXT: %0 = memref.alloc() : memref<4xf32>
%c4 = constant 4 : index
- %a = alloc(%c4) : memref<?xf32>
+ %a = memref.alloc(%c4) : memref<?xf32>
- // CHECK-NEXT: %1 = memref_cast %0 : memref<4xf32> to memref<?xf32>
+ // CHECK-NEXT: %1 = memref.cast %0 : memref<4xf32> to memref<?xf32>
// CHECK-NEXT: return %1 : memref<?xf32>
return %a : memref<?xf32>
}
@@ -372,30 +372,30 @@ func @alloc_const_fold() -> memref<?xf32> {
func @dead_alloc_fold() {
// CHECK-NEXT: return
%c4 = constant 4 : index
- %a = alloc(%c4) : memref<?xf32>
+ %a = memref.alloc(%c4) : memref<?xf32>
return
}
// CHECK-LABEL: func @dead_dealloc_fold
func @dead_dealloc_fold() {
// CHECK-NEXT: return
- %a = alloc() : memref<4xf32>
- dealloc %a: memref<4xf32>
+ %a = memref.alloc() : memref<4xf32>
+ memref.dealloc %a: memref<4xf32>
return
}
// CHECK-LABEL: func @dead_dealloc_fold_multi_use
func @dead_dealloc_fold_multi_use(%cond : i1) {
// CHECK-NEXT: return
- %a = alloc() : memref<4xf32>
+ %a = memref.alloc() : memref<4xf32>
cond_br %cond, ^bb1, ^bb2
^bb1:
- dealloc %a: memref<4xf32>
+ memref.dealloc %a: memref<4xf32>
return
^bb2:
- dealloc %a: memref<4xf32>
+ memref.dealloc %a: memref<4xf32>
return
}
@@ -423,29 +423,29 @@ func @dyn_shape_fold(%L : index, %M : index) -> (memref<? x ? x i32>, memref<? x
%N = constant 1024 : index
%K = constant 512 : index
- // CHECK-NEXT: alloc(%arg0) : memref<?x1024xf32>
- %a = alloc(%L, %N) : memref<? x ? x f32>
+ // CHECK-NEXT: memref.alloc(%arg0) : memref<?x1024xf32>
+ %a = memref.alloc(%L, %N) : memref<? x ? x f32>
- // CHECK-NEXT: alloc(%arg1) : memref<4x1024x8x512x?xf32>
- %b = alloc(%N, %K, %M) : memref<4 x ? x 8 x ? x ? x f32>
+ // CHECK-NEXT: memref.alloc(%arg1) : memref<4x1024x8x512x?xf32>
+ %b = memref.alloc(%N, %K, %M) : memref<4 x ? x 8 x ? x ? x f32>
- // CHECK-NEXT: alloc() : memref<512x1024xi32>
- %c = alloc(%K, %N) : memref<? x ? x i32>
+ // CHECK-NEXT: memref.alloc() : memref<512x1024xi32>
+ %c = memref.alloc(%K, %N) : memref<? x ? x i32>
- // CHECK: alloc() : memref<9x9xf32>
- %d = alloc(%nine, %nine) : memref<? x ? x f32>
+ // CHECK: memref.alloc() : memref<9x9xf32>
+ %d = memref.alloc(%nine, %nine) : memref<? x ? x f32>
- // CHECK: alloca(%arg1) : memref<4x1024x8x512x?xf32>
- %e = alloca(%N, %K, %M) : memref<4 x ? x 8 x ? x ? x f32>
+ // CHECK: memref.alloca(%arg1) : memref<4x1024x8x512x?xf32>
+ %e = memref.alloca(%N, %K, %M) : memref<4 x ? x 8 x ? x ? x f32>
// CHECK: affine.for
affine.for %i = 0 to %L {
// CHECK-NEXT: affine.for
affine.for %j = 0 to 10 {
- // CHECK-NEXT: load %0[%arg2, %arg3] : memref<?x1024xf32>
- // CHECK-NEXT: store %{{.*}}, %1[%c0, %c0, %arg2, %arg3, %c0] : memref<4x1024x8x512x?xf32>
- %v = load %a[%i, %j] : memref<?x?xf32>
- store %v, %b[%zero, %zero, %i, %j, %zero] : memref<4x?x8x?x?xf32>
+ // CHECK-NEXT: memref.load %0[%arg2, %arg3] : memref<?x1024xf32>
+ // CHECK-NEXT: memref.store %{{.*}}, %1[%c0, %c0, %arg2, %arg3, %c0] : memref<4x1024x8x512x?xf32>
+ %v = memref.load %a[%i, %j] : memref<?x?xf32>
+ memref.store %v, %b[%zero, %zero, %i, %j, %zero] : memref<4x?x8x?x?xf32>
}
}
@@ -468,24 +468,24 @@ func @dim_op_fold(%arg0: index, %arg1: index, %arg2: index, %BUF: memref<?xi8>,
%c0 = constant 0 : index
%c1 = constant 1 : index
%c2 = constant 2 : index
- %0 = alloc(%arg0, %arg1) : memref<?x?xf32>
- %1 = alloc(%arg1, %arg2) : memref<?x8x?xf32>
- %2 = dim %1, %c2 : memref<?x8x?xf32>
+ %0 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
+ %1 = memref.alloc(%arg1, %arg2) : memref<?x8x?xf32>
+ %2 = memref.dim %1, %c2 : memref<?x8x?xf32>
affine.for %arg3 = 0 to %2 {
- %3 = alloc(%arg0) : memref<?xi8>
- %ub = dim %3, %c0 : memref<?xi8>
+ %3 = memref.alloc(%arg0) : memref<?xi8>
+ %ub = memref.dim %3, %c0 : memref<?xi8>
affine.for %arg4 = 0 to %ub {
- %s = dim %0, %c0 : memref<?x?xf32>
- %v = std.view %3[%c0][%arg4, %s] : memref<?xi8> to memref<?x?xf32>
- %sv = subview %0[%c0, %c0][%s,%arg4][%c1,%c1] : memref<?x?xf32> to memref<?x?xf32, #map1>
- %l = dim %v, %c1 : memref<?x?xf32>
- %u = dim %sv, %c0 : memref<?x?xf32, #map1>
+ %s = memref.dim %0, %c0 : memref<?x?xf32>
+ %v = memref.view %3[%c0][%arg4, %s] : memref<?xi8> to memref<?x?xf32>
+ %sv = memref.subview %0[%c0, %c0][%s,%arg4][%c1,%c1] : memref<?x?xf32> to memref<?x?xf32, #map1>
+ %l = memref.dim %v, %c1 : memref<?x?xf32>
+ %u = memref.dim %sv, %c0 : memref<?x?xf32, #map1>
affine.for %arg5 = %l to %u {
"foo"() : () -> ()
}
- %sv2 = subview %0[0, 0][17, %arg4][1, 1] : memref<?x?xf32> to memref<17x?xf32, #map3>
- %l2 = dim %v, %c1 : memref<?x?xf32>
- %u2 = dim %sv2, %c1 : memref<17x?xf32, #map3>
+ %sv2 = memref.subview %0[0, 0][17, %arg4][1, 1] : memref<?x?xf32> to memref<17x?xf32, #map3>
+ %l2 = memref.dim %v, %c1 : memref<?x?xf32>
+ %u2 = memref.dim %sv2, %c1 : memref<17x?xf32, #map3>
scf.for %arg5 = %l2 to %u2 step %c1 {
"foo"() : () -> ()
}
@@ -502,13 +502,13 @@ func @dim_op_fold(%arg0: index, %arg1: index, %arg2: index, %BUF: memref<?xi8>,
// CHECK-NEXT: }
// CHECK-NEXT: }
- %A = view %BUF[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
- %B = view %BUF[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
- %C = view %BUF[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
+ %A = memref.view %BUF[%c0][%M, %K] : memref<?xi8> to memref<?x?xf32>
+ %B = memref.view %BUF[%c0][%K, %N] : memref<?xi8> to memref<?x?xf32>
+ %C = memref.view %BUF[%c0][%M, %N] : memref<?xi8> to memref<?x?xf32>
- %M_ = dim %A, %c0 : memref<?x?xf32>
- %K_ = dim %A, %c1 : memref<?x?xf32>
- %N_ = dim %C, %c1 : memref<?x?xf32>
+ %M_ = memref.dim %A, %c0 : memref<?x?xf32>
+ %K_ = memref.dim %A, %c1 : memref<?x?xf32>
+ %N_ = memref.dim %C, %c1 : memref<?x?xf32>
scf.for %i = %c0 to %M_ step %c1 {
scf.for %j = %c0 to %N_ step %c1 {
scf.for %k = %c0 to %K_ step %c1 {
@@ -533,9 +533,9 @@ func @hoist_constant(%arg0: memref<8xi32>) {
// CHECK-NEXT: %c42_i32 = constant 42 : i32
// CHECK-NEXT: affine.for %arg1 = 0 to 8 {
affine.for %arg1 = 0 to 8 {
- // CHECK-NEXT: store %c42_i32, %arg0[%arg1]
+ // CHECK-NEXT: memref.store %c42_i32, %arg0[%arg1]
%c42_i32 = constant 42 : i32
- store %c42_i32, %arg0[%arg1] : memref<8xi32>
+ memref.store %c42_i32, %arg0[%arg1] : memref<8xi32>
}
return
}
@@ -547,8 +547,8 @@ func @const_fold_propagate() -> memref<?x?xf32> {
%VT_i_s = affine.apply affine_map<(d0) -> (d0 floordiv 8)> (%VT_i)
%VT_k_l = affine.apply affine_map<(d0) -> (d0 floordiv 16)> (%VT_i)
- // CHECK: = alloc() : memref<64x32xf32>
- %Av = alloc(%VT_i_s, %VT_k_l) : memref<?x?xf32>
+ // CHECK: = memref.alloc() : memref<64x32xf32>
+ %Av = memref.alloc(%VT_i_s, %VT_k_l) : memref<?x?xf32>
return %Av : memref<?x?xf32>
}
@@ -663,11 +663,11 @@ func @lowered_affine_ceildiv() -> (index, index) {
// CHECK-LABEL: cast_values
func @cast_values(%arg0: memref<?xi32>) -> memref<2xi32> {
// NOP cast
- %1 = memref_cast %arg0 : memref<?xi32> to memref<?xi32>
- // CHECK-NEXT: %[[RET:.*]] = memref_cast %arg0 : memref<?xi32> to memref<2xi32>
- %3 = memref_cast %1 : memref<?xi32> to memref<2xi32>
+ %1 = memref.cast %arg0 : memref<?xi32> to memref<?xi32>
+ // CHECK-NEXT: %[[RET:.*]] = memref.cast %arg0 : memref<?xi32> to memref<2xi32>
+ %3 = memref.cast %1 : memref<?xi32> to memref<2xi32>
// NOP cast
- %5 = memref_cast %3 : memref<2xi32> to memref<2xi32>
+ %5 = memref.cast %3 : memref<2xi32> to memref<2xi32>
// CHECK-NEXT: return %[[RET]] : memref<2xi32>
return %5 : memref<2xi32>
}
@@ -677,33 +677,33 @@ func @cast_values(%arg0: memref<?xi32>) -> memref<2xi32> {
// CHECK-LABEL: func @view
func @view(%arg0 : index) -> (f32, f32, f32, f32) {
// CHECK: %[[C15:.*]] = constant 15 : index
- // CHECK: %[[ALLOC_MEM:.*]] = alloc() : memref<2048xi8>
- %0 = alloc() : memref<2048xi8>
+ // CHECK: %[[ALLOC_MEM:.*]] = memref.alloc() : memref<2048xi8>
+ %0 = memref.alloc() : memref<2048xi8>
%c0 = constant 0 : index
%c7 = constant 7 : index
%c11 = constant 11 : index
%c15 = constant 15 : index
// Test: fold constant sizes.
- // CHECK: std.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<7x11xf32>
- %1 = view %0[%c15][%c7, %c11] : memref<2048xi8> to memref<?x?xf32>
- %r0 = load %1[%c0, %c0] : memref<?x?xf32>
+ // CHECK: memref.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<7x11xf32>
+ %1 = memref.view %0[%c15][%c7, %c11] : memref<2048xi8> to memref<?x?xf32>
+ %r0 = memref.load %1[%c0, %c0] : memref<?x?xf32>
// Test: fold one constant size.
- // CHECK: std.view %[[ALLOC_MEM]][%[[C15]]][%arg0, %arg0] : memref<2048xi8> to memref<?x?x7xf32>
- %2 = view %0[%c15][%arg0, %arg0, %c7] : memref<2048xi8> to memref<?x?x?xf32>
- %r1 = load %2[%c0, %c0, %c0] : memref<?x?x?xf32>
+ // CHECK: memref.view %[[ALLOC_MEM]][%[[C15]]][%arg0, %arg0] : memref<2048xi8> to memref<?x?x7xf32>
+ %2 = memref.view %0[%c15][%arg0, %arg0, %c7] : memref<2048xi8> to memref<?x?x?xf32>
+ %r1 = memref.load %2[%c0, %c0, %c0] : memref<?x?x?xf32>
// Test: preserve an existing static size.
- // CHECK: std.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<7x4xf32>
- %3 = view %0[%c15][%c7] : memref<2048xi8> to memref<?x4xf32>
- %r2 = load %3[%c0, %c0] : memref<?x4xf32>
-
- // Test: folding static alloc and memref_cast into a view.
- // CHECK: std.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<15x7xf32>
- %4 = memref_cast %0 : memref<2048xi8> to memref<?xi8>
- %5 = view %4[%c15][%c15, %c7] : memref<?xi8> to memref<?x?xf32>
- %r3 = load %5[%c0, %c0] : memref<?x?xf32>
+ // CHECK: memref.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<7x4xf32>
+ %3 = memref.view %0[%c15][%c7] : memref<2048xi8> to memref<?x4xf32>
+ %r2 = memref.load %3[%c0, %c0] : memref<?x4xf32>
+
+ // Test: folding static alloc and memref.cast into a view.
+ // CHECK memref.view %[[ALLOC_MEM]][%[[C15]]][] : memref<2048xi8> to memref<15x7xf32>
+ %4 = memref.cast %0 : memref<2048xi8> to memref<?xi8>
+ %5 = memref.view %4[%c15][%c15, %c7] : memref<?xi8> to memref<?x?xf32>
+ %r3 = memref.load %5[%c0, %c0] : memref<?x?xf32>
return %r0, %r1, %r2, %r3 : f32, f32, f32, f32
}
@@ -739,142 +739,142 @@ func @subview(%arg0 : index, %arg1 : index) -> (index, index) {
// CHECK-NOT: constant 15 : index
%c15 = constant 15 : index
- // CHECK: %[[ALLOC0:.*]] = alloc()
- %0 = alloc() : memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]>
+ // CHECK: %[[ALLOC0:.*]] = memref.alloc()
+ %0 = memref.alloc() : memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]>
// Test: subview with constant base memref and constant operands is folded.
// Note that the subview uses the base memrefs layout map because it used
// zero offset and unit stride arguments.
- // CHECK: subview %[[ALLOC0]][0, 0, 0] [7, 11, 2] [1, 1, 1] :
+ // CHECK: memref.subview %[[ALLOC0]][0, 0, 0] [7, 11, 2] [1, 1, 1] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<7x11x2xf32, #[[$BASE_MAP0]]>
- %1 = subview %0[%c0, %c0, %c0] [%c7, %c11, %c2] [%c1, %c1, %c1]
+ %1 = memref.subview %0[%c0, %c0, %c0] [%c7, %c11, %c2] [%c1, %c1, %c1]
: memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]> to
memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- %v0 = load %1[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ %v0 = memref.load %1[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
// Test: subview with one dynamic operand can also be folded.
- // CHECK: subview %[[ALLOC0]][0, %[[ARG0]], 0] [7, 11, 15] [1, 1, 1] :
+ // CHECK: memref.subview %[[ALLOC0]][0, %[[ARG0]], 0] [7, 11, 15] [1, 1, 1] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<7x11x15xf32, #[[$SUBVIEW_MAP0]]>
- %2 = subview %0[%c0, %arg0, %c0] [%c7, %c11, %c15] [%c1, %c1, %c1]
+ %2 = memref.subview %0[%c0, %arg0, %c0] [%c7, %c11, %c15] [%c1, %c1, %c1]
: memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]> to
memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- store %v0, %2[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ memref.store %v0, %2[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- // CHECK: %[[ALLOC1:.*]] = alloc(%[[ARG0]])
- %3 = alloc(%arg0) : memref<?x16x4xf32, offset : 0, strides : [64, 4, 1]>
+ // CHECK: %[[ALLOC1:.*]] = memref.alloc(%[[ARG0]])
+ %3 = memref.alloc(%arg0) : memref<?x16x4xf32, offset : 0, strides : [64, 4, 1]>
// Test: subview with constant operands but dynamic base memref is folded as long as the strides and offset of the base memref are static.
- // CHECK: subview %[[ALLOC1]][0, 0, 0] [7, 11, 15] [1, 1, 1] :
+ // CHECK: memref.subview %[[ALLOC1]][0, 0, 0] [7, 11, 15] [1, 1, 1] :
// CHECK-SAME: memref<?x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<7x11x15xf32, #[[$BASE_MAP0]]>
- %4 = subview %3[%c0, %c0, %c0] [%c7, %c11, %c15] [%c1, %c1, %c1]
+ %4 = memref.subview %3[%c0, %c0, %c0] [%c7, %c11, %c15] [%c1, %c1, %c1]
: memref<?x16x4xf32, offset : 0, strides : [64, 4, 1]> to
memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- store %v0, %4[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ memref.store %v0, %4[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
// Test: subview offset operands are folded correctly w.r.t. base strides.
- // CHECK: subview %[[ALLOC0]][1, 2, 7] [7, 11, 2] [1, 1, 1] :
+ // CHECK: memref.subview %[[ALLOC0]][1, 2, 7] [7, 11, 2] [1, 1, 1] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]> to
// CHECK-SAME: memref<7x11x2xf32, #[[$SUBVIEW_MAP1]]>
- %5 = subview %0[%c1, %c2, %c7] [%c7, %c11, %c2] [%c1, %c1, %c1]
+ %5 = memref.subview %0[%c1, %c2, %c7] [%c7, %c11, %c2] [%c1, %c1, %c1]
: memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]> to
memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- store %v0, %5[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ memref.store %v0, %5[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
// Test: subview stride operands are folded correctly w.r.t. base strides.
- // CHECK: subview %[[ALLOC0]][0, 0, 0] [7, 11, 2] [2, 7, 11] :
+ // CHECK: memref.subview %[[ALLOC0]][0, 0, 0] [7, 11, 2] [2, 7, 11] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]>
// CHECK-SAME: to memref<7x11x2xf32, #[[$SUBVIEW_MAP2]]>
- %6 = subview %0[%c0, %c0, %c0] [%c7, %c11, %c2] [%c2, %c7, %c11]
+ %6 = memref.subview %0[%c0, %c0, %c0] [%c7, %c11, %c2] [%c2, %c7, %c11]
: memref<8x16x4xf32, offset : 0, strides : [64, 4, 1]> to
memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- store %v0, %6[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ memref.store %v0, %6[%c0, %c0, %c0] : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
// Test: subview shape are folded, but offsets and strides are not even if base memref is static
- // CHECK: subview %[[ALLOC0]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [7, 11, 2] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
+ // CHECK: memref.subview %[[ALLOC0]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [7, 11, 2] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]> to
// CHECK-SAME: memref<7x11x2xf32, #[[$SUBVIEW_MAP3]]>
- %10 = subview %0[%arg0, %arg0, %arg0] [%c7, %c11, %c2] [%arg1, %arg1, %arg1] :
+ %10 = memref.subview %0[%arg0, %arg0, %arg0] [%c7, %c11, %c2] [%arg1, %arg1, %arg1] :
memref<8x16x4xf32, offset:0, strides:[64, 4, 1]> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %10[%arg1, %arg1, %arg1] :
+ memref.store %v0, %10[%arg1, %arg1, %arg1] :
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
// Test: subview strides are folded, but offsets and shape are not even if base memref is static
- // CHECK: subview %[[ALLOC0]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [2, 7, 11] :
+ // CHECK: memref.subview %[[ALLOC0]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [2, 7, 11] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]> to
// CHECK-SAME: memref<?x?x?xf32, #[[$SUBVIEW_MAP4]]
- %11 = subview %0[%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] [%c2, %c7, %c11] :
+ %11 = memref.subview %0[%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] [%c2, %c7, %c11] :
memref<8x16x4xf32, offset:0, strides:[64, 4, 1]> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %11[%arg0, %arg0, %arg0] :
+ memref.store %v0, %11[%arg0, %arg0, %arg0] :
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
// Test: subview offsets are folded, but strides and shape are not even if base memref is static
- // CHECK: subview %[[ALLOC0]][1, 2, 7] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [%[[ARG0]], %[[ARG0]], %[[ARG0]]] :
+ // CHECK: memref.subview %[[ALLOC0]][1, 2, 7] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [%[[ARG0]], %[[ARG0]], %[[ARG0]]] :
// CHECK-SAME: memref<8x16x4xf32, #[[$BASE_MAP0]]> to
// CHECK-SAME: memref<?x?x?xf32, #[[$SUBVIEW_MAP5]]
- %13 = subview %0[%c1, %c2, %c7] [%arg1, %arg1, %arg1] [%arg0, %arg0, %arg0] :
+ %13 = memref.subview %0[%c1, %c2, %c7] [%arg1, %arg1, %arg1] [%arg0, %arg0, %arg0] :
memref<8x16x4xf32, offset:0, strides:[64, 4, 1]> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %13[%arg1, %arg1, %arg1] :
+ memref.store %v0, %13[%arg1, %arg1, %arg1] :
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- // CHECK: %[[ALLOC2:.*]] = alloc(%[[ARG0]], %[[ARG0]], %[[ARG1]])
- %14 = alloc(%arg0, %arg0, %arg1) : memref<?x?x?xf32>
+ // CHECK: %[[ALLOC2:.*]] = memref.alloc(%[[ARG0]], %[[ARG0]], %[[ARG1]])
+ %14 = memref.alloc(%arg0, %arg0, %arg1) : memref<?x?x?xf32>
// Test: subview shape are folded, even if base memref is not static
- // CHECK: subview %[[ALLOC2]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [7, 11, 2] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
+ // CHECK: memref.subview %[[ALLOC2]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [7, 11, 2] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
// CHECK-SAME: memref<?x?x?xf32> to
// CHECK-SAME: memref<7x11x2xf32, #[[$SUBVIEW_MAP3]]>
- %15 = subview %14[%arg0, %arg0, %arg0] [%c7, %c11, %c2] [%arg1, %arg1, %arg1] :
+ %15 = memref.subview %14[%arg0, %arg0, %arg0] [%c7, %c11, %c2] [%arg1, %arg1, %arg1] :
memref<?x?x?xf32> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %15[%arg1, %arg1, %arg1] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
+ memref.store %v0, %15[%arg1, %arg1, %arg1] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
// TEST: subview strides are folded, in the type only the most minor stride is folded.
- // CHECK: subview %[[ALLOC2]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [2, 2, 2] :
+ // CHECK: memref.subview %[[ALLOC2]][%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] [2, 2, 2] :
// CHECK-SAME: memref<?x?x?xf32> to
// CHECK-SAME: memref<?x?x?xf32, #[[$SUBVIEW_MAP6]]
- %16 = subview %14[%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] [%c2, %c2, %c2] :
+ %16 = memref.subview %14[%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] [%c2, %c2, %c2] :
memref<?x?x?xf32> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %16[%arg0, %arg0, %arg0] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
+ memref.store %v0, %16[%arg0, %arg0, %arg0] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
// TEST: subview offsets are folded but the type offset remains dynamic, when the base memref is not static
- // CHECK: subview %[[ALLOC2]][1, 1, 1] [%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
+ // CHECK: memref.subview %[[ALLOC2]][1, 1, 1] [%[[ARG0]], %[[ARG0]], %[[ARG0]]] [%[[ARG1]], %[[ARG1]], %[[ARG1]]] :
// CHECK-SAME: memref<?x?x?xf32> to
// CHECK-SAME: memref<?x?x?xf32, #[[$SUBVIEW_MAP3]]
- %17 = subview %14[%c1, %c1, %c1] [%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] :
+ %17 = memref.subview %14[%c1, %c1, %c1] [%arg0, %arg0, %arg0] [%arg1, %arg1, %arg1] :
memref<?x?x?xf32> to
memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- store %v0, %17[%arg0, %arg0, %arg0] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
+ memref.store %v0, %17[%arg0, %arg0, %arg0] : memref<?x?x?xf32, offset: ?, strides: [?, ?, ?]>
- // CHECK: %[[ALLOC3:.*]] = alloc() : memref<12x4xf32>
- %18 = alloc() : memref<12x4xf32>
+ // CHECK: %[[ALLOC3:.*]] = memref.alloc() : memref<12x4xf32>
+ %18 = memref.alloc() : memref<12x4xf32>
%c4 = constant 4 : index
// TEST: subview strides are maintained when sizes are folded
- // CHECK: subview %[[ALLOC3]][%arg1, %arg1] [2, 4] [1, 1] :
+ // CHECK: memref.subview %[[ALLOC3]][%arg1, %arg1] [2, 4] [1, 1] :
// CHECK-SAME: memref<12x4xf32> to
// CHECK-SAME: memref<2x4xf32, #[[$SUBVIEW_MAP7]]>
- %19 = subview %18[%arg1, %arg1] [%c2, %c4] [1, 1] :
+ %19 = memref.subview %18[%arg1, %arg1] [%c2, %c4] [1, 1] :
memref<12x4xf32> to
memref<?x?xf32, offset: ?, strides:[4, 1]>
- store %v0, %19[%arg1, %arg1] : memref<?x?xf32, offset: ?, strides:[4, 1]>
+ memref.store %v0, %19[%arg1, %arg1] : memref<?x?xf32, offset: ?, strides:[4, 1]>
// TEST: subview strides and sizes are maintained when offsets are folded
- // CHECK: subview %[[ALLOC3]][2, 4] [12, 4] [1, 1] :
+ // CHECK: memref.subview %[[ALLOC3]][2, 4] [12, 4] [1, 1] :
// CHECK-SAME: memref<12x4xf32> to
// CHECK-SAME: memref<12x4xf32, #[[$SUBVIEW_MAP8]]>
- %20 = subview %18[%c2, %c4] [12, 4] [1, 1] :
+ %20 = memref.subview %18[%c2, %c4] [12, 4] [1, 1] :
memref<12x4xf32> to
memref<12x4xf32, offset: ?, strides:[4, 1]>
- store %v0, %20[%arg1, %arg1] : memref<12x4xf32, offset: ?, strides:[4, 1]>
+ memref.store %v0, %20[%arg1, %arg1] : memref<12x4xf32, offset: ?, strides:[4, 1]>
// Test: dim on subview is rewritten to size operand.
- %7 = dim %4, %c0 : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
- %8 = dim %4, %c1 : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ %7 = memref.dim %4, %c0 : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
+ %8 = memref.dim %4, %c1 : memref<?x?x?xf32, offset : ?, strides : [?, ?, ?]>
// CHECK: return %[[C7]], %[[C11]]
return %7, %8 : index, index
@@ -1006,9 +1006,9 @@ func @tensor_ceildivi_signed_by_one(%arg0: tensor<4x5xi32>) -> tensor<4x5xi32> {
// CHECK-LABEL: func @memref_cast_folding_subview
func @memref_cast_folding_subview(%arg0: memref<4x5xf32>, %i: index) -> (memref<?x?xf32, offset:? , strides: [?, ?]>) {
- %0 = memref_cast %arg0 : memref<4x5xf32> to memref<?x?xf32>
- // CHECK-NEXT: subview %{{.*}}: memref<4x5xf32>
- %1 = subview %0[%i, %i][%i, %i][%i, %i]: memref<?x?xf32> to memref<?x?xf32, offset:? , strides: [?, ?]>
+ %0 = memref.cast %arg0 : memref<4x5xf32> to memref<?x?xf32>
+ // CHECK-NEXT: memref.subview %{{.*}}: memref<4x5xf32>
+ %1 = memref.subview %0[%i, %i][%i, %i][%i, %i]: memref<?x?xf32> to memref<?x?xf32, offset:? , strides: [?, ?]>
// CHECK-NEXT: return %{{.*}}
return %1: memref<?x?xf32, offset:? , strides: [?, ?]>
}
@@ -1022,11 +1022,10 @@ func @memref_cast_folding_subview(%arg0: memref<4x5xf32>, %i: index) -> (memref<
func @memref_cast_folding_subview_static(%V: memref<16x16xf32>, %a: index, %b: index)
-> memref<3x4xf32, offset:?, strides:[?, 1]>
{
- %0 = memref_cast %V : memref<16x16xf32> to memref<?x?xf32>
- %1 = subview %0[0, 0][3, 4][1, 1] : memref<?x?xf32> to memref<3x4xf32, offset:?, strides:[?, 1]>
+ %0 = memref.cast %V : memref<16x16xf32> to memref<?x?xf32>
+ %1 = memref.subview %0[0, 0][3, 4][1, 1] : memref<?x?xf32> to memref<3x4xf32, offset:?, strides:[?, 1]>
- // CHECK: subview{{.*}}: memref<16x16xf32> to memref<3x4xf32, #[[$map0]]>
- // CHECK: memref_cast{{.*}}: memref<3x4xf32, #[[$map0]]> to memref<3x4xf32, #[[$map1]]>
+ // CHECK: memref.subview{{.*}}: memref<16x16xf32> to memref<3x4xf32, #[[$map0]]>
return %1: memref<3x4xf32, offset:?, strides:[?, 1]>
}
diff --git a/mlir/test/Transforms/constant-fold.mlir b/mlir/test/Transforms/constant-fold.mlir
index 74dea976ab1f..dab0c4b645e5 100644
--- a/mlir/test/Transforms/constant-fold.mlir
+++ b/mlir/test/Transforms/constant-fold.mlir
@@ -13,8 +13,8 @@ func @affine_for(%p : memref<f32>) {
%2 = addf %0, %1 : f32
- // CHECK-NEXT: store [[C]], [[ARG]][]
- store %2, %p[] : memref<f32>
+ // CHECK-NEXT: memref.store [[C]], [[ARG]][]
+ memref.store %2, %p[] : memref<f32>
}
}
return
@@ -548,7 +548,7 @@ func @dim(%x : tensor<8x4xf32>) -> index {
// CHECK:[[C4:%.+]] = constant 4 : index
%c1 = constant 1 : index
- %0 = dim %x, %c1 : tensor<8x4xf32>
+ %0 = memref.dim %x, %c1 : tensor<8x4xf32>
// CHECK-NEXT: return [[C4]]
return %0 : index
@@ -793,7 +793,7 @@ func @splat_fold() -> (vector<4xf32>, tensor<4xf32>) {
// CHECK-LABEL: func @subview_scalar_fold
func @subview_scalar_fold(%arg0: memref<f32>) -> memref<f32> {
- // CHECK-NOT: subview
- %c = subview %arg0[] [] [] : memref<f32> to memref<f32>
+ // CHECK-NOT: memref.subview
+ %c = memref.subview %arg0[] [] [] : memref<f32> to memref<f32>
return %c : memref<f32>
}
diff --git a/mlir/test/Transforms/copy-removal.mlir b/mlir/test/Transforms/copy-removal.mlir
index 1432037f2110..a91c5c2b9528 100644
--- a/mlir/test/Transforms/copy-removal.mlir
+++ b/mlir/test/Transforms/copy-removal.mlir
@@ -6,65 +6,65 @@
// CHECK-LABEL: func @nested_region_control_flow_div_nested
func @nested_region_control_flow_div_nested(%arg0: index, %arg1: index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
// CHECK: %{{.*}} = scf.if
%2 = scf.if %0 -> (memref<?x?xf32>) {
// CHECK: %[[PERCENT3:.*]] = scf.if
%3 = scf.if %0 -> (memref<?x?xf32>) {
%c0_0 = constant 0 : index
- %7 = dim %1, %c0_0 : memref<?x?xf32>
+ %7 = memref.dim %1, %c0_0 : memref<?x?xf32>
%c1_1 = constant 1 : index
- %8 = dim %1, %c1_1 : memref<?x?xf32>
- %9 = alloc(%7, %8) : memref<?x?xf32>
+ %8 = memref.dim %1, %c1_1 : memref<?x?xf32>
+ %9 = memref.alloc(%7, %8) : memref<?x?xf32>
// CHECK: linalg.copy({{.*}}, %[[PERCENT9:.*]])
linalg.copy(%1, %9) : memref<?x?xf32>, memref<?x?xf32>
// CHECK: scf.yield %[[PERCENT9]]
scf.yield %9 : memref<?x?xf32>
} else {
- // CHECK: %[[PERCENT7:.*]] = alloc
- %7 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ // CHECK: %[[PERCENT7:.*]] = memref.alloc
+ %7 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
%c0_0 = constant 0 : index
- %8 = dim %7, %c0_0 : memref<?x?xf32>
+ %8 = memref.dim %7, %c0_0 : memref<?x?xf32>
%c1_1 = constant 1 : index
- %9 = dim %7, %c1_1 : memref<?x?xf32>
- // CHECK-NOT: %{{.*}} = alloc
+ %9 = memref.dim %7, %c1_1 : memref<?x?xf32>
+ // CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NOT: linalg.copy(%[[PERCENT7]], %{{.*}})
- // CHECK-NOT: dealloc %[[PERCENT7]]
- %10 = alloc(%8, %9) : memref<?x?xf32>
+ // CHECK-NOT: memref.dealloc %[[PERCENT7]]
+ %10 = memref.alloc(%8, %9) : memref<?x?xf32>
linalg.copy(%7, %10) : memref<?x?xf32>, memref<?x?xf32>
- dealloc %7 : memref<?x?xf32>
+ memref.dealloc %7 : memref<?x?xf32>
// CHECK: scf.yield %[[PERCENT7]]
scf.yield %10 : memref<?x?xf32>
}
%c0 = constant 0 : index
- %4 = dim %3, %c0 : memref<?x?xf32>
+ %4 = memref.dim %3, %c0 : memref<?x?xf32>
%c1 = constant 1 : index
- %5 = dim %3, %c1 : memref<?x?xf32>
- // CHECK-NOT: %{{.*}} = alloc
+ %5 = memref.dim %3, %c1 : memref<?x?xf32>
+ // CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NOT: linalg.copy(%[[PERCENT3]], %{{.*}})
- // CHECK-NOT: dealloc %[[PERCENT3]]
- %6 = alloc(%4, %5) : memref<?x?xf32>
+ // CHECK-NOT: memref.dealloc %[[PERCENT3]]
+ %6 = memref.alloc(%4, %5) : memref<?x?xf32>
linalg.copy(%3, %6) : memref<?x?xf32>, memref<?x?xf32>
- dealloc %3 : memref<?x?xf32>
+ memref.dealloc %3 : memref<?x?xf32>
// CHECK: scf.yield %[[PERCENT3]]
scf.yield %6 : memref<?x?xf32>
} else {
- // CHECK: %[[PERCENT3:.*]] = alloc
- %3 = alloc(%arg1, %arg1) : memref<?x?xf32>
+ // CHECK: %[[PERCENT3:.*]] = memref.alloc
+ %3 = memref.alloc(%arg1, %arg1) : memref<?x?xf32>
%c0 = constant 0 : index
- %4 = dim %3, %c0 : memref<?x?xf32>
+ %4 = memref.dim %3, %c0 : memref<?x?xf32>
%c1 = constant 1 : index
- %5 = dim %3, %c1 : memref<?x?xf32>
- // CHECK-NOT: %{{.*}} = alloc
+ %5 = memref.dim %3, %c1 : memref<?x?xf32>
+ // CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NOT: linalg.copy(%[[PERCENT3]], %{{.*}})
- // CHECK-NOT: dealloc %[[PERCENT3]]
- %6 = alloc(%4, %5) : memref<?x?xf32>
+ // CHECK-NOT: memref.dealloc %[[PERCENT3]]
+ %6 = memref.alloc(%4, %5) : memref<?x?xf32>
linalg.copy(%3, %6) : memref<?x?xf32>, memref<?x?xf32>
- dealloc %3 : memref<?x?xf32>
+ memref.dealloc %3 : memref<?x?xf32>
// CHECK: scf.yield %[[PERCENT3]]
scf.yield %6 : memref<?x?xf32>
}
- dealloc %1 : memref<?x?xf32>
+ memref.dealloc %1 : memref<?x?xf32>
return %2 : memref<?x?xf32>
}
@@ -72,16 +72,16 @@ func @nested_region_control_flow_div_nested(%arg0: index, %arg1: index) -> memre
// CHECK-LABEL: func @simple_test
func @simple_test() -> memref<5xf32> {
- %temp = alloc() : memref<5xf32>
- %ret = alloc() : memref<5xf32>
+ %temp = memref.alloc() : memref<5xf32>
+ %ret = memref.alloc() : memref<5xf32>
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
- dealloc %ret : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
return %temp : memref<5xf32>
}
// CHECK-SAME: () -> memref<5xf32>
-// CHECK-NEXT: %[[ret:.*]] = alloc()
+// CHECK-NEXT: %[[ret:.*]] = memref.alloc()
// CHECK-NOT: linalg.copy(%[[ret]], %{{.*}})
-// CHECK-NOT: dealloc %[[ret]]
+// CHECK-NOT: memref.dealloc %[[ret]]
// CHECK: return %[[ret]]
// -----
@@ -92,20 +92,20 @@ func @simple_test() -> memref<5xf32> {
// CHECK-LABEL: func @test_with_ret_usage_before_copy
func @test_with_ret_usage_before_copy() -> memref<5xf32> {
- %ret = alloc() : memref<5xf32>
- %temp = alloc() : memref<5xf32>
+ %ret = memref.alloc() : memref<5xf32>
+ %temp = memref.alloc() : memref<5xf32>
%c0 = constant 0 : index
- %dimension = dim %ret, %c0 : memref<5xf32>
+ %dimension = memref.dim %ret, %c0 : memref<5xf32>
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
- dealloc %ret : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
return %temp : memref<5xf32>
}
-// CHECK-NEXT: %[[ret:.*]] = alloc()
-// CHECK-NOT: %{{.*}} = alloc
+// CHECK-NEXT: %[[ret:.*]] = memref.alloc()
+// CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NEXT: %{{.*}} = constant
-// CHECK-NEXT: %[[DIM:.*]] = dim %[[ret]]
+// CHECK-NEXT: %[[DIM:.*]] = memref.dim %[[ret]]
// CHECK-NOT: linalg.copy(%[[ret]], %{{.*}})
-// CHECK-NOT: dealloc %[[ret]]
+// CHECK-NOT: memref.dealloc %[[ret]]
// CHECK: return %[[ret]]
// -----
@@ -115,13 +115,13 @@ func @test_with_ret_usage_before_copy() -> memref<5xf32> {
// CHECK-LABEL: func @test_with_ret_usage_after_copy
func @test_with_ret_usage_after_copy() -> memref<5xf32> {
- %ret = alloc() : memref<5xf32>
- %temp = alloc() : memref<5xf32>
+ %ret = memref.alloc() : memref<5xf32>
+ %temp = memref.alloc() : memref<5xf32>
// CHECK: linalg.copy
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
%c0 = constant 0 : index
- %dimension = dim %ret, %c0 : memref<5xf32>
- dealloc %ret : memref<5xf32>
+ %dimension = memref.dim %ret, %c0 : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
return %temp : memref<5xf32>
}
@@ -132,13 +132,13 @@ func @test_with_ret_usage_after_copy() -> memref<5xf32> {
// CHECK-LABEL: func @test_with_temp_usage_before_copy
func @test_with_temp_usage_before_copy() -> memref<5xf32> {
- %ret = alloc() : memref<5xf32>
- %temp = alloc() : memref<5xf32>
+ %ret = memref.alloc() : memref<5xf32>
+ %temp = memref.alloc() : memref<5xf32>
%c0 = constant 0 : index
- %dimension = dim %temp, %c0 : memref<5xf32>
+ %dimension = memref.dim %temp, %c0 : memref<5xf32>
// CHECK: linalg.copy
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
- dealloc %ret : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
return %temp : memref<5xf32>
}
@@ -149,11 +149,11 @@ func @test_with_temp_usage_before_copy() -> memref<5xf32> {
// removed.
// However the following pattern is not handled by copy removal.
-// %from = alloc()
-// %to = alloc()
+// %from = memref.alloc()
+// %to = memref.alloc()
// copy(%from, %to)
// read_from(%from) + write_to(%something_else)
-// dealloc(%from)
+// memref.dealloc(%from)
// return %to
// In particular, linalg.generic is a memoryEffectOp between copy and dealloc.
// Since no alias analysis is performed and no distinction is made between reads
@@ -163,9 +163,9 @@ func @test_with_temp_usage_before_copy() -> memref<5xf32> {
// CHECK-LABEL: func @test_with_temp_usage_after_copy
func @test_with_temp_usage_after_copy() -> memref<5xf32> {
- %ret = alloc() : memref<5xf32>
- %res = alloc() : memref<5xf32>
- %temp = alloc() : memref<5xf32>
+ %ret = memref.alloc() : memref<5xf32>
+ %res = memref.alloc() : memref<5xf32>
+ %temp = memref.alloc() : memref<5xf32>
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
linalg.generic {
indexing_maps = [#map0, #map0],
@@ -176,22 +176,22 @@ func @test_with_temp_usage_after_copy() -> memref<5xf32> {
%tmp1 = math.exp %gen1_arg0 : f32
linalg.yield %tmp1 : f32
}
- dealloc %ret : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
return %temp : memref<5xf32>
}
-// CHECK-NEXT: %[[ret:.*]] = alloc()
-// CHECK-NEXT: %[[res:.*]] = alloc()
-// CHECK-NEXT: %[[temp:.*]] = alloc()
+// CHECK-NEXT: %[[ret:.*]] = memref.alloc()
+// CHECK-NEXT: %[[res:.*]] = memref.alloc()
+// CHECK-NEXT: %[[temp:.*]] = memref.alloc()
// CHECK-NEXT: linalg.copy(%[[ret]], %[[temp]])
// CHECK-NEXT: linalg.generic
-// CHECK: dealloc %[[ret]]
+// CHECK: memref.dealloc %[[ret]]
// CHECK: return %[[temp]]
// -----
// CHECK-LABEL: func @make_allocation
func @make_allocation() -> memref<5xf32> {
- %mem = alloc() : memref<5xf32>
+ %mem = memref.alloc() : memref<5xf32>
return %mem : memref<5xf32>
}
@@ -199,12 +199,12 @@ func @make_allocation() -> memref<5xf32> {
func @test_with_function_call() -> memref<5xf32> {
// CHECK-NEXT: %[[ret:.*]] = call @make_allocation() : () -> memref<5xf32>
%ret = call @make_allocation() : () -> (memref<5xf32>)
- // CHECK-NOT: %{{.*}} = alloc
+ // CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NOT: linalg.copy(%[[ret]], %{{.*}})
- // CHECK-NOT: dealloc %[[ret]]
- %temp = alloc() : memref<5xf32>
+ // CHECK-NOT: memref.dealloc %[[ret]]
+ %temp = memref.alloc() : memref<5xf32>
linalg.copy(%ret, %temp) : memref<5xf32>, memref<5xf32>
- dealloc %ret : memref<5xf32>
+ memref.dealloc %ret : memref<5xf32>
// CHECK: return %[[ret]]
return %temp : memref<5xf32>
}
@@ -213,20 +213,20 @@ func @test_with_function_call() -> memref<5xf32> {
// CHECK-LABEL: func @multiple_deallocs_in_
diff erent_blocks
func @multiple_deallocs_in_
diff erent_blocks(%cond : i1) -> memref<5xf32> {
- // CHECK-NEXT: %[[PERCENT0:.*]] = alloc()
- %0 = alloc() : memref<5xf32>
+ // CHECK-NEXT: %[[PERCENT0:.*]] = memref.alloc()
+ %0 = memref.alloc() : memref<5xf32>
cond_br %cond, ^bb1, ^bb2
^bb1:
- dealloc %0 : memref<5xf32>
+ memref.dealloc %0 : memref<5xf32>
// CHECK: br ^[[BB3:.*]](%[[PERCENT0]]
br ^bb3(%0 : memref<5xf32>)
^bb2:
- // CHECK-NOT: %{{.*}} = alloc
+ // CHECK-NOT: %{{.*}} = memref.alloc
// CHECK-NOT: linalg.copy(%[[PERCENT0]], %{{.*}})
- // CHECK-NOT: dealloc %[[PERCENT0]]
- %temp = alloc() : memref<5xf32>
+ // CHECK-NOT: memref.dealloc %[[PERCENT0]]
+ %temp = memref.alloc() : memref<5xf32>
linalg.copy(%0, %temp) : memref<5xf32>, memref<5xf32>
- dealloc %0 : memref<5xf32>
+ memref.dealloc %0 : memref<5xf32>
// CHECK: br ^[[BB3]](%[[PERCENT0]]
br ^bb3(%temp : memref<5xf32>)
^bb3(%res : memref<5xf32>):
@@ -240,12 +240,12 @@ func @multiple_deallocs_in_
diff erent_blocks(%cond : i1) -> memref<5xf32> {
// CHECK-LABEL: func @test_ReuseCopyTargetAsSource
func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>, %result: memref<2xf32>){
// CHECK-SAME: (%[[ARG0:.*]]: memref<2xf32>, %[[RES:.*]]: memref<2xf32>)
- // CHECK-NOT: %{{.*}} = alloc
- %temp = alloc() : memref<2xf32>
+ // CHECK-NOT: %{{.*}} = memref.alloc
+ %temp = memref.alloc() : memref<2xf32>
// CHECK-NEXT: linalg.generic
// CHECK-SAME: ins(%[[ARG0]]{{.*}}outs(%[[RES]]
// CHECK-NOT: linalg.copy(%{{.*}}, %[[RES]])
- // CHECK-NOT: dealloc %{{.*}}
+ // CHECK-NOT: memref.dealloc %{{.*}}
linalg.generic {
indexing_maps = [#map0, #map0],
iterator_types = ["parallel"]}
@@ -256,7 +256,7 @@ func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>, %result: memref<2xf32>)
linalg.yield %tmp2 : f32
}
linalg.copy(%temp, %result) : memref<2xf32>, memref<2xf32>
- dealloc %temp : memref<2xf32>
+ memref.dealloc %temp : memref<2xf32>
// CHECK: return
return
}
@@ -270,8 +270,8 @@ func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>, %result: memref<2xf32>)
// CHECK-LABEL: func @test_ReuseCopyTargetAsSource
func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>){
- %to = alloc() : memref<2xf32>
- %temp = alloc() : memref<2xf32>
+ %to = memref.alloc() : memref<2xf32>
+ %temp = memref.alloc() : memref<2xf32>
linalg.generic {
indexing_maps = [#map0, #map0],
iterator_types = ["parallel"]}
@@ -292,7 +292,7 @@ func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>){
}
// CHECK: linalg.copy
linalg.copy(%temp, %to) : memref<2xf32>, memref<2xf32>
- dealloc %temp : memref<2xf32>
+ memref.dealloc %temp : memref<2xf32>
return
}
@@ -302,34 +302,34 @@ func @test_ReuseCopyTargetAsSource(%arg0: memref<2xf32>){
// CHECK-LABEL: func @loop_alloc
func @loop_alloc(%arg0: index, %arg1: index, %arg2: index, %arg3: memref<2xf32>, %arg4: memref<2xf32>) {
- // CHECK: %{{.*}} = alloc()
- %0 = alloc() : memref<2xf32>
- dealloc %0 : memref<2xf32>
- // CHECK: %{{.*}} = alloc()
- %1 = alloc() : memref<2xf32>
+ // CHECK: %{{.*}} = memref.alloc()
+ %0 = memref.alloc() : memref<2xf32>
+ memref.dealloc %0 : memref<2xf32>
+ // CHECK: %{{.*}} = memref.alloc()
+ %1 = memref.alloc() : memref<2xf32>
// CHECK: linalg.copy
linalg.copy(%arg3, %1) : memref<2xf32>, memref<2xf32>
%2 = scf.for %arg5 = %arg0 to %arg1 step %arg2 iter_args(%arg6 = %1) -> (memref<2xf32>) {
%3 = cmpi eq, %arg5, %arg1 : index
- // CHECK: dealloc
- dealloc %arg6 : memref<2xf32>
- // CHECK: %[[PERCENT4:.*]] = alloc()
- %4 = alloc() : memref<2xf32>
- // CHECK-NOT: alloc
+ // CHECK: memref.dealloc
+ memref.dealloc %arg6 : memref<2xf32>
+ // CHECK: %[[PERCENT4:.*]] = memref.alloc()
+ %4 = memref.alloc() : memref<2xf32>
+ // CHECK-NOT: memref.alloc
// CHECK-NOT: linalg.copy
- // CHECK-NOT: dealloc
- %5 = alloc() : memref<2xf32>
+ // CHECK-NOT: memref.dealloc
+ %5 = memref.alloc() : memref<2xf32>
linalg.copy(%4, %5) : memref<2xf32>, memref<2xf32>
- dealloc %4 : memref<2xf32>
- // CHECK: %[[PERCENT6:.*]] = alloc()
- %6 = alloc() : memref<2xf32>
+ memref.dealloc %4 : memref<2xf32>
+ // CHECK: %[[PERCENT6:.*]] = memref.alloc()
+ %6 = memref.alloc() : memref<2xf32>
// CHECK: linalg.copy(%[[PERCENT4]], %[[PERCENT6]])
linalg.copy(%5, %6) : memref<2xf32>, memref<2xf32>
scf.yield %6 : memref<2xf32>
}
// CHECK: linalg.copy
linalg.copy(%2, %arg4) : memref<2xf32>, memref<2xf32>
- dealloc %2 : memref<2xf32>
+ memref.dealloc %2 : memref<2xf32>
return
}
@@ -341,8 +341,8 @@ func @loop_alloc(%arg0: index, %arg1: index, %arg2: index, %arg3: memref<2xf32>,
// CHECK-LABEL: func @check_with_affine_dialect
func @check_with_affine_dialect(%arg0: memref<4xf32>, %arg1: memref<4xf32>, %arg2: memref<4xf32>) {
// CHECK-SAME: (%[[ARG0:.*]]: memref<4xf32>, %[[ARG1:.*]]: memref<4xf32>, %[[RES:.*]]: memref<4xf32>)
- // CHECK-NOT: alloc
- %0 = alloc() : memref<4xf32>
+ // CHECK-NOT: memref.alloc
+ %0 = memref.alloc() : memref<4xf32>
affine.for %arg3 = 0 to 4 {
%5 = affine.load %arg0[%arg3] : memref<4xf32>
%6 = affine.load %arg1[%arg3] : memref<4xf32>
@@ -355,7 +355,7 @@ func @check_with_affine_dialect(%arg0: memref<4xf32>, %arg1: memref<4xf32>, %arg
// CHECK-NOT: linalg.copy
// CHECK-NOT: dealloc
linalg.copy(%0, %arg2) : memref<4xf32>, memref<4xf32>
- dealloc %0 : memref<4xf32>
+ memref.dealloc %0 : memref<4xf32>
//CHECK: return
return
}
diff --git a/mlir/test/Transforms/cse.mlir b/mlir/test/Transforms/cse.mlir
index 1c9b7650ac97..455d560d9471 100644
--- a/mlir/test/Transforms/cse.mlir
+++ b/mlir/test/Transforms/cse.mlir
@@ -96,11 +96,11 @@ func @
diff erent_attributes(index, index) -> (i1, i1, i1) {
/// Check that operations with side effects are not eliminated.
// CHECK-LABEL: @side_effect
func @side_effect() -> (memref<2x1xf32>, memref<2x1xf32>) {
- // CHECK: %0 = alloc() : memref<2x1xf32>
- %0 = alloc() : memref<2x1xf32>
+ // CHECK: %0 = memref.alloc() : memref<2x1xf32>
+ %0 = memref.alloc() : memref<2x1xf32>
- // CHECK-NEXT: %1 = alloc() : memref<2x1xf32>
- %1 = alloc() : memref<2x1xf32>
+ // CHECK-NEXT: %1 = memref.alloc() : memref<2x1xf32>
+ %1 = memref.alloc() : memref<2x1xf32>
// CHECK-NEXT: return %0, %1 : memref<2x1xf32>, memref<2x1xf32>
return %0, %1 : memref<2x1xf32>, memref<2x1xf32>
diff --git a/mlir/test/Transforms/finalizing-bufferize.mlir b/mlir/test/Transforms/finalizing-bufferize.mlir
index 5c09664776ea..0188217968d0 100644
--- a/mlir/test/Transforms/finalizing-bufferize.mlir
+++ b/mlir/test/Transforms/finalizing-bufferize.mlir
@@ -4,24 +4,24 @@
// CHECK-SAME: %[[ARG:.*]]: memref<f32>) -> memref<f32> {
// CHECK: return %[[ARG]] : memref<f32>
func @eliminate_materializations(%arg0: memref<f32>) -> memref<f32> {
- %0 = tensor_load %arg0 : memref<f32>
- %1 = tensor_to_memref %0 : memref<f32>
+ %0 = memref.tensor_load %arg0 : memref<f32>
+ %1 = memref.buffer_cast %0 : memref<f32>
return %1 : memref<f32>
}
// -----
-func @unable_to_convert_lone_tensor_to_memref() -> memref<f32> {
+func @unable_to_convert_lone_buffer_cast() -> memref<f32> {
// expected-error @+1 {{failed to legalize operation 'test.source'}}
%0 = "test.source"() : () -> tensor<f32>
- %1 = tensor_to_memref %0 : memref<f32>
+ %1 = memref.buffer_cast %0 : memref<f32>
return %1 : memref<f32>
}
// -----
func @unable_to_convert_lone_tensor_load(%arg0: memref<f32>) {
- %0 = tensor_load %arg0 : memref<f32>
+ %0 = memref.tensor_load %arg0 : memref<f32>
// expected-error @+1 {{failed to legalize operation 'test.sink'}}
"test.sink"(%0) : (tensor<f32>) -> ()
return
diff --git a/mlir/test/Transforms/loop-fusion-dependence-check.mlir b/mlir/test/Transforms/loop-fusion-dependence-check.mlir
index 3d8f8017826f..95bb68c268cf 100644
--- a/mlir/test/Transforms/loop-fusion-dependence-check.mlir
+++ b/mlir/test/Transforms/loop-fusion-dependence-check.mlir
@@ -4,9 +4,9 @@
// CHECK-LABEL: func @cannot_fuse_would_create_cycle() {
func @cannot_fuse_would_create_cycle() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -37,9 +37,9 @@ func @cannot_fuse_would_create_cycle() {
// CHECK-LABEL: func @can_fuse_rar_dependence() {
func @can_fuse_rar_dependence() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -69,10 +69,10 @@ func @can_fuse_rar_dependence() {
// CHECK-LABEL: func @can_fuse_
diff erent_memrefs() {
func @can_fuse_
diff erent_memrefs() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
- %d = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
+ %d = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -102,7 +102,7 @@ func @can_fuse_
diff erent_memrefs() {
// CHECK-LABEL: func @should_not_fuse_across_intermediate_store() {
func @should_not_fuse_across_intermediate_store() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -127,7 +127,7 @@ func @should_not_fuse_across_intermediate_store() {
// CHECK-LABEL: func @should_not_fuse_across_intermediate_load() {
func @should_not_fuse_across_intermediate_load() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -152,8 +152,8 @@ func @should_not_fuse_across_intermediate_load() {
// CHECK-LABEL: func @should_not_fuse_across_ssa_value_def() {
func @should_not_fuse_across_ssa_value_def() {
- %0 = alloc() : memref<10xf32>
- %1 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
+ %1 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -182,7 +182,7 @@ func @should_not_fuse_across_ssa_value_def() {
// CHECK-LABEL: func @should_not_fuse_store_before_load() {
func @should_not_fuse_store_before_load() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -208,7 +208,7 @@ func @should_not_fuse_store_before_load() {
// CHECK-LABEL: func @should_not_fuse_across_load_at_depth1() {
func @should_not_fuse_across_load_at_depth1() {
- %0 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -232,7 +232,7 @@ func @should_not_fuse_across_load_at_depth1() {
// CHECK-LABEL: func @should_not_fuse_across_load_in_loop_at_depth1() {
func @should_not_fuse_across_load_in_loop_at_depth1() {
- %0 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -258,7 +258,7 @@ func @should_not_fuse_across_load_in_loop_at_depth1() {
// CHECK-LABEL: func @should_not_fuse_across_store_at_depth1() {
func @should_not_fuse_across_store_at_depth1() {
- %0 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -282,7 +282,7 @@ func @should_not_fuse_across_store_at_depth1() {
// CHECK-LABEL: func @should_not_fuse_across_store_in_loop_at_depth1() {
func @should_not_fuse_across_store_in_loop_at_depth1() {
- %0 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
@@ -308,8 +308,8 @@ func @should_not_fuse_across_store_in_loop_at_depth1() {
// CHECK-LABEL: func @should_not_fuse_across_ssa_value_def_at_depth1() {
func @should_not_fuse_across_ssa_value_def_at_depth1() {
- %0 = alloc() : memref<10x10xf32>
- %1 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
+ %1 = memref.alloc() : memref<10x10xf32>
%c0 = constant 0 : index
%cf7 = constant 7.0 : f32
diff --git a/mlir/test/Transforms/loop-fusion-slice-computation.mlir b/mlir/test/Transforms/loop-fusion-slice-computation.mlir
index cc54a09ce0c7..7a29ebab932d 100644
--- a/mlir/test/Transforms/loop-fusion-slice-computation.mlir
+++ b/mlir/test/Transforms/loop-fusion-slice-computation.mlir
@@ -4,7 +4,7 @@
// CHECK-LABEL: func @slice_depth1_loop_nest() {
func @slice_depth1_loop_nest() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
// expected-remark at -1 {{slice ( src loop: 1, dst loop: 0, depth: 1 : insert point: (1, 1) loop bounds: [(d0) -> (d0), (d0) -> (d0 + 1)] )}}
@@ -24,7 +24,7 @@ func @slice_depth1_loop_nest() {
// same location.
// CHECK-LABEL: func @slice_depth1_loop_nest_with_offsets() {
func @slice_depth1_loop_nest_with_offsets() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
// expected-remark at -1 {{slice ( src loop: 1, dst loop: 0, depth: 1 : insert point: (1, 2) loop bounds: [(d0) -> (d0 + 3), (d0) -> (d0 + 4)] )}}
@@ -45,7 +45,7 @@ func @slice_depth1_loop_nest_with_offsets() {
// Slices at loop depth 2 should slice loop bounds of both loops.
// CHECK-LABEL: func @slice_depth2_loop_nest() {
func @slice_depth2_loop_nest() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
// expected-remark at -1 {{slice ( src loop: 1, dst loop: 0, depth: 1 : insert point: (1, 1) loop bounds: [(d0) -> (d0), (d0) -> (d0 + 1)] [(d0) -> (0), (d0) -> (8)] )}}
@@ -71,7 +71,7 @@ func @slice_depth2_loop_nest() {
// depths 1 and 2 because the dependent store in loop nest %i0 is at depth 2.
// CHECK-LABEL: func @slice_depth2_loop_nest_two_loads() {
func @slice_depth2_loop_nest_two_loads() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%c0 = constant 0 : index
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
@@ -99,7 +99,7 @@ func @slice_depth2_loop_nest_two_loads() {
// loop nest %i2 is at depth 2.
// CHECK-LABEL: func @slice_depth2_loop_nest_two_stores() {
func @slice_depth2_loop_nest_two_stores() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%c0 = constant 0 : index
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
@@ -124,7 +124,7 @@ func @slice_depth2_loop_nest_two_stores() {
// Test loop nest which has a smaller outer trip count than its inner scf.
// CHECK-LABEL: func @slice_loop_nest_with_smaller_outer_trip_count() {
func @slice_loop_nest_with_smaller_outer_trip_count() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%c0 = constant 0 : index
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
diff --git a/mlir/test/Transforms/loop-fusion-transformation.mlir b/mlir/test/Transforms/loop-fusion-transformation.mlir
index 1b335fa1669a..c978218b532b 100644
--- a/mlir/test/Transforms/loop-fusion-transformation.mlir
+++ b/mlir/test/Transforms/loop-fusion-transformation.mlir
@@ -2,7 +2,7 @@
// CHECK-LABEL: func @slice_depth1_loop_nest() {
func @slice_depth1_loop_nest() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 16 {
affine.store %cst, %0[%i0] : memref<100xf32>
@@ -24,9 +24,9 @@ func @slice_depth1_loop_nest() {
// CHECK-LABEL: func @should_fuse_reduction_to_pointwise() {
func @should_fuse_reduction_to_pointwise() {
- %a = alloc() : memref<10x10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10x10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -64,9 +64,9 @@ func @should_fuse_reduction_to_pointwise() {
// CHECK-LABEL: func @should_fuse_avoiding_dependence_cycle() {
func @should_fuse_avoiding_dependence_cycle() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
diff --git a/mlir/test/Transforms/loop-fusion.mlir b/mlir/test/Transforms/loop-fusion.mlir
index e800076a8fe3..74f60e2e1e73 100644
--- a/mlir/test/Transforms/loop-fusion.mlir
+++ b/mlir/test/Transforms/loop-fusion.mlir
@@ -12,7 +12,7 @@
// CHECK-LABEL: func @should_fuse_raw_dep_for_locality() {
func @should_fuse_raw_dep_for_locality() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -33,9 +33,9 @@ func @should_fuse_raw_dep_for_locality() {
// CHECK-LABEL: func @should_fuse_reduction_to_pointwise() {
func @should_fuse_reduction_to_pointwise() {
- %a = alloc() : memref<10x10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10x10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -76,7 +76,7 @@ func @should_fuse_reduction_to_pointwise() {
// CHECK-LABEL: func @should_fuse_loop_nests_with_shifts() {
func @should_fuse_loop_nests_with_shifts() {
- %a = alloc() : memref<10x10xf32>
+ %a = memref.alloc() : memref<10x10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 9 {
@@ -116,8 +116,8 @@ func @should_fuse_loop_nests_with_shifts() {
// CHECK-LABEL: func @should_fuse_loop_nest() {
func @should_fuse_loop_nest() {
- %a = alloc() : memref<10x10xf32>
- %b = alloc() : memref<10x10xf32>
+ %a = memref.alloc() : memref<10x10xf32>
+ %b = memref.alloc() : memref<10x10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -137,8 +137,8 @@ func @should_fuse_loop_nest() {
}
}
// Expecting private memref for '%a' first, then private memref for '%b'.
- // CHECK-DAG: [[NEWA:%[0-9]+]] = alloc() : memref<1x1xf32>
- // CHECK-DAG: [[NEWB:%[0-9]+]] = alloc() : memref<1x1xf32>
+ // CHECK-DAG: [[NEWA:%[0-9]+]] = memref.alloc() : memref<1x1xf32>
+ // CHECK-DAG: [[NEWB:%[0-9]+]] = memref.alloc() : memref<1x1xf32>
// CHECK: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, [[NEWA]][0, 0] : memref<1x1xf32>
@@ -155,9 +155,9 @@ func @should_fuse_loop_nest() {
// CHECK-LABEL: func @should_fuse_across_intermediate_loop_with_no_deps() {
func @should_fuse_across_intermediate_loop_with_no_deps() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -190,8 +190,8 @@ func @should_fuse_across_intermediate_loop_with_no_deps() {
// CHECK-LABEL: func @should_fuse_all_loops() {
func @should_fuse_all_loops() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
// Set up flow dependences from first and second loops to third.
@@ -208,8 +208,8 @@ func @should_fuse_all_loops() {
// Should fuse first and second loops into third.
// Expecting private memref for '%a' first, then private memref for '%b'.
- // CHECK-DAG: [[NEWA:%[0-9]+]] = alloc() : memref<1xf32>
- // CHECK-DAG: [[NEWB:%[0-9]+]] = alloc() : memref<1xf32>
+ // CHECK-DAG: [[NEWA:%[0-9]+]] = memref.alloc() : memref<1xf32>
+ // CHECK-DAG: [[NEWB:%[0-9]+]] = memref.alloc() : memref<1xf32>
// CHECK: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, [[NEWA]][0] : memref<1xf32>
// CHECK-NEXT: affine.store %{{.*}}, [[NEWB]][0] : memref<1xf32>
@@ -224,9 +224,9 @@ func @should_fuse_all_loops() {
// CHECK-LABEL: func @should_fuse_first_and_second_loops() {
func @should_fuse_first_and_second_loops() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -260,9 +260,9 @@ func @should_fuse_first_and_second_loops() {
// CHECK-LABEL: func @should_not_fuse_would_create_cycle() {
func @should_not_fuse_would_create_cycle() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -303,7 +303,7 @@ func @should_not_fuse_would_create_cycle() {
// CHECK-LABEL: func @should_fuse_producer_consumer() {
func @should_fuse_producer_consumer() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -319,7 +319,7 @@ func @should_fuse_producer_consumer() {
// %i1, but OK to fuse %i1 into %i2.
// TODO: When the fusion pass is run to a fixed-point, it should
// fuse all three of these loop nests.
- // CHECK: alloc() : memref<1xf32>
+ // CHECK: memref.alloc() : memref<1xf32>
// CHECK: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[0] : memref<1xf32>
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[0] : memref<1xf32>
@@ -333,8 +333,8 @@ func @should_fuse_producer_consumer() {
// CHECK-LABEL: func @should_fuse_and_move_to_preserve_war_dep() {
func @should_fuse_and_move_to_preserve_war_dep() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -366,7 +366,7 @@ func @should_fuse_and_move_to_preserve_war_dep() {
// CHECK-LABEL: func @should_fuse_if_top_level_access() {
func @should_fuse_if_top_level_access() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -380,8 +380,8 @@ func @should_fuse_if_top_level_access() {
%v1 = affine.load %m[%c0] : memref<10xf32>
// Top-level load to '%m' should prevent creating a private memref but
// loop nests should be fused and '%i0' should be removed.
- // CHECK: %[[m:.*]] = alloc() : memref<10xf32>
- // CHECK-NOT: alloc
+ // CHECK: %[[m:.*]] = memref.alloc() : memref<10xf32>
+ // CHECK-NOT: memref.alloc
// CHECK: affine.for %[[i1:.*]] = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, %[[m]][%[[i1]]] : memref<10xf32>
@@ -395,7 +395,7 @@ func @should_fuse_if_top_level_access() {
// CHECK-LABEL: func @should_fuse_but_not_remove_src() {
func @should_fuse_but_not_remove_src() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 100 {
@@ -424,7 +424,7 @@ func @should_fuse_but_not_remove_src() {
// CHECK-LABEL: func @should_fuse_no_top_level_access() {
func @should_fuse_no_top_level_access() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -447,7 +447,7 @@ func @should_fuse_no_top_level_access() {
// CHECK-LABEL: func @should_not_fuse_if_inst_at_top_level() {
func @should_not_fuse_if_inst_at_top_level() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -475,7 +475,7 @@ func @should_not_fuse_if_inst_at_top_level() {
// CHECK-LABEL: func @should_not_fuse_if_inst_in_loop_nest() {
func @should_not_fuse_if_inst_in_loop_nest() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%c4 = constant 4 : index
@@ -504,7 +504,7 @@ func @should_not_fuse_if_inst_in_loop_nest() {
// CHECK-LABEL: func @permute_and_fuse() {
func @permute_and_fuse() {
- %m = alloc() : memref<10x20x30xf32>
+ %m = memref.alloc() : memref<10x20x30xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -545,7 +545,7 @@ func @permute_and_fuse() {
// Reshape from a 64 x f32 to 16 x 4 x f32.
// CHECK-LABEL: func @fuse_reshape_64_16_4
func @fuse_reshape_64_16_4(%in : memref<64xf32>) {
- %out = alloc() : memref<16x4xf32>
+ %out = memref.alloc() : memref<16x4xf32>
affine.for %i0 = 0 to 64 {
%v = affine.load %in[%i0] : memref<64xf32>
@@ -575,8 +575,8 @@ func @fuse_reshape_64_16_4(%in : memref<64xf32>) {
// Reshape a 16x4xf32 to 64xf32.
// CHECK-LABEL: func @fuse_reshape_16_4_64
func @fuse_reshape_16_4_64() {
- %in = alloc() : memref<16x4xf32>
- %out = alloc() : memref<64xf32>
+ %in = memref.alloc() : memref<16x4xf32>
+ %out = memref.alloc() : memref<64xf32>
affine.for %i0 = 0 to 16 {
affine.for %i1 = 0 to 4 {
@@ -608,9 +608,9 @@ func @fuse_reshape_16_4_64() {
// All three loop nests below (6-d one, 2-d one, 2-d one is fused into a single
// 2-d loop nest).
func @R6_to_R2_reshape_square() -> memref<64x9xi32> {
- %in = alloc() : memref<2x2x3x3x16x1xi32>
- %out = alloc() : memref<64x9xi32>
- %live_out = alloc() : memref<64x9xi32>
+ %in = memref.alloc() : memref<2x2x3x3x16x1xi32>
+ %out = memref.alloc() : memref<64x9xi32>
+ %live_out = memref.alloc() : memref<64x9xi32>
// Initialize input.
affine.for %i0 = 0 to 2 {
@@ -670,9 +670,9 @@ func @R6_to_R2_reshape_square() -> memref<64x9xi32> {
//
// CHECK-LABEL: func @R6_to_R2_reshape
-// CHECK: alloc() : memref<1x2x3x3x16x1xi32>
-// CHECK: alloc() : memref<1x1xi32>
-// CHECK: alloc() : memref<64x9xi32>
+// CHECK: memref.alloc() : memref<1x2x3x3x16x1xi32>
+// CHECK: memref.alloc() : memref<1x1xi32>
+// CHECK: memref.alloc() : memref<64x9xi32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 64 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 9 {
// CHECK-NEXT: affine.apply [[$MAP0]](%{{.*}}, %{{.*}})
@@ -703,7 +703,7 @@ func @R6_to_R2_reshape_square() -> memref<64x9xi32> {
// CHECK-LABEL: func @fuse_symbolic_bounds
func @fuse_symbolic_bounds(%M : index, %N : index) {
%N_plus_5 = affine.apply affine_map<(d0) -> (d0 + 5)>(%N)
- %m = alloc(%M, %N_plus_5) : memref<? x ? x f32>
+ %m = memref.alloc(%M, %N_plus_5) : memref<? x ? x f32>
%c0 = constant 0.0 : f32
%s = constant 5 : index
@@ -727,8 +727,8 @@ func @fuse_symbolic_bounds(%M : index, %N : index) {
// CHECK-LABEL: func @should_fuse_reduction_at_depth_of_one
func @should_fuse_reduction_at_depth_of_one() {
- %a = alloc() : memref<10x100xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10x100xf32>
+ %b = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 100 {
@@ -772,8 +772,8 @@ func @should_fuse_reduction_at_depth_of_one() {
// CHECK-LABEL: func @should_fuse_at_src_depth1_and_dst_depth1
func @should_fuse_at_src_depth1_and_dst_depth1() {
- %a = alloc() : memref<100x16xf32>
- %b = alloc() : memref<100x16xf32>
+ %a = memref.alloc() : memref<100x16xf32>
+ %b = memref.alloc() : memref<100x16xf32>
affine.for %i0 = 0 to 100 {
affine.for %i1 = 0 to 16 {
@@ -820,7 +820,7 @@ func @should_fuse_at_src_depth1_and_dst_depth1() {
// CHECK-LABEL: func @should_fuse_src_depth1_at_dst_depth2
func @should_fuse_src_depth1_at_dst_depth2() {
- %a = alloc() : memref<100xf32>
+ %a = memref.alloc() : memref<100xf32>
%c0 = constant 0.0 : f32
affine.for %i0 = 0 to 100 {
@@ -851,7 +851,7 @@ func @should_fuse_src_depth1_at_dst_depth2() {
// CHECK-LABEL: func @fusion_at_depth0_not_currently_supported
func @fusion_at_depth0_not_currently_supported() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%c0 = constant 0 : index
%cst = constant 0.000000e+00 : f32
affine.for %i0 = 0 to 10 {
@@ -862,7 +862,7 @@ func @fusion_at_depth0_not_currently_supported() {
}
// NOTE: Should shrink memref size to 1 element access by load in dst loop
// nest, and make the store in the slice store to the same element.
- // CHECK-DAG: alloc() : memref<1xf32>
+ // CHECK-DAG: memref.alloc() : memref<1xf32>
// CHECK: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[0] : memref<1xf32>
// CHECK-NEXT: affine.load %{{.*}}[0] : memref<1xf32>
@@ -875,9 +875,9 @@ func @fusion_at_depth0_not_currently_supported() {
// CHECK-LABEL: func @should_fuse_deep_loop_nests
func @should_fuse_deep_loop_nests() {
- %0 = alloc() : memref<2x2x3x3x16x10xf32, 2>
- %1 = alloc() : memref<2x2x3x3x16x10xf32, 2>
- %2 = alloc() : memref<3x3x3x3x16x10xf32, 2>
+ %0 = memref.alloc() : memref<2x2x3x3x16x10xf32, 2>
+ %1 = memref.alloc() : memref<2x2x3x3x16x10xf32, 2>
+ %2 = memref.alloc() : memref<3x3x3x3x16x10xf32, 2>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c1_0 = constant 1 : index
@@ -934,7 +934,7 @@ func @should_fuse_deep_loop_nests() {
// bounds which are a function of the first four loops of destination loop nest,
// where the destination loops nests have been interchanged.
-// CHECK-DAG: alloc() : memref<1x1x1x1x16x10xf32, 2>
+// CHECK-DAG: memref.alloc() : memref<1x1x1x1x16x10xf32, 2>
// CHECK: affine.for %{{.*}} = 0 to 3 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 3 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 2 {
@@ -979,8 +979,8 @@ func @should_fuse_deep_loop_nests() {
// CHECK-LABEL: func @should_fuse_at_depth1_and_reduce_slice_trip_count
func @should_fuse_at_depth1_and_reduce_slice_trip_count() {
- %a = alloc() : memref<4x256xf32>
- %b = alloc() : memref<4x256xf32>
+ %a = memref.alloc() : memref<4x256xf32>
+ %b = memref.alloc() : memref<4x256xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
@@ -1008,7 +1008,7 @@ func @should_fuse_at_depth1_and_reduce_slice_trip_count() {
// NOTE: the size of the private memref created for the fused loop nest
// is reduced from the original shape from 4x256 to 4x16 because of the
// data accessed by the load.
- // CHECK-DAG: alloc() : memref<1x16xf32>
+ // CHECK-DAG: memref.alloc() : memref<1x16xf32>
// CHECK: affine.for %{{.*}} = 0 to 4 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 256 {
// CHECK-NEXT: affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<4x256xf32>
@@ -1028,7 +1028,7 @@ func @should_fuse_at_depth1_and_reduce_slice_trip_count() {
// CHECK-LABEL: func @should_fuse_at_depth1_with_trip_count_20
func @should_fuse_at_depth1_with_trip_count_20() {
- %a = alloc() : memref<100xf32>
+ %a = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
@@ -1047,7 +1047,7 @@ func @should_fuse_at_depth1_with_trip_count_20() {
}
}
// NOTE: The size of the private memref created for fusion is shrunk to 20xf32
- // CHECK-DAG: alloc() : memref<20xf32>
+ // CHECK-DAG: memref.alloc() : memref<20xf32>
// CHECK: affine.for %{{.*}} = 0 to 5 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 20 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<20xf32>
@@ -1069,7 +1069,7 @@ func @should_fuse_at_depth1_with_trip_count_20() {
// CHECK-LABEL: func @should_fuse_at_depth1_with_trip_count_19
func @should_fuse_at_depth1_with_trip_count_19() {
- %a = alloc() : memref<100xf32>
+ %a = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
@@ -1088,7 +1088,7 @@ func @should_fuse_at_depth1_with_trip_count_19() {
}
}
// NOTE: The size of the private memref created for fusion is shrunk to 19xf32
- // CHECK-DAG: alloc() : memref<19xf32>
+ // CHECK-DAG: memref.alloc() : memref<19xf32>
// CHECK: affine.for %{{.*}} = 0 to 5 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 19 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<19xf32>
@@ -1111,7 +1111,7 @@ func @should_fuse_at_depth1_with_trip_count_19() {
// CHECK-LABEL: func @should_fuse_with_private_memrefs_with_
diff _shapes() {
func @should_fuse_with_private_memrefs_with_
diff _shapes() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 100 {
@@ -1125,8 +1125,8 @@ func @should_fuse_with_private_memrefs_with_
diff _shapes() {
}
// Should create two new private memrefs customized to the shapes accessed
// by loops %{{.*}} and %{{.*}}.
- // CHECK-DAG: alloc() : memref<1xf32>
- // CHECK-DAG: alloc() : memref<1xf32>
+ // CHECK-DAG: memref.alloc() : memref<1xf32>
+ // CHECK-DAG: memref.alloc() : memref<1xf32>
// CHECK: affine.for %{{.*}} = 0 to 17 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[0] : memref<1xf32>
// CHECK-NEXT: affine.load %{{.*}}[0] : memref<1xf32>
@@ -1195,7 +1195,7 @@ func @should_fuse_live_out_arg(%arg0: memref<10xf32>) {
// CHECK-LABEL: func @should_fuse_escaping_memref_but_preserve_src_loop() -> memref<10xf32>
func @should_fuse_escaping_memref_but_preserve_src_loop() -> memref<10xf32> {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
}
@@ -1206,7 +1206,7 @@ func @should_fuse_escaping_memref_but_preserve_src_loop() -> memref<10xf32> {
// because it writes to memref '%m', which is returned by the function, and
// the '%i1' memory region does not cover '%i0' memory region.
- // CHECK-DAG: alloc() : memref<10xf32>
+ // CHECK-DAG: memref.alloc() : memref<10xf32>
// CHECK: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: }
@@ -1221,7 +1221,7 @@ func @should_fuse_escaping_memref_but_preserve_src_loop() -> memref<10xf32> {
// This should fuse with the %in becoming a 1x1x1.
func @R3_to_R2_reshape() {
- %in = alloc() : memref<2x3x16xi32>
+ %in = memref.alloc() : memref<2x3x16xi32>
%c0 = constant 0 : index
@@ -1249,7 +1249,7 @@ func @R3_to_R2_reshape() {
// CHECK-DAG: [[$MAP2:#map[0-9]+]] = affine_map<(d0) -> (d0 floordiv 48)>
// CHECK-LABEL: func @R3_to_R2_reshape()
-// CHECK-DAG: alloc() : memref<1x1x1xi32>
+// CHECK-DAG: memref.alloc() : memref<1x1x1xi32>
// CHECK: affine.for %{{.*}} = 0 to 32 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 3 {
// CHECK-NEXT: affine.apply [[$MAP0]](%{{.*}}, %{{.*}})
@@ -1265,8 +1265,8 @@ func @R3_to_R2_reshape() {
// -----
func @should_fuse_multi_output_producer() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -1293,9 +1293,9 @@ func @should_fuse_multi_output_producer() {
// CHECK-LABEL: func @fusion_preventing_deps_on_middle_loop() {
func @fusion_preventing_deps_on_middle_loop() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -1334,9 +1334,9 @@ func @fusion_preventing_deps_on_middle_loop() {
// CHECK-LABEL: func @should_fuse_and_move_to_preserve_war_dep() {
func @should_fuse_and_move_to_preserve_war_dep() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -1368,7 +1368,7 @@ func @should_fuse_and_move_to_preserve_war_dep() {
// It is possible to fuse loop '%i0' into '%i3' and preserve dependences
// if the fused loop nest is inserted between loops '%i1' and '%i2'.
- // CHECK-DAG: alloc() : memref<1xf32>
+ // CHECK-DAG: memref.alloc() : memref<1xf32>
// CHECK: affine.for %{{.*}} = 0 to 3 {
// CHECK-NEXT: affine.load %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: }
@@ -1389,9 +1389,9 @@ func @should_fuse_and_move_to_preserve_war_dep() {
// CHECK-LABEL: func @fusion_preventing_dep_on_constant() {
func @fusion_preventing_dep_on_constant() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -1430,9 +1430,9 @@ func @fusion_preventing_dep_on_constant() {
// CHECK-LABEL: func @should_fuse_and_preserve_dep_on_constant() {
func @should_fuse_and_preserve_dep_on_constant() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf11 = constant 11.0 : f32
@@ -1470,7 +1470,7 @@ func @should_fuse_and_preserve_dep_on_constant() {
// CHECK-LABEL: func @should_fuse_at_depth_above_loop_carried_dependence(%{{.*}}: memref<64x4xf32>, %{{.*}}: memref<64x4xf32>) {
func @should_fuse_at_depth_above_loop_carried_dependence(%arg0: memref<64x4xf32>, %arg1: memref<64x4xf32>) {
- %out = alloc() : memref<64x4xf32>
+ %out = memref.alloc() : memref<64x4xf32>
%0 = constant 0.0 : f32
affine.for %i0 = 0 to 64 {
affine.for %i1 = 0 to 4 {
@@ -1506,7 +1506,7 @@ func @should_fuse_at_depth_above_loop_carried_dependence(%arg0: memref<64x4xf32>
// loop nest iteration bounds on its loop '%i1' are reduced to 1, so the
// memref size can be reduced to 128x1xf32.
- // CHECK: alloc() : memref<64x1xf32>
+ // CHECK: memref.alloc() : memref<64x1xf32>
// CHECK: affine.for %{{.*}} = 0 to 4 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 64 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}, 0] : memref<64x1xf32>
@@ -1538,8 +1538,8 @@ func @should_fuse_at_depth_above_loop_carried_dependence(%arg0: memref<64x4xf32>
// CHECK-LABEL: func @should_fuse_only_two_loops_and_remove_producer() {
func @should_fuse_only_two_loops_and_remove_producer() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
@@ -1582,7 +1582,7 @@ func @should_fuse_only_two_loops_and_remove_producer() {
// CHECK-LABEL: func @should_fuse_after_one_loop_interchange() {
func @should_fuse_after_one_loop_interchange() {
- %a = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
%cf0 = constant 0.0 : f32
affine.for %i0 = 0 to 10 {
@@ -1617,7 +1617,7 @@ func @should_fuse_after_one_loop_interchange() {
// CHECK-LABEL: func @should_fuse_after_two_loop_interchanges() {
func @should_fuse_after_two_loop_interchanges() {
- %a = alloc() : memref<6x8xf32>
+ %a = memref.alloc() : memref<6x8xf32>
%cf0 = constant 0.0 : f32
affine.for %i0 = 0 to 6 {
@@ -1739,8 +1739,8 @@ func @slice_tile(%arg0: memref<128x8xf32>, %arg1: memref<32x8xf32>, %0 : f32) ->
// Test case which illustrates fix for b/126454413
func @test_add_slice_bounds() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%c0 = constant 0 : index
@@ -1785,7 +1785,7 @@ func @test_add_slice_bounds() {
// -----
func @should_fuse_init_loops_siblings_then_shared_producer(%arg0: memref<10x10xf32>, %arg1: memref<10x10xf32>) {
- %0 = alloc() : memref<10x10xf32>
+ %0 = memref.alloc() : memref<10x10xf32>
%cst = constant 0.000000e+00 : f32
%cst_0 = constant 1.000000e+00 : f32
%cst_1 = constant 7.000000e+00 : f32
@@ -1852,11 +1852,11 @@ func @should_fuse_init_loops_siblings_then_shared_producer(%arg0: memref<10x10xf
// -----
func @two_matrix_vector_products() {
- %in_matrix = alloc() : memref<10x10xf32>
- %in_vec0 = alloc() : memref<10xf32>
- %in_vec1 = alloc() : memref<10xf32>
- %out_vec0 = alloc() : memref<10xf32>
- %out_vec1 = alloc() : memref<10xf32>
+ %in_matrix = memref.alloc() : memref<10x10xf32>
+ %in_vec0 = memref.alloc() : memref<10xf32>
+ %in_vec1 = memref.alloc() : memref<10xf32>
+ %out_vec0 = memref.alloc() : memref<10xf32>
+ %out_vec1 = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
// Populate input matrix.
@@ -1916,7 +1916,7 @@ func @two_matrix_vector_products() {
// -----
func @should_not_slice_past_slice_barrier() {
- %0 = alloc() : memref<100x16xf32>
+ %0 = memref.alloc() : memref<100x16xf32>
affine.for %i0 = 0 to 100 {
affine.for %i1 = 0 to 16 {
%1 = "op1"() : () -> f32
@@ -1948,7 +1948,7 @@ func @should_not_slice_past_slice_barrier() {
#map0 = affine_map<(d0, d1) -> (d0 * 16 + d1)>
func @fuse_across_dim_mismatch(%arg0: memref<4x4x16x1xf32>, %arg1: memref<144x9xf32>, %arg2: memref<9xf32>) {
- %1 = alloc() : memref<144x4xf32>
+ %1 = memref.alloc() : memref<144x4xf32>
%2 = constant 0.0 : f32
affine.for %i2 = 0 to 9 {
affine.for %i3 = 0 to 4 {
@@ -1972,7 +1972,7 @@ func @fuse_across_dim_mismatch(%arg0: memref<4x4x16x1xf32>, %arg1: memref<144x9x
}
// MAXIMAL: #map = affine_map<(d0, d1) -> (d0 * 16 + d1)>
// MAXIMAL-LABEL: func @fuse_across_dim_mismatch
-// MAXIMAL: alloc() : memref<1x1xf32>
+// MAXIMAL: memref.alloc() : memref<1x1xf32>
// MAXIMAL: affine.for %{{.*}} = 0 to 9 {
// MAXIMAL-NEXT: affine.for %{{.*}} = 0 to 9 {
// MAXIMAL-NEXT: affine.for %{{.*}} = 0 to 4 {
@@ -1998,9 +1998,9 @@ func @fuse_across_dim_mismatch(%arg0: memref<4x4x16x1xf32>, %arg1: memref<144x9x
#map12 = affine_map<(d0, d1) -> (d0 * 16 - d1 + 15)>
func @fuse_across_varying_dims_complex(%arg0: f32) {
%c0 = constant 0 : index
- %0 = alloc() : memref<2x2x3x3x16x1xf32>
- %1 = alloc() : memref<64x9xf32>
- %2 = alloc() : memref<144x4xf32>
+ %0 = memref.alloc() : memref<2x2x3x3x16x1xf32>
+ %1 = memref.alloc() : memref<64x9xf32>
+ %2 = memref.alloc() : memref<144x4xf32>
affine.for %i0 = 0 to 64 {
affine.for %i1 = 0 to 9 {
%4 = affine.apply #map3(%i0, %i1)
@@ -2044,10 +2044,10 @@ func @fuse_across_varying_dims_complex(%arg0: f32) {
// MAXIMAL-DAG: [[$MAP7:#map[0-9]+]] = affine_map<(d0, d1) -> (d0 * 16 + d1)>
// MAXIMAL-DAG: [[$MAP8:#map[0-9]+]] = affine_map<(d0, d1) -> (d0 * 16 - d1 + 15)>
// MAXIMAL-LABEL: func @fuse_across_varying_dims_complex
-// MAXIMAL-NEXT: alloc() : memref<64x1xf32>
+// MAXIMAL-NEXT: memref.alloc() : memref<64x1xf32>
// MAXIMAL-NEXT: constant 0 : index
-// MAXIMAL-NEXT: alloc() : memref<2x2x3x3x16x1xf32>
-// MAXIMAL-NEXT: alloc() : memref<144x4xf32>
+// MAXIMAL-NEXT: memref.alloc() : memref<2x2x3x3x16x1xf32>
+// MAXIMAL-NEXT: memref.alloc() : memref<144x4xf32>
// MAXIMAL-NEXT: affine.for %{{.*}} = 0 to 9 {
// MAXIMAL-NEXT: affine.for %{{.*}} = 0 to 9 {
// MAXIMAL-NEXT: affine.for %{{.*}} = 0 to 4 {
@@ -2081,7 +2081,7 @@ func @fuse_across_varying_dims_complex(%arg0: f32) {
// -----
func @should_fuse_with_slice_union() {
- %a = alloc() : memref<100xf32>
+ %a = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
@@ -2284,8 +2284,8 @@ func @affine_2_dependent_mm_fused(%arg0: memref<1024x1024xf32>, %arg1: memref<10
// CHECK-LABEL: func @should_fuse_self_dependence_multi_store_producer() {
func @should_fuse_self_dependence_multi_store_producer() {
- %m = alloc() : memref<10xf32>
- %local_m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
+ %local_m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -2310,8 +2310,8 @@ func @should_fuse_self_dependence_multi_store_producer() {
// CHECK-LABEL: func @should_fuse_dead_multi_store_producer() {
func @should_fuse_dead_multi_store_producer() {
- %m = alloc() : memref<10xf32>
- %dead_m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
+ %dead_m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -2334,7 +2334,7 @@ func @should_fuse_dead_multi_store_producer() {
// CHECK-LABEL: func @should_fuse_function_live_out_multi_store_producer
func @should_fuse_function_live_out_multi_store_producer(%live_in_out_m : memref<10xf32>) {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -2359,7 +2359,7 @@ func @should_fuse_function_live_out_multi_store_producer(%live_in_out_m : memref
// CHECK-LABEL: func @mul_add_0
func @mul_add_0(%arg0: memref<3x4xf32>, %arg1: memref<4x3xf32>, %arg2: memref<3x3xf32>, %arg3: memref<3x3xf32>) {
%cst = constant 0.000000e+00 : f32
- %0 = alloc() : memref<3x3xf32>
+ %0 = memref.alloc() : memref<3x3xf32>
affine.for %arg4 = 0 to 3 {
affine.for %arg5 = 0 to 3 {
affine.store %cst, %0[%arg4, %arg5] : memref<3x3xf32>
@@ -2468,7 +2468,7 @@ func @should_fuse_producer_with_multi_outgoing_edges(%a : memref<1xf32>, %b : me
// MAXIMAL-LABEL: func @reshape_into_matmul
func @reshape_into_matmul(%lhs : memref<1024x1024xf32>,
%R: memref<16x64x1024xf32>, %out: memref<1024x1024xf32>) {
- %rhs = alloc() : memref<1024x1024xf32>
+ %rhs = memref.alloc() : memref<1024x1024xf32>
// Reshape from 3-d to 2-d.
affine.for %i0 = 0 to 16 {
@@ -2495,7 +2495,7 @@ func @reshape_into_matmul(%lhs : memref<1024x1024xf32>,
}
return
}
-// MAXIMAL-NEXT: alloc
+// MAXIMAL-NEXT: memref.alloc
// MAXIMAL-NEXT: affine.for
// MAXIMAL-NEXT: affine.for
// MAXIMAL-NEXT: affine.for
@@ -2580,7 +2580,7 @@ func @multi_outgoing_edges(%in0 : memref<32xf32>,
// CHECK-LABEL: func @calc
func @calc(%arg0: memref<?xf32>, %arg1: memref<?xf32>, %arg2: memref<?xf32>, %len: index) {
%c1 = constant 1 : index
- %1 = alloc(%len) : memref<?xf32>
+ %1 = memref.alloc(%len) : memref<?xf32>
affine.for %arg4 = 1 to 10 {
%7 = affine.load %arg0[%arg4] : memref<?xf32>
%8 = affine.load %arg1[%arg4] : memref<?xf32>
@@ -2595,7 +2595,7 @@ func @calc(%arg0: memref<?xf32>, %arg1: memref<?xf32>, %arg2: memref<?xf32>, %le
}
return
}
-// CHECK: alloc() : memref<1xf32>
+// CHECK: memref.alloc() : memref<1xf32>
// CHECK: affine.for %arg{{.*}} = 1 to 10 {
// CHECK-NEXT: affine.load %arg{{.*}}
// CHECK-NEXT: affine.load %arg{{.*}}
@@ -2620,10 +2620,10 @@ func @should_not_fuse_since_non_affine_users(%in0 : memref<32xf32>,
affine.store %add, %in0[%d] : memref<32xf32>
}
affine.for %d = 0 to 32 {
- %lhs = load %in0[%d] : memref<32xf32>
- %rhs = load %in1[%d] : memref<32xf32>
+ %lhs = memref.load %in0[%d] : memref<32xf32>
+ %rhs = memref.load %in1[%d] : memref<32xf32>
%add = subf %lhs, %rhs : f32
- store %add, %in0[%d] : memref<32xf32>
+ memref.store %add, %in0[%d] : memref<32xf32>
}
affine.for %d = 0 to 32 {
%lhs = affine.load %in0[%d] : memref<32xf32>
@@ -2646,15 +2646,15 @@ func @should_not_fuse_since_non_affine_users(%in0 : memref<32xf32>,
// CHECK-LABEL: func @should_not_fuse_since_top_level_non_affine_users
func @should_not_fuse_since_top_level_non_affine_users(%in0 : memref<32xf32>,
%in1 : memref<32xf32>) {
- %sum = alloc() : memref<f32>
+ %sum = memref.alloc() : memref<f32>
affine.for %d = 0 to 32 {
%lhs = affine.load %in0[%d] : memref<32xf32>
%rhs = affine.load %in1[%d] : memref<32xf32>
%add = addf %lhs, %rhs : f32
- store %add, %sum[] : memref<f32>
+ memref.store %add, %sum[] : memref<f32>
affine.store %add, %in0[%d] : memref<32xf32>
}
- %load_sum = load %sum[] : memref<f32>
+ %load_sum = memref.load %sum[] : memref<f32>
affine.for %d = 0 to 32 {
%lhs = affine.load %in0[%d] : memref<32xf32>
%rhs = affine.load %in1[%d] : memref<32xf32>
@@ -2662,7 +2662,7 @@ func @should_not_fuse_since_top_level_non_affine_users(%in0 : memref<32xf32>,
%sub = subf %add, %load_sum: f32
affine.store %sub, %in0[%d] : memref<32xf32>
}
- dealloc %sum : memref<f32>
+ memref.dealloc %sum : memref<f32>
return
}
@@ -2686,7 +2686,7 @@ func @should_not_fuse_since_top_level_non_affine_mem_write_users(
%add = addf %lhs, %rhs : f32
affine.store %add, %in0[%d] : memref<32xf32>
}
- store %cst_0, %in0[%c0] : memref<32xf32>
+ memref.store %cst_0, %in0[%c0] : memref<32xf32>
affine.for %d = 0 to 32 {
%lhs = affine.load %in0[%d] : memref<32xf32>
%rhs = affine.load %in1[%d] : memref<32xf32>
@@ -2705,7 +2705,7 @@ func @should_not_fuse_since_top_level_non_affine_mem_write_users(
// MAXIMAL-LABEL: func @fuse_minor_affine_map
func @fuse_minor_affine_map(%in: memref<128xf32>, %out: memref<20x512xf32>) {
- %tmp = alloc() : memref<128xf32>
+ %tmp = memref.alloc() : memref<128xf32>
affine.for %arg4 = 0 to 128 {
%ld = affine.load %in[%arg4] : memref<128xf32>
@@ -2725,7 +2725,7 @@ func @fuse_minor_affine_map(%in: memref<128xf32>, %out: memref<20x512xf32>) {
// TODO: The size of the private memref is not properly computed in the presence
// of the 'mod' operation. It should be memref<1xf32> instead of
// memref<128xf32>: https://bugs.llvm.org/show_bug.cgi?id=46973
-// MAXIMAL: alloc() : memref<128xf32>
+// MAXIMAL: memref.alloc() : memref<128xf32>
// MAXIMAL: affine.for
// MAXIMAL-NEXT: affine.for
// MAXIMAL-NOT: affine.for
@@ -2735,9 +2735,9 @@ func @fuse_minor_affine_map(%in: memref<128xf32>, %out: memref<20x512xf32>) {
// CHECK-LABEL: func @should_fuse_multi_store_producer_and_privatize_memfefs
func @should_fuse_multi_store_producer_and_privatize_memfefs() {
- %a = alloc() : memref<10xf32>
- %b = alloc() : memref<10xf32>
- %c = alloc() : memref<10xf32>
+ %a = memref.alloc() : memref<10xf32>
+ %b = memref.alloc() : memref<10xf32>
+ %c = memref.alloc() : memref<10xf32>
%cst = constant 0.000000e+00 : f32
affine.for %arg0 = 0 to 10 {
affine.store %cst, %a[%arg0] : memref<10xf32>
@@ -2839,22 +2839,22 @@ func @should_fuse_multi_store_producer_with_scaping_memrefs_and_preserve_src(
// -----
func @should_not_fuse_due_to_dealloc(%arg0: memref<16xf32>){
- %A = alloc() : memref<16xf32>
- %C = alloc() : memref<16xf32>
+ %A = memref.alloc() : memref<16xf32>
+ %C = memref.alloc() : memref<16xf32>
%cst_1 = constant 1.000000e+00 : f32
affine.for %arg1 = 0 to 16 {
%a = affine.load %arg0[%arg1] : memref<16xf32>
affine.store %a, %A[%arg1] : memref<16xf32>
affine.store %a, %C[%arg1] : memref<16xf32>
}
- dealloc %C : memref<16xf32>
- %B = alloc() : memref<16xf32>
+ memref.dealloc %C : memref<16xf32>
+ %B = memref.alloc() : memref<16xf32>
affine.for %arg1 = 0 to 16 {
%a = affine.load %A[%arg1] : memref<16xf32>
%b = addf %cst_1, %a : f32
affine.store %b, %B[%arg1] : memref<16xf32>
}
- dealloc %A : memref<16xf32>
+ memref.dealloc %A : memref<16xf32>
return
}
// CHECK-LABEL: func @should_not_fuse_due_to_dealloc
@@ -2862,7 +2862,7 @@ func @should_not_fuse_due_to_dealloc(%arg0: memref<16xf32>){
// CHECK-NEXT: affine.load
// CHECK-NEXT: affine.store
// CHECK-NEXT: affine.store
-// CHECK: dealloc
+// CHECK: memref.dealloc
// CHECK: affine.for
// CHECK-NEXT: affine.load
// CHECK-NEXT: addf
@@ -3021,14 +3021,14 @@ func @should_not_fuse_src_loop_nest_return_value(
func private @some_function(memref<16xf32>)
func @call_op_prevents_fusion(%arg0: memref<16xf32>){
- %A = alloc() : memref<16xf32>
+ %A = memref.alloc() : memref<16xf32>
%cst_1 = constant 1.000000e+00 : f32
affine.for %arg1 = 0 to 16 {
%a = affine.load %arg0[%arg1] : memref<16xf32>
affine.store %a, %A[%arg1] : memref<16xf32>
}
call @some_function(%A) : (memref<16xf32>) -> ()
- %B = alloc() : memref<16xf32>
+ %B = memref.alloc() : memref<16xf32>
affine.for %arg1 = 0 to 16 {
%a = affine.load %A[%arg1] : memref<16xf32>
%b = addf %cst_1, %a : f32
@@ -3050,14 +3050,14 @@ func @call_op_prevents_fusion(%arg0: memref<16xf32>){
func private @some_function()
func @call_op_does_not_prevent_fusion(%arg0: memref<16xf32>){
- %A = alloc() : memref<16xf32>
+ %A = memref.alloc() : memref<16xf32>
%cst_1 = constant 1.000000e+00 : f32
affine.for %arg1 = 0 to 16 {
%a = affine.load %arg0[%arg1] : memref<16xf32>
affine.store %a, %A[%arg1] : memref<16xf32>
}
call @some_function() : () -> ()
- %B = alloc() : memref<16xf32>
+ %B = memref.alloc() : memref<16xf32>
affine.for %arg1 = 0 to 16 {
%a = affine.load %A[%arg1] : memref<16xf32>
%b = addf %cst_1, %a : f32
diff --git a/mlir/test/Transforms/loop-invariant-code-motion.mlir b/mlir/test/Transforms/loop-invariant-code-motion.mlir
index 7f299b6b645d..198b925f876f 100644
--- a/mlir/test/Transforms/loop-invariant-code-motion.mlir
+++ b/mlir/test/Transforms/loop-invariant-code-motion.mlir
@@ -1,7 +1,7 @@
// RUN: mlir-opt %s -split-input-file -loop-invariant-code-motion | FileCheck %s
func @nested_loops_both_having_invariant_code() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -13,7 +13,7 @@ func @nested_loops_both_having_invariant_code() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %[[CST0:.*]] = constant 7.000000e+00 : f32
// CHECK-NEXT: %[[CST1:.*]] = constant 8.000000e+00 : f32
// CHECK-NEXT: %[[ADD0:.*]] = addf %[[CST0]], %[[CST1]] : f32
@@ -28,7 +28,7 @@ func @nested_loops_both_having_invariant_code() {
// -----
func @nested_loops_code_invariant_to_both() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
@@ -38,7 +38,7 @@ func @nested_loops_code_invariant_to_both() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -49,8 +49,8 @@ func @nested_loops_code_invariant_to_both() {
// -----
func @single_loop_nothing_invariant() {
- %m1 = alloc() : memref<10xf32>
- %m2 = alloc() : memref<10xf32>
+ %m1 = memref.alloc() : memref<10xf32>
+ %m2 = memref.alloc() : memref<10xf32>
affine.for %arg0 = 0 to 10 {
%v0 = affine.load %m1[%arg0] : memref<10xf32>
%v1 = affine.load %m2[%arg0] : memref<10xf32>
@@ -58,8 +58,8 @@ func @single_loop_nothing_invariant() {
affine.store %v2, %m1[%arg0] : memref<10xf32>
}
- // CHECK: %0 = alloc() : memref<10xf32>
- // CHECK-NEXT: %1 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
+ // CHECK-NEXT: %1 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: %2 = affine.load %0[%arg0] : memref<10xf32>
// CHECK-NEXT: %3 = affine.load %1[%arg0] : memref<10xf32>
@@ -72,7 +72,7 @@ func @single_loop_nothing_invariant() {
// -----
func @invariant_code_inside_affine_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
@@ -84,7 +84,7 @@ func @invariant_code_inside_affine_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %arg0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply #map(%arg0)
@@ -100,7 +100,7 @@ func @invariant_code_inside_affine_if() {
// -----
func @invariant_affine_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -110,7 +110,7 @@ func @invariant_affine_if() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %[[CST:.*]] = constant 8.000000e+00 : f32
// CHECK-NEXT: affine.for %[[ARG:.*]] = 0 to 10 {
// CHECK-NEXT: }
@@ -125,7 +125,7 @@ func @invariant_affine_if() {
// -----
func @invariant_affine_if2() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -136,7 +136,7 @@ func @invariant_affine_if2() {
}
}
- // CHECK: alloc
+ // CHECK: memref.alloc
// CHECK-NEXT: constant
// CHECK-NEXT: affine.for
// CHECK-NEXT: affine.for
@@ -152,7 +152,7 @@ func @invariant_affine_if2() {
// -----
func @invariant_affine_nested_if() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -165,7 +165,7 @@ func @invariant_affine_nested_if() {
}
}
- // CHECK: alloc
+ // CHECK: memref.alloc
// CHECK-NEXT: constant
// CHECK-NEXT: affine.for
// CHECK-NEXT: affine.for
@@ -184,7 +184,7 @@ func @invariant_affine_nested_if() {
// -----
func @invariant_affine_nested_if_else() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf8 = constant 8.0 : f32
affine.for %arg0 = 0 to 10 {
affine.for %arg1 = 0 to 10 {
@@ -200,7 +200,7 @@ func @invariant_affine_nested_if_else() {
}
}
- // CHECK: alloc
+ // CHECK: memref.alloc
// CHECK-NEXT: constant
// CHECK-NEXT: affine.for
// CHECK-NEXT: affine.for
@@ -225,7 +225,7 @@ func @invariant_loop_dialect() {
%ci0 = constant 0 : index
%ci10 = constant 10 : index
%ci1 = constant 1 : index
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
scf.for %arg0 = %ci0 to %ci10 step %ci1 {
@@ -234,7 +234,7 @@ func @invariant_loop_dialect() {
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: %cst = constant 7.000000e+00 : f32
// CHECK-NEXT: %cst_0 = constant 8.000000e+00 : f32
// CHECK-NEXT: %1 = addf %cst, %cst_0 : f32
@@ -248,14 +248,14 @@ func @variant_loop_dialect() {
%ci0 = constant 0 : index
%ci10 = constant 10 : index
%ci1 = constant 1 : index
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
scf.for %arg0 = %ci0 to %ci10 step %ci1 {
scf.for %arg1 = %ci0 to %ci10 step %ci1 {
%v0 = addi %arg0, %arg1 : index
}
}
- // CHECK: %0 = alloc() : memref<10xf32>
+ // CHECK: %0 = memref.alloc() : memref<10xf32>
// CHECK-NEXT: scf.for
// CHECK-NEXT: scf.for
// CHECK-NEXT: addi
diff --git a/mlir/test/Transforms/memref-bound-check.mlir b/mlir/test/Transforms/memref-bound-check.mlir
index a8d83cc65a2c..25aa343ad3c2 100644
--- a/mlir/test/Transforms/memref-bound-check.mlir
+++ b/mlir/test/Transforms/memref-bound-check.mlir
@@ -8,8 +8,8 @@ func @test() {
%minusone = constant -1 : index
%sym = constant 111 : index
- %A = alloc() : memref<9 x 9 x i32>
- %B = alloc() : memref<111 x i32>
+ %A = memref.alloc() : memref<9 x 9 x i32>
+ %B = memref.alloc() : memref<111 x i32>
affine.for %i = -1 to 10 {
affine.for %j = -1 to 10 {
@@ -41,7 +41,7 @@ func @test() {
// CHECK-LABEL: func @test_mod_floordiv_ceildiv
func @test_mod_floordiv_ceildiv() {
%zero = constant 0 : index
- %A = alloc() : memref<128 x 64 x 64 x i32>
+ %A = memref.alloc() : memref<128 x 64 x 64 x i32>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 256 {
@@ -64,9 +64,9 @@ func @test_mod_floordiv_ceildiv() {
// CHECK-LABEL: func @test_no_out_of_bounds()
func @test_no_out_of_bounds() {
%zero = constant 0 : index
- %A = alloc() : memref<257 x 256 x i32>
- %C = alloc() : memref<257 x i32>
- %B = alloc() : memref<1 x i32>
+ %A = memref.alloc() : memref<257 x 256 x i32>
+ %C = memref.alloc() : memref<257 x i32>
+ %B = memref.alloc() : memref<1 x i32>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 256 {
@@ -90,7 +90,7 @@ func @test_no_out_of_bounds() {
// CHECK-LABEL: func @mod_div
func @mod_div() {
%zero = constant 0 : index
- %A = alloc() : memref<128 x 64 x 64 x i32>
+ %A = memref.alloc() : memref<128 x 64 x 64 x i32>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 256 {
@@ -113,7 +113,7 @@ func @mod_div() {
// Tests with nested mod's and floordiv's.
// CHECK-LABEL: func @mod_floordiv_nested() {
func @mod_floordiv_nested() {
- %A = alloc() : memref<256 x 256 x i32>
+ %A = memref.alloc() : memref<256 x 256 x i32>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 256 {
%idx0 = affine.apply affine_map<(d0, d1) -> ((d0 mod 1024) floordiv 4)>(%i, %j)
@@ -126,7 +126,7 @@ func @mod_floordiv_nested() {
// CHECK-LABEL: func @test_semi_affine_bailout
func @test_semi_affine_bailout(%N : index) {
- %B = alloc() : memref<10 x i32>
+ %B = memref.alloc() : memref<10 x i32>
affine.for %i = 0 to 10 {
%idx = affine.apply affine_map<(d0)[s0] -> (d0 * s0)>(%i)[%N]
%y = affine.load %B[%idx] : memref<10 x i32>
@@ -137,7 +137,7 @@ func @test_semi_affine_bailout(%N : index) {
// CHECK-LABEL: func @multi_mod_floordiv
func @multi_mod_floordiv() {
- %A = alloc() : memref<2x2xi32>
+ %A = memref.alloc() : memref<2x2xi32>
affine.for %ii = 0 to 64 {
%idx0 = affine.apply affine_map<(d0) -> ((d0 mod 147456) floordiv 1152)> (%ii)
%idx1 = affine.apply affine_map<(d0) -> (((d0 mod 147456) mod 1152) floordiv 384)> (%ii)
@@ -149,8 +149,8 @@ func @multi_mod_floordiv() {
// CHECK-LABEL: func @delinearize_mod_floordiv
func @delinearize_mod_floordiv() {
%c0 = constant 0 : index
- %in = alloc() : memref<2x2x3x3x16x1xi32>
- %out = alloc() : memref<64x9xi32>
+ %in = memref.alloc() : memref<2x2x3x3x16x1xi32>
+ %out = memref.alloc() : memref<64x9xi32>
// Reshape '%in' into '%out'.
affine.for %ii = 0 to 64 {
@@ -186,7 +186,7 @@ func @zero_d_memref(%arg0: memref<i32>) {
// CHECK-LABEL: func @out_of_bounds
func @out_of_bounds() {
- %in = alloc() : memref<1xi32>
+ %in = memref.alloc() : memref<1xi32>
%c9 = constant 9 : i32
affine.for %i0 = 10 to 11 {
@@ -208,7 +208,7 @@ func @out_of_bounds() {
// CHECK-LABEL: func @test_complex_mod_floordiv
func @test_complex_mod_floordiv(%arg0: memref<4x4x16x1xf32>) {
%c0 = constant 0 : index
- %0 = alloc() : memref<1x2x3x3x16x1xf32>
+ %0 = memref.alloc() : memref<1x2x3x3x16x1xf32>
affine.for %i0 = 0 to 64 {
affine.for %i1 = 0 to 9 {
%2 = affine.apply #map3(%i0, %i1)
@@ -228,8 +228,8 @@ func @test_complex_mod_floordiv(%arg0: memref<4x4x16x1xf32>) {
// CHECK-LABEL: func @test_mod_bound
func @test_mod_bound() {
- %0 = alloc() : memref<7 x f32>
- %1 = alloc() : memref<6 x f32>
+ %0 = memref.alloc() : memref<7 x f32>
+ %1 = memref.alloc() : memref<6 x f32>
affine.for %i0 = 0 to 4096 {
affine.for %i1 = #map0(%i0) to #map1(%i0) {
affine.load %0[%i1] : memref<7 x f32>
@@ -248,9 +248,9 @@ func @test_mod_bound() {
// CHECK-LABEL: func @test_floordiv_bound
func @test_floordiv_bound() {
- %0 = alloc() : memref<1027 x f32>
- %1 = alloc() : memref<1026 x f32>
- %2 = alloc() : memref<4096 x f32>
+ %0 = memref.alloc() : memref<1027 x f32>
+ %1 = memref.alloc() : memref<1026 x f32>
+ %2 = memref.alloc() : memref<4096 x f32>
%N = constant 2048 : index
affine.for %i0 = 0 to 4096 {
affine.for %i1 = #map0(%i0) to #map1(%i0) {
@@ -287,7 +287,7 @@ func @non_composed_bound_operand(%arg0: memref<1024xf32>) {
// CHECK-LABEL: func @zero_d_memref
func @zero_d_memref() {
- %Z = alloc() : memref<f32>
+ %Z = memref.alloc() : memref<f32>
affine.for %i = 0 to 100 {
affine.load %Z[] : memref<f32>
}
diff --git a/mlir/test/Transforms/memref-dataflow-opt.mlir b/mlir/test/Transforms/memref-dataflow-opt.mlir
index dfda19372813..49a3dcbd2018 100644
--- a/mlir/test/Transforms/memref-dataflow-opt.mlir
+++ b/mlir/test/Transforms/memref-dataflow-opt.mlir
@@ -9,7 +9,7 @@
// CHECK-LABEL: func @simple_store_load() {
func @simple_store_load() {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
%v0 = affine.load %m[%i0] : memref<10xf32>
@@ -29,7 +29,7 @@ func @multi_store_load() {
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
%cf9 = constant 9.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
%v0 = affine.load %m[%i0] : memref<10xf32>
@@ -58,7 +58,7 @@ func @multi_store_load() {
// CHECK-LABEL: func @store_load_affine_apply
func @store_load_affine_apply() -> memref<10x10xf32> {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
%t0 = affine.apply affine_map<(d0, d1) -> (d1 + 1)>(%i0, %i1)
@@ -74,7 +74,7 @@ func @store_load_affine_apply() -> memref<10x10xf32> {
// The memref and its stores won't be erased due to this memref return.
return %m : memref<10x10xf32>
// CHECK: %{{.*}} = constant 7.000000e+00 : f32
-// CHECK-NEXT: %{{.*}} = alloc() : memref<10x10xf32>
+// CHECK-NEXT: %{{.*}} = memref.alloc() : memref<10x10xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: %{{.*}} = affine.apply [[$MAP0]](%{{.*}}, %{{.*}})
@@ -91,7 +91,7 @@ func @store_load_affine_apply() -> memref<10x10xf32> {
// CHECK-LABEL: func @store_load_nested
func @store_load_nested(%N : index) {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to %N {
@@ -116,7 +116,7 @@ func @store_load_nested(%N : index) {
func @multi_store_load_nested_no_fwd(%N : index) {
%cf7 = constant 7.0 : f32
%cf8 = constant 8.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to %N {
@@ -137,7 +137,7 @@ func @multi_store_load_nested_no_fwd(%N : index) {
func @store_load_store_nested_no_fwd(%N : index) {
%cf7 = constant 7.0 : f32
%cf9 = constant 9.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to %N {
@@ -158,7 +158,7 @@ func @multi_store_load_nested_fwd(%N : index) {
%cf8 = constant 8.0 : f32
%cf9 = constant 9.0 : f32
%cf10 = constant 10.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to %N {
@@ -181,7 +181,7 @@ func @multi_store_load_nested_fwd(%N : index) {
// CHECK-LABEL: func @store_load_no_fwd
func @store_load_no_fwd() {
%cf7 = constant 7.0 : f32
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to 10 {
@@ -200,7 +200,7 @@ func @store_load_no_fwd() {
func @store_load_fwd() {
%cf7 = constant 7.0 : f32
%c0 = constant 0 : index
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.store %cf7, %m[%c0] : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
@@ -222,7 +222,7 @@ func @store_load_store_nested_fwd(%N : index) -> f32 {
%cf9 = constant 9.0 : f32
%c0 = constant 0 : index
%c1 = constant 1 : index
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
affine.for %i0 = 0 to 10 {
affine.store %cf7, %m[%i0] : memref<10xf32>
affine.for %i1 = 0 to %N {
@@ -235,7 +235,7 @@ func @store_load_store_nested_fwd(%N : index) -> f32 {
// Due to this load, the memref isn't optimized away.
%v3 = affine.load %m[%c1] : memref<10xf32>
return %v3 : f32
-// CHECK: %{{.*}} = alloc() : memref<10xf32>
+// CHECK: %{{.*}} = memref.alloc() : memref<10xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to 10 {
// CHECK-NEXT: affine.store %{{.*}}, %{{.*}}[%{{.*}}] : memref<10xf32>
// CHECK-NEXT: affine.for %{{.*}} = 0 to %{{.*}} {
@@ -285,7 +285,7 @@ func @refs_not_known_to_be_equal(%A : memref<100 x 100 x f32>, %M : index) {
// The value loaded from %in can directly be stored to %out by eliminating
// store and load from %tmp.
func @vector_forwarding(%in : memref<512xf32>, %out : memref<512xf32>) {
- %tmp = alloc() : memref<512xf32>
+ %tmp = memref.alloc() : memref<512xf32>
affine.for %i = 0 to 16 {
%ld0 = affine.vector_load %in[32*%i] : memref<512xf32>, vector<32xf32>
affine.vector_store %ld0, %tmp[32*%i] : memref<512xf32>, vector<32xf32>
diff --git a/mlir/test/Transforms/memref-dependence-check.mlir b/mlir/test/Transforms/memref-dependence-check.mlir
index fbe0a149e9f4..fdd9c554a2b4 100644
--- a/mlir/test/Transforms/memref-dependence-check.mlir
+++ b/mlir/test/Transforms/memref-dependence-check.mlir
@@ -6,7 +6,7 @@
// CHECK-LABEL: func @store_may_execute_before_load() {
func @store_may_execute_before_load() {
- %m = alloc() : memref<10xf32>
+ %m = memref.alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%c0 = constant 4 : index
// There is no dependence from store 0 to load 1 at depth if we take into account
@@ -33,7 +33,7 @@ func @store_may_execute_before_load() {
// CHECK-LABEL: func @dependent_loops() {
func @dependent_loops() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%cst = constant 7.000000e+00 : f32
// There is a dependence from 0 to 1 at depth 1 (common surrounding loops 0)
// because the first loop with the store dominates the second scf.
@@ -55,8 +55,8 @@ func @dependent_loops() {
// -----
// CHECK-LABEL: func @
diff erent_memrefs() {
func @
diff erent_memrefs() {
- %m.a = alloc() : memref<100xf32>
- %m.b = alloc() : memref<100xf32>
+ %m.a = memref.alloc() : memref<100xf32>
+ %m.b = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%c1 = constant 1.0 : f32
affine.store %c1, %m.a[%c0] : memref<100xf32>
@@ -71,7 +71,7 @@ func @
diff erent_memrefs() {
// -----
// CHECK-LABEL: func @store_load_
diff erent_elements() {
func @store_load_
diff erent_elements() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c7 = constant 7.0 : f32
@@ -87,7 +87,7 @@ func @store_load_
diff erent_elements() {
// -----
// CHECK-LABEL: func @load_store_
diff erent_elements() {
func @load_store_
diff erent_elements() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c7 = constant 7.0 : f32
@@ -103,7 +103,7 @@ func @load_store_
diff erent_elements() {
// -----
// CHECK-LABEL: func @store_load_same_element() {
func @store_load_same_element() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c11 = constant 11 : index
%c7 = constant 7.0 : f32
affine.store %c7, %m[%c11] : memref<100xf32>
@@ -118,7 +118,7 @@ func @store_load_same_element() {
// -----
// CHECK-LABEL: func @load_load_same_element() {
func @load_load_same_element() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c11 = constant 11 : index
%c7 = constant 7.0 : f32
%v0 = affine.load %m[%c11] : memref<100xf32>
@@ -133,7 +133,7 @@ func @load_load_same_element() {
// -----
// CHECK-LABEL: func @store_load_same_symbol(%arg0: index) {
func @store_load_same_symbol(%arg0: index) {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.store %c7, %m[%arg0] : memref<100xf32>
// expected-remark at above {{dependence from 0 to 0 at depth 1 = false}}
@@ -147,7 +147,7 @@ func @store_load_same_symbol(%arg0: index) {
// -----
// CHECK-LABEL: func @store_load_
diff erent_symbols(%arg0: index, %arg1: index) {
func @store_load_
diff erent_symbols(%arg0: index, %arg1: index) {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.store %c7, %m[%arg0] : memref<100xf32>
// expected-remark at above {{dependence from 0 to 0 at depth 1 = false}}
@@ -161,7 +161,7 @@ func @store_load_
diff erent_symbols(%arg0: index, %arg1: index) {
// -----
// CHECK-LABEL: func @store_load_
diff _element_affine_apply_const() {
func @store_load_
diff _element_affine_apply_const() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c1 = constant 1 : index
%c8 = constant 8.0 : f32
%a0 = affine.apply affine_map<(d0) -> (d0)> (%c1)
@@ -178,7 +178,7 @@ func @store_load_
diff _element_affine_apply_const() {
// -----
// CHECK-LABEL: func @store_load_same_element_affine_apply_const() {
func @store_load_same_element_affine_apply_const() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c9 = constant 9 : index
%c11 = constant 11 : index
@@ -196,7 +196,7 @@ func @store_load_same_element_affine_apply_const() {
// -----
// CHECK-LABEL: func @store_load_affine_apply_symbol(%arg0: index) {
func @store_load_affine_apply_symbol(%arg0: index) {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%a0 = affine.apply affine_map<(d0) -> (d0)> (%arg0)
affine.store %c7, %m[%a0] : memref<100xf32>
@@ -212,7 +212,7 @@ func @store_load_affine_apply_symbol(%arg0: index) {
// -----
// CHECK-LABEL: func @store_load_affine_apply_symbol_offset(%arg0: index) {
func @store_load_affine_apply_symbol_offset(%arg0: index) {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%a0 = affine.apply affine_map<(d0) -> (d0)> (%arg0)
affine.store %c7, %m[%a0] : memref<100xf32>
@@ -228,7 +228,7 @@ func @store_load_affine_apply_symbol_offset(%arg0: index) {
// -----
// CHECK-LABEL: func @store_range_load_after_range() {
func @store_range_load_after_range() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c10 = constant 10 : index
affine.for %i0 = 0 to 10 {
@@ -251,7 +251,7 @@ func @store_range_load_after_range() {
// -----
// CHECK-LABEL: func @store_load_func_symbol(%arg0: index, %arg1: index) {
func @store_load_func_symbol(%arg0: index, %arg1: index) {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c10 = constant 10 : index
affine.for %i0 = 0 to %arg1 {
@@ -274,7 +274,7 @@ func @store_load_func_symbol(%arg0: index, %arg1: index) {
// -----
// CHECK-LABEL: func @store_range_load_last_in_range() {
func @store_range_load_last_in_range() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c10 = constant 10 : index
affine.for %i0 = 0 to 10 {
@@ -302,7 +302,7 @@ func @store_range_load_last_in_range() {
// -----
// CHECK-LABEL: func @store_range_load_before_range() {
func @store_range_load_before_range() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c0 = constant 0 : index
affine.for %i0 = 1 to 11 {
@@ -325,7 +325,7 @@ func @store_range_load_before_range() {
// -----
// CHECK-LABEL: func @store_range_load_first_in_range() {
func @store_range_load_first_in_range() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
%c0 = constant 0 : index
affine.for %i0 = 1 to 11 {
@@ -351,7 +351,7 @@ func @store_range_load_first_in_range() {
// -----
// CHECK-LABEL: func @store_plus_3() {
func @store_plus_3() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 1 to 11 {
%a0 = affine.apply affine_map<(d0) -> (d0 + 3)> (%i0)
@@ -373,7 +373,7 @@ func @store_plus_3() {
// -----
// CHECK-LABEL: func @load_minus_2() {
func @load_minus_2() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 2 to 11 {
%a0 = affine.apply affine_map<(d0) -> (d0)> (%i0)
@@ -395,7 +395,7 @@ func @load_minus_2() {
// -----
// CHECK-LABEL: func @perfectly_nested_loops_loop_independent() {
func @perfectly_nested_loops_loop_independent() {
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 11 {
affine.for %i1 = 0 to 11 {
@@ -426,7 +426,7 @@ func @perfectly_nested_loops_loop_independent() {
// -----
// CHECK-LABEL: func @perfectly_nested_loops_loop_carried_at_depth1() {
func @perfectly_nested_loops_loop_carried_at_depth1() {
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 9 {
affine.for %i1 = 0 to 9 {
@@ -457,7 +457,7 @@ func @perfectly_nested_loops_loop_carried_at_depth1() {
// -----
// CHECK-LABEL: func @perfectly_nested_loops_loop_carried_at_depth2() {
func @perfectly_nested_loops_loop_carried_at_depth2() {
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
@@ -488,7 +488,7 @@ func @perfectly_nested_loops_loop_carried_at_depth2() {
// -----
// CHECK-LABEL: func @one_common_loop() {
func @one_common_loop() {
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
%c7 = constant 7.0 : f32
// There is a loop-independent dependence from access 0 to 1 at depth 2.
affine.for %i0 = 0 to 10 {
@@ -519,8 +519,8 @@ func @one_common_loop() {
// -----
// CHECK-LABEL: func @dependence_cycle() {
func @dependence_cycle() {
- %m.a = alloc() : memref<100xf32>
- %m.b = alloc() : memref<100xf32>
+ %m.a = memref.alloc() : memref<100xf32>
+ %m.b = memref.alloc() : memref<100xf32>
// Dependences:
// *) loop-independent dependence from access 1 to 2 at depth 2.
@@ -573,7 +573,7 @@ func @dependence_cycle() {
// -----
// CHECK-LABEL: func @negative_and_positive_direction_vectors(%arg0: index, %arg1: index) {
func @negative_and_positive_direction_vectors(%arg0: index, %arg1: index) {
- %m = alloc() : memref<10x10xf32>
+ %m = memref.alloc() : memref<10x10xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to %arg0 {
affine.for %i1 = 0 to %arg1 {
@@ -603,7 +603,7 @@ func @negative_and_positive_direction_vectors(%arg0: index, %arg1: index) {
// -----
// CHECK-LABEL: func @war_raw_waw_deps() {
func @war_raw_waw_deps() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
affine.for %i1 = 0 to 10 {
@@ -631,7 +631,7 @@ func @war_raw_waw_deps() {
// -----
// CHECK-LABEL: func @mod_deps() {
func @mod_deps() {
- %m = alloc() : memref<100xf32>
+ %m = memref.alloc() : memref<100xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
%a0 = affine.apply affine_map<(d0) -> (d0 mod 2)> (%i0)
@@ -655,7 +655,7 @@ func @mod_deps() {
// -----
// CHECK-LABEL: func @loop_nest_depth() {
func @loop_nest_depth() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%c7 = constant 7.0 : f32
affine.for %i0 = 0 to 128 {
@@ -691,7 +691,7 @@ func @loop_nest_depth() {
// mod/div's successively.
// CHECK-LABEL: func @mod_div_3d() {
func @mod_div_3d() {
- %M = alloc() : memref<2x2x2xi32>
+ %M = memref.alloc() : memref<2x2x2xi32>
%c0 = constant 0 : i32
affine.for %i0 = 0 to 8 {
affine.for %i1 = 0 to 8 {
@@ -716,8 +716,8 @@ func @mod_div_3d() {
func @delinearize_mod_floordiv() {
%c0 = constant 0 : index
%val = constant 0 : i32
- %in = alloc() : memref<2x2x3x3x16x1xi32>
- %out = alloc() : memref<64x9xi32>
+ %in = memref.alloc() : memref<2x2x3x3x16x1xi32>
+ %out = memref.alloc() : memref<64x9xi32>
affine.for %i0 = 0 to 2 {
affine.for %i1 = 0 to 2 {
@@ -788,7 +788,7 @@ func @delinearize_mod_floordiv() {
// Load and store ops access the same elements in strided scf.
// CHECK-LABEL: func @strided_loop_with_dependence_at_depth2
func @strided_loop_with_dependence_at_depth2() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%cf0 = constant 0.0 : f32
affine.for %i0 = 0 to 8 step 2 {
affine.store %cf0, %0[%i0] : memref<10xf32>
@@ -810,7 +810,7 @@ func @strided_loop_with_dependence_at_depth2() {
// Load and store ops access alternating memref elements: no dependence.
// CHECK-LABEL: func @strided_loop_with_no_dependence
func @strided_loop_with_no_dependence() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%cf0 = constant 0.0 : f32
affine.for %i0 = 0 to 8 step 2 {
%a0 = affine.apply affine_map<(d0) -> (d0 + 1)>(%i0)
@@ -833,7 +833,7 @@ func @strided_loop_with_no_dependence() {
// Affine.Store op accesses memref elements at offset causing loop-carried dependence.
// CHECK-LABEL: func @strided_loop_with_loop_carried_dependence_at_depth1
func @strided_loop_with_loop_carried_dependence_at_depth1() {
- %0 = alloc() : memref<10xf32>
+ %0 = memref.alloc() : memref<10xf32>
%cf0 = constant 0.0 : f32
affine.for %i0 = 0 to 8 step 2 {
%a0 = affine.apply affine_map<(d0) -> (d0 + 4)>(%i0)
@@ -857,7 +857,7 @@ func @strided_loop_with_loop_carried_dependence_at_depth1() {
// properly computed when the load and store are at
diff erent loop depths.
// CHECK-LABEL: func @test_dep_store_depth1_load_depth2
func @test_dep_store_depth1_load_depth2() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 10 {
%a0 = affine.apply affine_map<(d0) -> (d0 - 1)>(%i0)
@@ -884,7 +884,7 @@ func @test_dep_store_depth1_load_depth2() {
// properly computed when the load and store are at
diff erent loop depths.
// CHECK-LABEL: func @test_dep_store_depth2_load_depth1
func @test_dep_store_depth2_load_depth1() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cst = constant 7.000000e+00 : f32
affine.for %i0 = 0 to 10 {
affine.for %i1 = affine_map<(d0) -> (d0)>(%i0) to affine_map<(d0) -> (d0 + 1)>(%i0) {
@@ -912,7 +912,7 @@ func @test_dep_store_depth2_load_depth1() {
// CHECK-LABEL: func @test_affine_for_if_same_block() {
func @test_affine_for_if_same_block() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 100 {
@@ -940,7 +940,7 @@ func @test_affine_for_if_same_block() {
// CHECK-LABEL: func @test_affine_for_if_separated() {
func @test_affine_for_if_separated() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 10 {
@@ -970,7 +970,7 @@ func @test_affine_for_if_separated() {
// CHECK-LABEL: func @test_affine_for_if_partially_joined() {
func @test_affine_for_if_partially_joined() {
- %0 = alloc() : memref<100xf32>
+ %0 = memref.alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 100 {
@@ -1001,7 +1001,7 @@ func @test_affine_for_if_partially_joined() {
// CHECK-LABEL: func @test_interleaved_affine_for_if() {
func @test_interleaved_affine_for_if() {
- %0 = alloc() : memref<100x100xf32>
+ %0 = memref.alloc() : memref<100x100xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 100 {
@@ -1039,9 +1039,9 @@ func @test_interleaved_affine_for_if() {
// CHECK-LABEL: func @test_interleaved_affine_for_if() {
func @test_interleaved_affine_for_if() {
- %0 = alloc() : memref<101xf32>
+ %0 = memref.alloc() : memref<101xf32>
%c0 = constant 0 : index
- %N = dim %0, %c0 : memref<101xf32>
+ %N = memref.dim %0, %c0 : memref<101xf32>
%cf7 = constant 7.0 : f32
affine.for %i0 = 0 to 101 {
diff --git a/mlir/test/Transforms/normalize-memrefs-ops.mlir b/mlir/test/Transforms/normalize-memrefs-ops.mlir
index 84b817aa7f24..01c694c62e1b 100644
--- a/mlir/test/Transforms/normalize-memrefs-ops.mlir
+++ b/mlir/test/Transforms/normalize-memrefs-ops.mlir
@@ -16,13 +16,13 @@
// CHECK-LABEL: test_norm
// CHECK-SAME: (%[[ARG0:.*]]: memref<1x16x1x1x32x64xf32>)
func @test_norm(%arg0 : memref<1x16x14x14xf32, #map0>) -> () {
- %0 = alloc() : memref<1x16x14x14xf32, #map0>
+ %0 = memref.alloc() : memref<1x16x14x14xf32, #map0>
"test.op_norm"(%arg0, %0) : (memref<1x16x14x14xf32, #map0>, memref<1x16x14x14xf32, #map0>) -> ()
- dealloc %0 : memref<1x16x14x14xf32, #map0>
+ memref.dealloc %0 : memref<1x16x14x14xf32, #map0>
- // CHECK: %[[v0:.*]] = alloc() : memref<1x16x1x1x32x64xf32>
+ // CHECK: %[[v0:.*]] = memref.alloc() : memref<1x16x1x1x32x64xf32>
// CHECK: "test.op_norm"(%[[ARG0]], %[[v0]]) : (memref<1x16x1x1x32x64xf32>, memref<1x16x1x1x32x64xf32>) -> ()
- // CHECK: dealloc %[[v0]] : memref<1x16x1x1x32x64xf32>
+ // CHECK: memref.dealloc %[[v0]] : memref<1x16x1x1x32x64xf32>
return
}
@@ -31,13 +31,13 @@ func @test_norm(%arg0 : memref<1x16x14x14xf32, #map0>) -> () {
// CHECK-LABEL: test_nonnorm
// CHECK-SAME: (%[[ARG0:.*]]: memref<1x16x14x14xf32, #map>)
func @test_nonnorm(%arg0 : memref<1x16x14x14xf32, #map0>) -> () {
- %0 = alloc() : memref<1x16x14x14xf32, #map0>
+ %0 = memref.alloc() : memref<1x16x14x14xf32, #map0>
"test.op_nonnorm"(%arg0, %0) : (memref<1x16x14x14xf32, #map0>, memref<1x16x14x14xf32, #map0>) -> ()
- dealloc %0 : memref<1x16x14x14xf32, #map0>
+ memref.dealloc %0 : memref<1x16x14x14xf32, #map0>
- // CHECK: %[[v0:.*]] = alloc() : memref<1x16x14x14xf32, #map>
+ // CHECK: %[[v0:.*]] = memref.alloc() : memref<1x16x14x14xf32, #map>
// CHECK: "test.op_nonnorm"(%[[ARG0]], %[[v0]]) : (memref<1x16x14x14xf32, #map>, memref<1x16x14x14xf32, #map>) -> ()
- // CHECK: dealloc %[[v0]] : memref<1x16x14x14xf32, #map>
+ // CHECK: memref.dealloc %[[v0]] : memref<1x16x14x14xf32, #map>
return
}
@@ -46,13 +46,13 @@ func @test_nonnorm(%arg0 : memref<1x16x14x14xf32, #map0>) -> () {
// CHECK-LABEL: test_norm_mix
// CHECK-SAME: (%[[ARG0:.*]]: memref<1x16x1x1x32x64xf32>
func @test_norm_mix(%arg0 : memref<1x16x1x1x32x64xf32>) -> () {
- %0 = alloc() : memref<1x16x14x14xf32, #map0>
+ %0 = memref.alloc() : memref<1x16x14x14xf32, #map0>
"test.op_norm"(%arg0, %0) : (memref<1x16x1x1x32x64xf32>, memref<1x16x14x14xf32, #map0>) -> ()
- dealloc %0 : memref<1x16x14x14xf32, #map0>
+ memref.dealloc %0 : memref<1x16x14x14xf32, #map0>
- // CHECK: %[[v0:.*]] = alloc() : memref<1x16x1x1x32x64xf32>
+ // CHECK: %[[v0:.*]] = memref.alloc() : memref<1x16x1x1x32x64xf32>
// CHECK: "test.op_norm"(%[[ARG0]], %[[v0]]) : (memref<1x16x1x1x32x64xf32>, memref<1x16x1x1x32x64xf32>) -> ()
- // CHECK: dealloc %[[v0]] : memref<1x16x1x1x32x64xf32>
+ // CHECK: memref.dealloc %[[v0]] : memref<1x16x1x1x32x64xf32>
return
}
@@ -63,10 +63,10 @@ func @test_norm_mix(%arg0 : memref<1x16x1x1x32x64xf32>) -> () {
// CHECK-LABEL: test_load_store
// CHECK-SAME: (%[[ARG0:.*]]: memref<1x16x14x14xf32>
func @test_load_store(%arg0 : memref<1x16x14x14xf32>) -> () {
- %0 = alloc() : memref<1x16x14x14xf32, #map_tile>
- // CHECK: %[[v0:.*]] = alloc() : memref<1x16x1x1x32x32xf32>
- %1 = alloc() : memref<1x16x14x14xf32>
- // CHECK: %[[v1:.*]] = alloc() : memref<1x16x14x14xf32>
+ %0 = memref.alloc() : memref<1x16x14x14xf32, #map_tile>
+ // CHECK: %[[v0:.*]] = memref.alloc() : memref<1x16x1x1x32x32xf32>
+ %1 = memref.alloc() : memref<1x16x14x14xf32>
+ // CHECK: %[[v1:.*]] = memref.alloc() : memref<1x16x14x14xf32>
"test.op_norm"(%0, %1) : (memref<1x16x14x14xf32, #map_tile>, memref<1x16x14x14xf32>) -> ()
// CHECK: "test.op_norm"(%[[v0]], %[[v1]]) : (memref<1x16x1x1x32x32xf32>, memref<1x16x14x14xf32>) -> ()
%cst = constant 3.0 : f32
@@ -74,19 +74,19 @@ func @test_load_store(%arg0 : memref<1x16x14x14xf32>) -> () {
affine.for %j = 0 to 16 {
affine.for %k = 0 to 14 {
affine.for %l = 0 to 14 {
- %2 = load %1[%i, %j, %k, %l] : memref<1x16x14x14xf32>
+ %2 = memref.load %1[%i, %j, %k, %l] : memref<1x16x14x14xf32>
// CHECK: memref<1x16x14x14xf32>
%3 = addf %2, %cst : f32
- store %3, %arg0[%i, %j, %k, %l] : memref<1x16x14x14xf32>
+ memref.store %3, %arg0[%i, %j, %k, %l] : memref<1x16x14x14xf32>
// CHECK: memref<1x16x14x14xf32>
}
}
}
}
- dealloc %0 : memref<1x16x14x14xf32, #map_tile>
- // CHECK: dealloc %[[v0]] : memref<1x16x1x1x32x32xf32>
- dealloc %1 : memref<1x16x14x14xf32>
- // CHECK: dealloc %[[v1]] : memref<1x16x14x14xf32>
+ memref.dealloc %0 : memref<1x16x14x14xf32, #map_tile>
+ // CHECK: memref.dealloc %[[v0]] : memref<1x16x1x1x32x32xf32>
+ memref.dealloc %1 : memref<1x16x14x14xf32>
+ // CHECK: memref.dealloc %[[v1]] : memref<1x16x14x14xf32>
return
}
@@ -95,16 +95,16 @@ func @test_load_store(%arg0 : memref<1x16x14x14xf32>) -> () {
// CHECK-LABEL: test_norm_ret
// CHECK-SAME: (%[[ARG0:.*]]: memref<1x16x1x1x32x32xf32>) -> (memref<1x16x1x1x32x32xf32>, memref<1x16x14x14xf32>) {
func @test_norm_ret(%arg0: memref<1x16x14x14xf32, #map_tile>) -> (memref<1x16x14x14xf32, #map_tile>, memref<1x16x14x14xf32>) {
- %0 = alloc() : memref<1x16x14x14xf32, #map_tile>
- // CHECK-NEXT: %[[v0:.*]] = alloc() : memref<1x16x1x1x32x32xf32>
+ %0 = memref.alloc() : memref<1x16x14x14xf32, #map_tile>
+ // CHECK-NEXT: %[[v0:.*]] = memref.alloc() : memref<1x16x1x1x32x32xf32>
%1, %2 = "test.op_norm_ret"(%arg0) : (memref<1x16x14x14xf32, #map_tile>) -> (memref<1x16x14x14xf32, #map_tile>, memref<1x16x14x14xf32>)
// CHECK-NEXT: %[[v1:.*]], %[[v2:.*]] = "test.op_norm_ret"
// CHECK-SAME: (memref<1x16x1x1x32x32xf32>) -> (memref<1x16x1x1x32x32xf32>, memref<1x16x14x14xf32>)
"test.op_norm"(%1, %0) : (memref<1x16x14x14xf32, #map_tile>, memref<1x16x14x14xf32, #map_tile>) -> ()
// CHECK-NEXT: "test.op_norm"
// CHECK-SAME: : (memref<1x16x1x1x32x32xf32>, memref<1x16x1x1x32x32xf32>) -> ()
- dealloc %0 : memref<1x16x14x14xf32, #map_tile>
- // CHECK-NEXT: dealloc %[[v0]] : memref<1x16x1x1x32x32xf32>
+ memref.dealloc %0 : memref<1x16x14x14xf32, #map_tile>
+ // CHECK-NEXT: memref.dealloc %[[v0]] : memref<1x16x1x1x32x32xf32>
return %1, %2 : memref<1x16x14x14xf32, #map_tile>, memref<1x16x14x14xf32>
// CHECK-NEXT: return %[[v1]], %[[v2]] : memref<1x16x1x1x32x32xf32>, memref<1x16x14x14xf32>
}
diff --git a/mlir/test/Transforms/normalize-memrefs.mlir b/mlir/test/Transforms/normalize-memrefs.mlir
index c186423b9ba2..086cff3c931f 100644
--- a/mlir/test/Transforms/normalize-memrefs.mlir
+++ b/mlir/test/Transforms/normalize-memrefs.mlir
@@ -5,32 +5,32 @@
// CHECK-LABEL: func @permute()
func @permute() {
- %A = alloc() : memref<64x256xf32, affine_map<(d0, d1) -> (d1, d0)>>
+ %A = memref.alloc() : memref<64x256xf32, affine_map<(d0, d1) -> (d1, d0)>>
affine.for %i = 0 to 64 {
affine.for %j = 0 to 256 {
%1 = affine.load %A[%i, %j] : memref<64x256xf32, affine_map<(d0, d1) -> (d1, d0)>>
"prevent.dce"(%1) : (f32) -> ()
}
}
- dealloc %A : memref<64x256xf32, affine_map<(d0, d1) -> (d1, d0)>>
+ memref.dealloc %A : memref<64x256xf32, affine_map<(d0, d1) -> (d1, d0)>>
return
}
// The old memref alloc should disappear.
// CHECK-NOT: memref<64x256xf32>
-// CHECK: [[MEM:%[0-9]+]] = alloc() : memref<256x64xf32>
+// CHECK: [[MEM:%[0-9]+]] = memref.alloc() : memref<256x64xf32>
// CHECK-NEXT: affine.for %[[I:arg[0-9]+]] = 0 to 64 {
// CHECK-NEXT: affine.for %[[J:arg[0-9]+]] = 0 to 256 {
// CHECK-NEXT: affine.load [[MEM]][%[[J]], %[[I]]] : memref<256x64xf32>
// CHECK-NEXT: "prevent.dce"
// CHECK-NEXT: }
// CHECK-NEXT: }
-// CHECK-NEXT: dealloc [[MEM]]
+// CHECK-NEXT: memref.dealloc [[MEM]]
// CHECK-NEXT: return
// CHECK-LABEL: func @shift
func @shift(%idx : index) {
- // CHECK-NEXT: alloc() : memref<65xf32>
- %A = alloc() : memref<64xf32, affine_map<(d0) -> (d0 + 1)>>
+ // CHECK-NEXT: memref.alloc() : memref<65xf32>
+ %A = memref.alloc() : memref<64xf32, affine_map<(d0) -> (d0 + 1)>>
// CHECK-NEXT: affine.load %{{.*}}[symbol(%arg0) + 1] : memref<65xf32>
affine.load %A[%idx] : memref<64xf32, affine_map<(d0) -> (d0 + 1)>>
affine.for %i = 0 to 64 {
@@ -44,7 +44,7 @@ func @shift(%idx : index) {
// CHECK-LABEL: func @high_dim_permute()
func @high_dim_permute() {
// CHECK-NOT: memref<64x128x256xf32,
- %A = alloc() : memref<64x128x256xf32, affine_map<(d0, d1, d2) -> (d2, d0, d1)>>
+ %A = memref.alloc() : memref<64x128x256xf32, affine_map<(d0, d1, d2) -> (d2, d0, d1)>>
// CHECK: %[[I:arg[0-9]+]]
affine.for %i = 0 to 64 {
// CHECK: %[[J:arg[0-9]+]]
@@ -62,16 +62,16 @@ func @high_dim_permute() {
// CHECK-LABEL: func @invalid_map
func @invalid_map() {
- %A = alloc() : memref<64x128xf32, affine_map<(d0, d1) -> (d0, -d1 - 10)>>
- // CHECK: %{{.*}} = alloc() : memref<64x128xf32,
+ %A = memref.alloc() : memref<64x128xf32, affine_map<(d0, d1) -> (d0, -d1 - 10)>>
+ // CHECK: %{{.*}} = memref.alloc() : memref<64x128xf32,
return
}
// A tiled layout.
// CHECK-LABEL: func @data_tiling
func @data_tiling(%idx : index) {
- // CHECK: alloc() : memref<8x32x8x16xf32>
- %A = alloc() : memref<64x512xf32, affine_map<(d0, d1) -> (d0 floordiv 8, d1 floordiv 16, d0 mod 8, d1 mod 16)>>
+ // CHECK: memref.alloc() : memref<8x32x8x16xf32>
+ %A = memref.alloc() : memref<64x512xf32, affine_map<(d0, d1) -> (d0 floordiv 8, d1 floordiv 16, d0 mod 8, d1 mod 16)>>
// CHECK: affine.load %{{.*}}[symbol(%arg0) floordiv 8, symbol(%arg0) floordiv 16, symbol(%arg0) mod 8, symbol(%arg0) mod 16]
%1 = affine.load %A[%idx, %idx] : memref<64x512xf32, affine_map<(d0, d1) -> (d0 floordiv 8, d1 floordiv 16, d0 mod 8, d1 mod 16)>>
"prevent.dce"(%1) : (f32) -> ()
@@ -81,7 +81,7 @@ func @data_tiling(%idx : index) {
// Strides 2 and 4 along respective dimensions.
// CHECK-LABEL: func @strided
func @strided() {
- %A = alloc() : memref<64x128xf32, affine_map<(d0, d1) -> (2*d0, 4*d1)>>
+ %A = memref.alloc() : memref<64x128xf32, affine_map<(d0, d1) -> (2*d0, 4*d1)>>
// CHECK: affine.for %[[IV0:.*]] =
affine.for %i = 0 to 64 {
// CHECK: affine.for %[[IV1:.*]] =
@@ -97,7 +97,7 @@ func @strided() {
// Strided, but the strides are in the linearized space.
// CHECK-LABEL: func @strided_cumulative
func @strided_cumulative() {
- %A = alloc() : memref<2x5xf32, affine_map<(d0, d1) -> (3*d0 + 17*d1)>>
+ %A = memref.alloc() : memref<2x5xf32, affine_map<(d0, d1) -> (3*d0 + 17*d1)>>
// CHECK: affine.for %[[IV0:.*]] =
affine.for %i = 0 to 2 {
// CHECK: affine.for %[[IV1:.*]] =
@@ -114,8 +114,8 @@ func @strided_cumulative() {
// when the index remap has symbols.
// CHECK-LABEL: func @symbolic_operands
func @symbolic_operands(%s : index) {
- // CHECK: alloc() : memref<100xf32>
- %A = alloc()[%s] : memref<10x10xf32, affine_map<(d0,d1)[s0] -> (10*d0 + d1)>>
+ // CHECK: memref.alloc() : memref<100xf32>
+ %A = memref.alloc()[%s] : memref<10x10xf32, affine_map<(d0,d1)[s0] -> (10*d0 + d1)>>
affine.for %i = 0 to 10 {
affine.for %j = 0 to 10 {
// CHECK: affine.load %{{.*}}[%{{.*}} * 10 + %{{.*}}] : memref<100xf32>
@@ -129,7 +129,7 @@ func @symbolic_operands(%s : index) {
// Semi-affine maps, normalization not implemented yet.
// CHECK-LABEL: func @semi_affine_layout_map
func @semi_affine_layout_map(%s0: index, %s1: index) {
- %A = alloc()[%s0, %s1] : memref<256x1024xf32, affine_map<(d0, d1)[s0, s1] -> (d0*s0 + d1*s1)>>
+ %A = memref.alloc()[%s0, %s1] : memref<256x1024xf32, affine_map<(d0, d1)[s0, s1] -> (d0*s0 + d1*s1)>>
affine.for %i = 0 to 256 {
affine.for %j = 0 to 1024 {
// CHECK: memref<256x1024xf32, #map{{[0-9]+}}>
@@ -141,8 +141,8 @@ func @semi_affine_layout_map(%s0: index, %s1: index) {
// CHECK-LABEL: func @alignment
func @alignment() {
- %A = alloc() {alignment = 32 : i64}: memref<64x128x256xf32, affine_map<(d0, d1, d2) -> (d2, d0, d1)>>
- // CHECK-NEXT: alloc() {alignment = 32 : i64} : memref<256x64x128xf32>
+ %A = memref.alloc() {alignment = 32 : i64}: memref<64x128x256xf32, affine_map<(d0, d1, d2) -> (d2, d0, d1)>>
+ // CHECK-NEXT: memref.alloc() {alignment = 32 : i64} : memref<256x64x128xf32>
return
}
@@ -171,20 +171,20 @@ func @multiple_argument_type(%A: memref<16xf64, #tile>, %B: f64, %C: memref<8xf6
// CHECK-LABEL: func @single_argument_type
// CHECK-SAME: (%[[C:arg[0-9]+]]: memref<2x4xf64>)
func @single_argument_type(%C : memref<8xf64, #tile>) {
- %a = alloc(): memref<8xf64, #tile>
- %b = alloc(): memref<16xf64, #tile>
+ %a = memref.alloc(): memref<8xf64, #tile>
+ %b = memref.alloc(): memref<16xf64, #tile>
%d = constant 23.0 : f64
- %e = alloc(): memref<24xf64>
+ %e = memref.alloc(): memref<24xf64>
call @single_argument_type(%a): (memref<8xf64, #tile>) -> ()
call @single_argument_type(%C): (memref<8xf64, #tile>) -> ()
call @multiple_argument_type(%b, %d, %a, %e): (memref<16xf64, #tile>, f64, memref<8xf64, #tile>, memref<24xf64>) -> f64
return
}
-// CHECK: %[[a:[0-9]+]] = alloc() : memref<2x4xf64>
-// CHECK: %[[b:[0-9]+]] = alloc() : memref<4x4xf64>
+// CHECK: %[[a:[0-9]+]] = memref.alloc() : memref<2x4xf64>
+// CHECK: %[[b:[0-9]+]] = memref.alloc() : memref<4x4xf64>
// CHECK: %cst = constant 2.300000e+01 : f64
-// CHECK: %[[e:[0-9]+]] = alloc() : memref<24xf64>
+// CHECK: %[[e:[0-9]+]] = memref.alloc() : memref<24xf64>
// CHECK: call @single_argument_type(%[[a]]) : (memref<2x4xf64>) -> ()
// CHECK: call @single_argument_type(%[[C]]) : (memref<2x4xf64>) -> ()
// CHECK: call @multiple_argument_type(%[[b]], %cst, %[[a]], %[[e]]) : (memref<4x4xf64>, f64, memref<2x4xf64>, memref<24xf64>) -> f64
@@ -227,8 +227,8 @@ func @ret_multiple_argument_type(%A: memref<16xf64, #tile>, %B: f64, %C: memref<
// CHECK-LABEL: func @ret_single_argument_type
// CHECK-SAME: (%[[C:arg[0-9]+]]: memref<2x4xf64>) -> (memref<4x4xf64>, memref<2x4xf64>)
func @ret_single_argument_type(%C: memref<8xf64, #tile>) -> (memref<16xf64, #tile>, memref<8xf64, #tile>){
- %a = alloc() : memref<8xf64, #tile>
- %b = alloc() : memref<16xf64, #tile>
+ %a = memref.alloc() : memref<8xf64, #tile>
+ %b = memref.alloc() : memref<16xf64, #tile>
%d = constant 23.0 : f64
call @ret_single_argument_type(%a) : (memref<8xf64, #tile>) -> (memref<16xf64, #tile>, memref<8xf64, #tile>)
call @ret_single_argument_type(%C) : (memref<8xf64, #tile>) -> (memref<16xf64, #tile>, memref<8xf64, #tile>)
@@ -237,8 +237,8 @@ func @ret_single_argument_type(%C: memref<8xf64, #tile>) -> (memref<16xf64, #til
return %b, %a: memref<16xf64, #tile>, memref<8xf64, #tile>
}
-// CHECK: %[[a:[0-9]+]] = alloc() : memref<2x4xf64>
-// CHECK: %[[b:[0-9]+]] = alloc() : memref<4x4xf64>
+// CHECK: %[[a:[0-9]+]] = memref.alloc() : memref<2x4xf64>
+// CHECK: %[[b:[0-9]+]] = memref.alloc() : memref<4x4xf64>
// CHECK: %cst = constant 2.300000e+01 : f64
// CHECK: %[[resA:[0-9]+]]:2 = call @ret_single_argument_type(%[[a]]) : (memref<2x4xf64>) -> (memref<4x4xf64>, memref<2x4xf64>)
// CHECK: %[[resB:[0-9]+]]:2 = call @ret_single_argument_type(%[[C]]) : (memref<2x4xf64>) -> (memref<4x4xf64>, memref<2x4xf64>)
@@ -304,11 +304,11 @@ func private @external_func_B(memref<16xf64, #tile>, f64) -> (memref<8xf64, #til
// CHECK-LABEL: func @simply_call_external()
func @simply_call_external() {
- %a = alloc() : memref<16xf64, #tile>
+ %a = memref.alloc() : memref<16xf64, #tile>
call @external_func_A(%a) : (memref<16xf64, #tile>) -> ()
return
}
-// CHECK: %[[a:[0-9]+]] = alloc() : memref<4x4xf64>
+// CHECK: %[[a:[0-9]+]] = memref.alloc() : memref<4x4xf64>
// CHECK: call @external_func_A(%[[a]]) : (memref<4x4xf64>) -> ()
// CHECK-LABEL: func @use_value_of_external
@@ -323,7 +323,7 @@ func @use_value_of_external(%A: memref<16xf64, #tile>, %B: f64) -> (memref<8xf64
// CHECK-LABEL: func @affine_parallel_norm
func @affine_parallel_norm() -> memref<8xf32, #tile> {
%c = constant 23.0 : f32
- %a = alloc() : memref<8xf32, #tile>
+ %a = memref.alloc() : memref<8xf32, #tile>
// CHECK: affine.parallel (%{{.*}}) = (0) to (8) reduce ("assign") -> (memref<2x4xf32>)
%1 = affine.parallel (%i) = (0) to (8) reduce ("assign") -> memref<8xf32, #tile> {
affine.store %c, %a[%i] : memref<8xf32, #tile>
diff --git a/mlir/test/Transforms/parametric-tiling.mlir b/mlir/test/Transforms/parametric-tiling.mlir
index 1022a740ee22..bae657c3537e 100644
--- a/mlir/test/Transforms/parametric-tiling.mlir
+++ b/mlir/test/Transforms/parametric-tiling.mlir
@@ -57,9 +57,9 @@ func @rectangular(%arg0: memref<?x?xf32>) {
// TILE_74:scf.for %[[jj:.*]] = %[[j]] to %[[ub2]] step %c2
scf.for %j = %c1 to %c44 step %c2 {
// The right iterator are used.
- // TILE_7: load %arg0[%[[ii]], %[[j]]]
- // TILE_74: load %arg0[%[[ii]], %[[jj]]]
- load %arg0[%i, %j]: memref<?x?xf32>
+ // TILE_7: memref.load %arg0[%[[ii]], %[[j]]]
+ // TILE_74: memref.load %arg0[%[[ii]], %[[jj]]]
+ memref.load %arg0[%i, %j]: memref<?x?xf32>
}
}
return
@@ -124,9 +124,9 @@ func @triangular(%arg0: memref<?x?xf32>) {
// TILE_74:scf.for %[[jj:.*]] = %[[j]] to %[[ub2]] step %c2
scf.for %j = %c1 to %i step %c2 {
// The right iterator are used.
- // TILE_7: load %arg0[%[[ii]], %[[j]]]
- // TILE_74: load %arg0[%[[ii]], %[[jj]]]
- load %arg0[%i, %j]: memref<?x?xf32>
+ // TILE_7: memref.load %arg0[%[[ii]], %[[j]]]
+ // TILE_74: memref.load %arg0[%[[ii]], %[[jj]]]
+ memref.load %arg0[%i, %j]: memref<?x?xf32>
}
}
return
diff --git a/mlir/test/Transforms/pipeline-data-transfer.mlir b/mlir/test/Transforms/pipeline-data-transfer.mlir
index 80256a17d0a0..8ee0e5c4dbb2 100644
--- a/mlir/test/Transforms/pipeline-data-transfer.mlir
+++ b/mlir/test/Transforms/pipeline-data-transfer.mlir
@@ -8,10 +8,10 @@
// CHECK-LABEL: func @loop_nest_dma() {
func @loop_nest_dma() {
- %A = alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
- %Ah = alloc() : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
+ %A = memref.alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
+ %Ah = memref.alloc() : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
- %tag = alloc() : memref<1 x f32>
+ %tag = memref.alloc() : memref<1 x f32>
%zero = constant 0 : index
%num_elts = constant 32 : index
@@ -26,13 +26,13 @@ func @loop_nest_dma() {
"do_more_compute"(%i, %j) : (index, index) -> ()
}
}
- dealloc %tag : memref<1 x f32>
- dealloc %Ah : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
+ memref.dealloc %tag : memref<1 x f32>
+ memref.dealloc %Ah : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
return
}
-// CHECK: %{{.*}} = alloc() : memref<256xf32>
-// CHECK: %{{.*}} = alloc() : memref<2x32xf32, 1>
-// CHECK-NEXT: %{{.*}} = alloc() : memref<2x1xf32>
+// CHECK: %{{.*}} = memref.alloc() : memref<256xf32>
+// CHECK: %{{.*}} = memref.alloc() : memref<2x32xf32, 1>
+// CHECK-NEXT: %{{.*}} = memref.alloc() : memref<2x1xf32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}} mod 2, %{{.*}}], %{{.*}}[%{{.*}} mod 2, 0], %{{.*}} : memref<256xf32>, memref<2x32xf32, 1>, memref<2x1xf32>
// CHECK-NEXT: affine.for %{{.*}} = 1 to 8 {
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[%{{.*}} mod 2, %{{.*}}], %{{.*}}[%{{.*}} mod 2, 0], %{{.*}} : memref<256xf32>, memref<2x32xf32, 1>, memref<2x1xf32>
@@ -57,8 +57,8 @@ func @loop_nest_dma() {
// CHECK-NEXT: affine.for %{{.*}} = 0 to 32 {
// CHECK-NEXT: "do_more_compute"(%{{.*}}, %{{.*}}) : (index, index) -> ()
// CHECK-NEXT: }
-// CHECK-NEXT: dealloc %{{.*}} : memref<2x1xf32>
-// CHECK-NEXT: dealloc %{{.*}} : memref<2x32xf32, 1>
+// CHECK-NEXT: memref.dealloc %{{.*}} : memref<2x1xf32>
+// CHECK-NEXT: memref.dealloc %{{.*}} : memref<2x32xf32, 1>
// CHECK-NEXT: return
// CHECK-NEXT:}
@@ -73,19 +73,19 @@ func @loop_step(%arg0: memref<512xf32>,
%c0 = constant 0 : index
%c4 = constant 4 : index
affine.for %i0 = 0 to 512 step 4 {
- %1 = alloc() : memref<4xf32, 1>
- %2 = alloc() : memref<1xi32>
+ %1 = memref.alloc() : memref<4xf32, 1>
+ %2 = memref.alloc() : memref<1xi32>
affine.dma_start %arg0[%i0], %1[%c0], %2[%c0], %c4,
: memref<512xf32>, memref<4xf32, 1>, memref<1xi32>
affine.dma_wait %2[%c0], %c4 : memref<1xi32>
"compute"(%i0) : (index) -> ()
- dealloc %2 : memref<1xi32>
- dealloc %1 : memref<4xf32, 1>
+ memref.dealloc %2 : memref<1xi32>
+ memref.dealloc %1 : memref<4xf32, 1>
}
return
}
-// CHECK: [[BUF:%[0-9]+]] = alloc() : memref<2x4xf32, 1>
-// CHECK: [[TAG:%[0-9]+]] = alloc() : memref<2x1xi32>
+// CHECK: [[BUF:%[0-9]+]] = memref.alloc() : memref<2x4xf32, 1>
+// CHECK: [[TAG:%[0-9]+]] = memref.alloc() : memref<2x1xi32>
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[(%{{.*}} floordiv 4) mod 2, 0], [[TAG]][(%{{.*}} floordiv 4) mod 2, 0], %{{.*}} : memref<512xf32>, memref<2x4xf32, 1>, memref<2x1xi32>
// CHECK-NEXT: affine.for %{{.*}} = 4 to 512 step 4 {
// CHECK-NEXT: affine.dma_start %{{.*}}[%{{.*}}], %{{.*}}[(%{{.*}} floordiv 4) mod 2, 0], [[TAG]][(%{{.*}} floordiv 4) mod 2, 0], %{{.*}} : memref<512xf32>, memref<2x4xf32, 1>, memref<2x1xi32>
@@ -98,8 +98,8 @@ func @loop_step(%arg0: memref<512xf32>,
// CHECK-NEXT: %{{.*}} = affine.apply [[$FLOOR_MOD_2]]([[SHIFTED]])
// CHECK: affine.dma_wait [[TAG]][(%{{.*}} floordiv 4) mod 2, 0], %{{.*}} : memref<2x1xi32>
// CHECK-NEXT: "compute"(%{{.*}}) : (index) -> ()
-// CHECK-NEXT: dealloc [[TAG]] : memref<2x1xi32>
-// CHECK-NEXT: dealloc [[BUF]] : memref<2x4xf32, 1>
+// CHECK-NEXT: memref.dealloc [[TAG]] : memref<2x1xi32>
+// CHECK-NEXT: memref.dealloc [[BUF]] : memref<2x4xf32, 1>
// CHECK-NEXT: return
// CHECK-NEXT: }
@@ -111,15 +111,15 @@ func @loop_step(%arg0: memref<512xf32>,
func @loop_dma_nested(%arg0: memref<512x32xvector<8xf32>>, %arg1: memref<512x32xvector<8xf32>>, %arg2: memref<512x32xvector<8xf32>>) {
%num_elts = constant 256 : index
%c0 = constant 0 : index
- %0 = alloc() : memref<64x4xvector<8xf32>, 2>
- %1 = alloc() : memref<64x4xvector<8xf32>, 2>
- %2 = alloc() : memref<64x4xvector<8xf32>, 2>
- %3 = alloc() : memref<2xi32>
- %4 = alloc() : memref<2xi32>
- %5 = alloc() : memref<2xi32>
+ %0 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %1 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %2 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %3 = memref.alloc() : memref<2xi32>
+ %4 = memref.alloc() : memref<2xi32>
+ %5 = memref.alloc() : memref<2xi32>
// Prologue for DMA overlap on arg2.
- // CHECK-DAG: [[BUF_ARG2:%[0-9]+]] = alloc() : memref<2x64x4xvector<8xf32>, 2>
- // CHECK-DAG: [[TAG_ARG2:%[0-9]+]] = alloc() : memref<2x2xi32>
+ // CHECK-DAG: [[BUF_ARG2:%[0-9]+]] = memref.alloc() : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK-DAG: [[TAG_ARG2:%[0-9]+]] = memref.alloc() : memref<2x2xi32>
// CHECK: affine.dma_start %{{.*}}[
// CHECK: affine.for %{{.*}} = 1 to 8 {
affine.for %i0 = 0 to 8 {
@@ -130,10 +130,10 @@ func @loop_dma_nested(%arg0: memref<512x32xvector<8xf32>>, %arg1: memref<512x32x
// CHECK: affine.dma_start %{{.*}}[
// CHECK: affine.dma_wait [[TAG_ARG2]]
// Prologue for DMA overlap on arg0, arg1 nested within i0
- // CHECK: [[BUF_ARG0:%[0-9]+]] = alloc() : memref<2x64x4xvector<8xf32>, 2>
- // CHECK: [[BUF_ARG1:%[0-9]+]] = alloc() : memref<2x64x4xvector<8xf32>, 2>
- // CHECK: [[TAG_ARG0:%[0-9]+]] = alloc() : memref<2x2xi32>
- // CHECK: [[TAG_ARG1:%[0-9]+]] = alloc() : memref<2x2xi32>
+ // CHECK: [[BUF_ARG0:%[0-9]+]] = memref.alloc() : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK: [[BUF_ARG1:%[0-9]+]] = memref.alloc() : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK: [[TAG_ARG0:%[0-9]+]] = memref.alloc() : memref<2x2xi32>
+ // CHECK: [[TAG_ARG1:%[0-9]+]] = memref.alloc() : memref<2x2xi32>
// CHECK: affine.dma_start %{{.*}}[
// CHECK: affine.dma_start %{{.*}}[
// CHECK-NEXT: affine.for %{{.*}} = 1 to 8 {
@@ -157,17 +157,17 @@ func @loop_dma_nested(%arg0: memref<512x32xvector<8xf32>>, %arg1: memref<512x32x
// epilogue for arg0, arg1
// CHECK: affine.dma_wait [[TAG_ARG0]]
// CHECK: affine.dma_wait [[TAG_ARG1]]
- // CHECK-DAG: dealloc [[TAG_ARG1]] : memref<2x2xi32>
- // CHECK-DAG: dealloc [[TAG_ARG0]] : memref<2x2xi32>
- // CHECK-DAG: dealloc [[BUF_ARG1]] : memref<2x64x4xvector<8xf32>, 2>
- // CHECK-DAG: dealloc [[BUF_ARG0]] : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK-DAG: memref.dealloc [[TAG_ARG1]] : memref<2x2xi32>
+ // CHECK-DAG: memref.dealloc [[TAG_ARG0]] : memref<2x2xi32>
+ // CHECK-DAG: memref.dealloc [[BUF_ARG1]] : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK-DAG: memref.dealloc [[BUF_ARG0]] : memref<2x64x4xvector<8xf32>, 2>
// epilogue for DMA overlap on %arg2
// CHECK: affine.dma_wait [[TAG_ARG2]]
// Within the epilogue for arg2's DMA, we have the DMAs on %arg1, %arg2 nested.
- // CHECK: [[BUF_ARG0_NESTED:%[0-9]+]] = alloc() : memref<2x64x4xvector<8xf32>, 2>
- // CHECK: [[BUF_ARG1_NESTED:%[0-9]+]] = alloc() : memref<2x64x4xvector<8xf32>, 2>
- // CHECK: [[TAG_ARG0_NESTED:%[0-9]+]] = alloc() : memref<2x2xi32>
- // CHECK: [[TAG_ARG1_NESTED:%[0-9]+]] = alloc() : memref<2x2xi32>
+ // CHECK: [[BUF_ARG0_NESTED:%[0-9]+]] = memref.alloc() : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK: [[BUF_ARG1_NESTED:%[0-9]+]] = memref.alloc() : memref<2x64x4xvector<8xf32>, 2>
+ // CHECK: [[TAG_ARG0_NESTED:%[0-9]+]] = memref.alloc() : memref<2x2xi32>
+ // CHECK: [[TAG_ARG1_NESTED:%[0-9]+]] = memref.alloc() : memref<2x2xi32>
// CHECK: affine.dma_start %{{.*}}[
// CHECK: affine.dma_start %{{.*}}[
// CHECK: affine.for %{{.*}} = 1 to 8 {
@@ -181,20 +181,20 @@ func @loop_dma_nested(%arg0: memref<512x32xvector<8xf32>>, %arg1: memref<512x32x
// CHECK: affine.dma_wait [[TAG_ARG1_NESTED]]
// CHECK: affine.for %{{.*}} = 0 to 4 {
}
- dealloc %5 : memref<2xi32>
- dealloc %4 : memref<2xi32>
- dealloc %3 : memref<2xi32>
- dealloc %2 : memref<64x4xvector<8xf32>, 2>
- dealloc %1 : memref<64x4xvector<8xf32>, 2>
- dealloc %0 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %5 : memref<2xi32>
+ memref.dealloc %4 : memref<2xi32>
+ memref.dealloc %3 : memref<2xi32>
+ memref.dealloc %2 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %1 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %0 : memref<64x4xvector<8xf32>, 2>
return
// CHECK: }
-// CHECK-DAG: dealloc [[TAG_ARG1_NESTED]] : memref<2x2xi32>
-// CHECK-DAG: dealloc [[TAG_ARG0_NESTED]] : memref<2x2xi32>
-// CHECK-DAG: dealloc [[BUF_ARG1_NESTED]] : memref<2x64x4xvector<8xf32>, 2>
-// CHECK-DAG: dealloc [[BUF_ARG0_NESTED]] : memref<2x64x4xvector<8xf32>, 2>
-// CHECK-DAG: dealloc [[TAG_ARG2]] : memref<2x2xi32>
-// CHECK-DAG: dealloc [[BUF_ARG2]] : memref<2x64x4xvector<8xf32>, 2>
+// CHECK-DAG: memref.dealloc [[TAG_ARG1_NESTED]] : memref<2x2xi32>
+// CHECK-DAG: memref.dealloc [[TAG_ARG0_NESTED]] : memref<2x2xi32>
+// CHECK-DAG: memref.dealloc [[BUF_ARG1_NESTED]] : memref<2x64x4xvector<8xf32>, 2>
+// CHECK-DAG: memref.dealloc [[BUF_ARG0_NESTED]] : memref<2x64x4xvector<8xf32>, 2>
+// CHECK-DAG: memref.dealloc [[TAG_ARG2]] : memref<2x2xi32>
+// CHECK-DAG: memref.dealloc [[BUF_ARG2]] : memref<2x64x4xvector<8xf32>, 2>
// CHECK-NEXT: return
}
@@ -205,12 +205,12 @@ func @loop_dma_nested(%arg0: memref<512x32xvector<8xf32>>, %arg1: memref<512x32x
func @loop_dma_dependent(%arg2: memref<512x32xvector<8xf32>>) {
%num_elts = constant 256 : index
%c0 = constant 0 : index
- %0 = alloc() : memref<64x4xvector<8xf32>, 2>
- %1 = alloc() : memref<64x4xvector<8xf32>, 2>
- %2 = alloc() : memref<64x4xvector<8xf32>, 2>
- %3 = alloc() : memref<2xi32>
- %4 = alloc() : memref<2xi32>
- %5 = alloc() : memref<2xi32>
+ %0 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %1 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %2 = memref.alloc() : memref<64x4xvector<8xf32>, 2>
+ %3 = memref.alloc() : memref<2xi32>
+ %4 = memref.alloc() : memref<2xi32>
+ %5 = memref.alloc() : memref<2xi32>
// The two DMAs below are dependent (incoming and outgoing on the same
// memref) in the same iteration; so no pipelining here.
@@ -224,12 +224,12 @@ func @loop_dma_dependent(%arg2: memref<512x32xvector<8xf32>>) {
affine.dma_start %2[%c0, %c0], %arg2[%6, %c0], %5[%c0], %num_elts : memref<64x4xvector<8xf32>, 2>, memref<512x32xvector<8xf32>>, memref<2xi32>
affine.dma_wait %5[%c0], %num_elts : memref<2xi32>
}
- dealloc %5 : memref<2xi32>
- dealloc %4 : memref<2xi32>
- dealloc %3 : memref<2xi32>
- dealloc %2 : memref<64x4xvector<8xf32>, 2>
- dealloc %1 : memref<64x4xvector<8xf32>, 2>
- dealloc %0 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %5 : memref<2xi32>
+ memref.dealloc %4 : memref<2xi32>
+ memref.dealloc %3 : memref<2xi32>
+ memref.dealloc %2 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %1 : memref<64x4xvector<8xf32>, 2>
+ memref.dealloc %0 : memref<64x4xvector<8xf32>, 2>
return
}
@@ -240,8 +240,8 @@ func @escaping_use(%arg0: memref<512 x 32 x f32>) {
%c32 = constant 32 : index
%num_elt = constant 512 : index
%zero = constant 0 : index
- %Av = alloc() : memref<32 x 32 x f32, 2>
- %tag = alloc() : memref<1 x i32>
+ %Av = memref.alloc() : memref<32 x 32 x f32, 2>
+ %tag = memref.alloc() : memref<1 x i32>
// CHECK-NOT: affine.dma_start
// CHECK: affine.for %{{.*}} = 0 to 16 {
@@ -253,8 +253,8 @@ func @escaping_use(%arg0: memref<512 x 32 x f32>) {
// escaping use; no DMA pipelining / double buffering will be done.
"foo"(%Av) : (memref<32 x 32 x f32, 2>) -> ()
}
- dealloc %tag : memref<1 x i32>
- dealloc %Av : memref<32 x 32 x f32, 2>
+ memref.dealloc %tag : memref<1 x i32>
+ memref.dealloc %Av : memref<32 x 32 x f32, 2>
return
// CHECK: "foo"(%{{[0-9]+}}) : (memref<32x32xf32, 2>) -> ()
// CHECK: }
@@ -268,8 +268,8 @@ func @escaping_tag(%arg0: memref<512 x 32 x f32>) {
%c32 = constant 32 : index
%num_elt = constant 512 : index
%zero = constant 0 : index
- %Av = alloc() : memref<32 x 32 x f32, 2>
- %tag = alloc() : memref<1 x i32>
+ %Av = memref.alloc() : memref<32 x 32 x f32, 2>
+ %tag = memref.alloc() : memref<1 x i32>
// CHECK-NOT: affine.dma_start
// CHECK: affine.for %{{.*}} = 0 to 16 {
@@ -281,8 +281,8 @@ func @escaping_tag(%arg0: memref<512 x 32 x f32>) {
// escaping use; no DMA pipelining / double buffering will be done.
"foo"(%tag) : (memref<1 x i32>) -> ()
}
- dealloc %tag : memref<1 x i32>
- dealloc %Av : memref<32 x 32 x f32, 2>
+ memref.dealloc %tag : memref<1 x i32>
+ memref.dealloc %Av : memref<32 x 32 x f32, 2>
return
// CHECK: "foo"(%{{[0-9]+}}) : (memref<1xi32>) -> ()
// CHECK: }
@@ -297,8 +297,8 @@ func @live_out_use(%arg0: memref<512 x 32 x f32>) -> f32 {
%c32 = constant 32 : index
%num_elt = constant 512 : index
%zero = constant 0 : index
- %Av = alloc() : memref<32 x 32 x f32, 2>
- %tag = alloc() : memref<1 x i32>
+ %Av = memref.alloc() : memref<32 x 32 x f32, 2>
+ %tag = memref.alloc() : memref<1 x i32>
// CHECK-NOT: affine.dma_start
// CHECK: affine.for %{{.*}} = 0 to 16 {
@@ -310,8 +310,8 @@ func @live_out_use(%arg0: memref<512 x 32 x f32>) -> f32 {
}
// Use live out of 'affine.for' op; no DMA pipelining will be done.
%v = affine.load %Av[%zero, %zero] : memref<32 x 32 x f32, 2>
- dealloc %tag : memref<1 x i32>
- dealloc %Av : memref<32 x 32 x f32, 2>
+ memref.dealloc %tag : memref<1 x i32>
+ memref.dealloc %Av : memref<32 x 32 x f32, 2>
return %v : f32
// CHECK: affine.load %{{[0-9]+}}[%{{.*}}, %{{.*}}] : memref<32x32xf32, 2>
// CHECK: return
@@ -325,16 +325,16 @@ func @dynamic_shape_dma_buffer(%arg0: memref<512 x 32 x f32>) {
%num_elt = constant 512 : index
%zero = constant 0 : index
- %Av = alloc(%c32, %c32) : memref<? x ? x f32, 2>
- %tag = alloc() : memref<1 x i32>
+ %Av = memref.alloc(%c32, %c32) : memref<? x ? x f32, 2>
+ %tag = memref.alloc() : memref<1 x i32>
// Double buffering for dynamic shaped buffer.
-// CHECK: alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 2>
+// CHECK: memref.alloc(%{{.*}}, %{{.*}}) : memref<?x?xf32, 2>
// CHECK-NEXT: %[[C0:.*]] = constant 0 : index
-// CHECK-NEXT: dim %{{.*}}, %[[C0]] : memref<?x?xf32, 2>
+// CHECK-NEXT: memref.dim %{{.*}}, %[[C0]] : memref<?x?xf32, 2>
// CHECK-NEXT: %[[C1:.*]] = constant 1 : index
-// CHECK-NEXT: dim %{{.*}}, %[[C1]] : memref<?x?xf32, 2>
-// CHECK-NEXT: alloc(%{{.*}}, %{{.*}}) : memref<2x?x?xf32, 2>
+// CHECK-NEXT: memref.dim %{{.*}}, %[[C1]] : memref<?x?xf32, 2>
+// CHECK-NEXT: memref.alloc(%{{.*}}, %{{.*}}) : memref<2x?x?xf32, 2>
// CHECK: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}} mod 2, 0, 0], %{{.*}}[%{{.*}} mod 2, 0], %{{.*}}
affine.for %kTT = 0 to 16 {
affine.dma_start %arg0[%zero, %zero], %Av[%zero, %zero], %tag[%zero], %num_elt :
@@ -342,7 +342,7 @@ func @dynamic_shape_dma_buffer(%arg0: memref<512 x 32 x f32>) {
memref<? x ? x f32, 2>, memref<1 x i32>
affine.dma_wait %tag[%zero], %num_elt : memref<1 x i32>
}
- dealloc %Av : memref<? x ? x f32, 2>
+ memref.dealloc %Av : memref<? x ? x f32, 2>
return
// CHECK-NEXT: affine.for %{{.*}} = 1 to 16 {
// CHECK: affine.dma_start %{{.*}}[%{{.*}}, %{{.*}}], %{{.*}}[%{{.*}} mod 2, 0, 0], %{{.*}}[%{{.*}} mod 2, 0], %{{.*}}
@@ -358,9 +358,9 @@ func @dynamic_shape_dma_buffer(%arg0: memref<512 x 32 x f32>) {
// before performing any replacement.
// CHECK-LABEL: func @escaping_and_indexed_use_mix
func @escaping_and_indexed_use_mix() {
- %A = alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
- %Ah = alloc() : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
- %tag = alloc() : memref<1 x f32>
+ %A = memref.alloc() : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
+ %Ah = memref.alloc() : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
+ %tag = memref.alloc() : memref<1 x f32>
%zero = constant 0 : index
%num_elts = constant 32 : index
@@ -372,8 +372,8 @@ func @escaping_and_indexed_use_mix() {
%v = affine.load %Ah[%i] : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
"foo"(%v) : (f32) -> ()
}
- dealloc %A : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
- dealloc %Ah : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
+ memref.dealloc %A : memref<256 x f32, affine_map<(d0) -> (d0)>, 0>
+ memref.dealloc %Ah : memref<32 x f32, affine_map<(d0) -> (d0)>, 1>
return
}
// No replacement.
diff --git a/mlir/test/Transforms/promote-buffers-to-stack.mlir b/mlir/test/Transforms/promote-buffers-to-stack.mlir
index bf07a1290de3..b92c9ff4f668 100644
--- a/mlir/test/Transforms/promote-buffers-to-stack.mlir
+++ b/mlir/test/Transforms/promote-buffers-to-stack.mlir
@@ -21,7 +21,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
@@ -31,7 +31,7 @@ func @condBranch(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-NEXT: cond_br {{.*}}
// CHECK: ^bb2
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: test.copy
// CHECK-NEXT: return
@@ -56,7 +56,7 @@ func @condBranchDynamicType(
^bb1:
br ^bb3(%arg1 : memref<?xf32>)
^bb2(%0: index):
- %1 = alloc(%0) : memref<?xf32>
+ %1 = memref.alloc(%0) : memref<?xf32>
test.buffer_based in(%arg1: memref<?xf32>) out(%1: memref<?xf32>)
br ^bb3(%1 : memref<?xf32>)
^bb3(%2: memref<?xf32>):
@@ -67,7 +67,7 @@ func @condBranchDynamicType(
// CHECK-NEXT: cond_br
// CHECK: ^bb2
// CHECK: ^bb2(%[[IDX:.*]]:{{.*}})
-// CHECK-NEXT: %[[ALLOC0:.*]] = alloc(%[[IDX]])
+// CHECK-NEXT: %[[ALLOC0:.*]] = memref.alloc(%[[IDX]])
// CHECK-NEXT: test.buffer_based
// CHECK: br ^bb3
// CHECK-NEXT: ^bb3(%[[ALLOC0:.*]]:{{.*}})
@@ -79,35 +79,35 @@ func @condBranchDynamicType(
// CHECK-LABEL: func @dynamicRanked
func @dynamicRanked(%tensor: tensor<*xf32>) {
%0 = rank %tensor : tensor<*xf32>
- %1 = alloc(%0) : memref<?xindex>
+ %1 = memref.alloc(%0) : memref<?xindex>
return
}
// CHECK-NEXT: %[[RANK:.*]] = rank
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca(%[[RANK]])
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca(%[[RANK]])
// -----
// CHECK-LABEL: func @dynamicRanked2D
func @dynamicRanked2D(%tensor: tensor<*xf32>) {
%0 = rank %tensor : tensor<*xf32>
- %1 = alloc(%0, %0) : memref<?x?xindex>
+ %1 = memref.alloc(%0, %0) : memref<?x?xindex>
return
}
// CHECK-NEXT: %[[RANK:.*]] = rank
-// RANK-NEXT: %[[ALLOC:.*]] = alloca(%[[RANK]], %[[RANK]])
-// DEFINDEX-NEXT: %[[ALLOC:.*]] = alloc(%[[RANK]], %[[RANK]])
+// RANK-NEXT: %[[ALLOC:.*]] = memref.alloca(%[[RANK]], %[[RANK]])
+// DEFINDEX-NEXT: %[[ALLOC:.*]] = memref.alloc(%[[RANK]], %[[RANK]])
// -----
// CHECK-LABEL: func @dynamicNoRank
func @dynamicNoRank(%arg0: index) {
- %0 = alloc(%arg0) : memref<?xindex>
+ %0 = memref.alloc(%arg0) : memref<?xindex>
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc
// -----
@@ -117,10 +117,10 @@ func @dynamicNoRank(%arg0: index) {
// CHECK-LABEL: func @emptyUsesValue
func @emptyUsesValue(%arg0: memref<4xf32>) {
- %0 = alloc() : memref<4xf32>
+ %0 = memref.alloc() : memref<4xf32>
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: return
// -----
@@ -138,7 +138,7 @@ func @emptyUsesValue(%arg0: memref<4xf32>) {
func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
cond_br %arg0, ^bb1, ^bb2(%arg1 : memref<2xf32>)
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
br ^bb2(%0 : memref<2xf32>)
^bb2(%1: memref<2xf32>):
@@ -148,7 +148,7 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-NEXT: cond_br {{.*}}
// CHECK: ^bb1
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: test.copy
// CHECK-NEXT: return
@@ -164,7 +164,7 @@ func @criticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @invCriticalEdge
func @invCriticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0, ^bb1, ^bb2(%arg1 : memref<2xf32>)
^bb1:
@@ -174,7 +174,7 @@ func @invCriticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: cond_br
// CHECK: test.copy
// CHECK-NEXT: return
@@ -191,7 +191,7 @@ func @invCriticalEdge(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElse
func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -201,15 +201,15 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb2(%3: memref<2xf32>, %4: memref<2xf32>):
br ^bb3(%3, %4 : memref<2xf32>, memref<2xf32>)
^bb3(%5: memref<2xf32>, %6: memref<2xf32>):
- %7 = alloc() : memref<2xf32>
+ %7 = memref.alloc() : memref<2xf32>
test.buffer_based in(%5: memref<2xf32>) out(%7: memref<2xf32>)
test.copy(%7, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOCA0:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA0:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOCA1:.*]] = alloca()
+// CHECK: %[[ALLOCA1:.*]] = memref.alloca()
// CHECK: test.buffer_based
// CHECK: test.copy(%[[ALLOCA1]]
// CHECK-NEXT: return
@@ -226,7 +226,7 @@ func @ifElse(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNoUsers
func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -240,7 +240,7 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: return
// -----
@@ -259,7 +259,7 @@ func @ifElseNoUsers(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @ifElseNested
func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>)
cond_br %arg0,
^bb1(%arg1, %0 : memref<2xf32>, memref<2xf32>),
@@ -273,15 +273,15 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb4(%6: memref<2xf32>):
br ^bb5(%3, %6 : memref<2xf32>, memref<2xf32>)
^bb5(%7: memref<2xf32>, %8: memref<2xf32>):
- %9 = alloc() : memref<2xf32>
+ %9 = memref.alloc() : memref<2xf32>
test.buffer_based in(%7: memref<2xf32>) out(%9: memref<2xf32>)
test.copy(%9, %arg2) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOCA0:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA0:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based
-// CHECK: %[[ALLOCA1:.*]] = alloca()
+// CHECK: %[[ALLOCA1:.*]] = memref.alloca()
// CHECK: test.buffer_based
// CHECK: test.copy(%[[ALLOCA1]]
// CHECK-NEXT: return
@@ -294,17 +294,17 @@ func @ifElseNested(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
// CHECK-LABEL: func @redundantOperations
func @redundantOperations(%arg0: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%0: memref<2xf32>) out(%1: memref<2xf32>)
return
}
// CHECK: (%[[ARG0:.*]]: {{.*}})
-// CHECK-NEXT: %[[ALLOCA0:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA0:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based in(%[[ARG0]]{{.*}} out(%[[ALLOCA0]]
-// CHECK: %[[ALLOCA1:.*]] = alloca()
+// CHECK: %[[ALLOCA1:.*]] = memref.alloca()
// CHECK-NEXT: test.buffer_based in(%[[ALLOCA0]]{{.*}} out(%[[ALLOCA1]]
// CHECK: return
@@ -326,11 +326,11 @@ func @moving_alloc_and_inserting_missing_dealloc(
%arg1: memref<2xf32>) {
cond_br %cond, ^bb1, ^bb2
^bb1:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%0: memref<2xf32>)
br ^exit(%0 : memref<2xf32>)
^bb2:
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg0: memref<2xf32>) out(%1: memref<2xf32>)
br ^exit(%1 : memref<2xf32>)
^exit(%arg2: memref<2xf32>):
@@ -340,9 +340,9 @@ func @moving_alloc_and_inserting_missing_dealloc(
// CHECK-NEXT: cond_br {{.*}}
// CHECK: ^bb1
-// CHECK-NEXT: %{{.*}} = alloca()
+// CHECK-NEXT: %{{.*}} = memref.alloca()
// CHECK: ^bb2
-// CHECK-NEXT: %{{.*}} = alloca()
+// CHECK-NEXT: %{{.*}} = memref.alloca()
// CHECK: test.copy
// CHECK-NEXT: return
@@ -362,10 +362,10 @@ func @nested_regions_and_cond_branch(
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
test.region_buffer_based in(%arg1: memref<2xf32>) out(%0: memref<2xf32>) {
^bb0(%gen1_arg0: f32, %gen1_arg1: f32):
- %1 = alloc() : memref<2xf32>
+ %1 = memref.alloc() : memref<2xf32>
test.buffer_based in(%arg1: memref<2xf32>) out(%1: memref<2xf32>)
%tmp1 = math.exp %gen1_arg0 : f32
test.region_yield %tmp1 : f32
@@ -378,9 +378,9 @@ func @nested_regions_and_cond_branch(
// CHECK-NEXT: cond_br {{.*}}
// CHECK: ^bb2
-// CHECK-NEXT: %[[ALLOCA0:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA0:.*]] = memref.alloca()
// CHECK: ^bb0
-// CHECK-NEXT: %[[ALLOCA1:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOCA1:.*]] = memref.alloc()
// -----
@@ -394,16 +394,16 @@ func @memref_in_function_results(
%arg0: memref<5xf32>,
%arg1: memref<10xf32>,
%arg2: memref<5xf32>) -> (memref<10xf32>, memref<15xf32>) {
- %x = alloc() : memref<15xf32>
- %y = alloc() : memref<5xf32>
+ %x = memref.alloc() : memref<15xf32>
+ %y = memref.alloc() : memref<5xf32>
test.buffer_based in(%arg0: memref<5xf32>) out(%y: memref<5xf32>)
test.copy(%y, %arg2) : (memref<5xf32>, memref<5xf32>)
return %arg1, %x : memref<10xf32>, memref<15xf32>
}
// CHECK: (%[[ARG0:.*]]: memref<5xf32>, %[[ARG1:.*]]: memref<10xf32>,
// CHECK-SAME: %[[RESULT:.*]]: memref<5xf32>)
-// CHECK: %[[ALLOC:.*]] = alloc()
-// CHECK: %[[ALLOCA:.*]] = alloca()
+// CHECK: %[[ALLOC:.*]] = memref.alloc()
+// CHECK: %[[ALLOCA:.*]] = memref.alloca()
// CHECK: test.copy
// CHECK: return %[[ARG1]], %[[ALLOC]]
@@ -418,20 +418,20 @@ func @nested_region_control_flow(
%arg0 : index,
%arg1 : index) -> memref<?x?xf32> {
%0 = cmpi eq, %arg0, %arg1 : index
- %1 = alloc(%arg0, %arg0) : memref<?x?xf32>
+ %1 = memref.alloc(%arg0, %arg0) : memref<?x?xf32>
%2 = scf.if %0 -> (memref<?x?xf32>) {
scf.yield %1 : memref<?x?xf32>
} else {
- %3 = alloc(%arg0, %arg1) : memref<?x?xf32>
+ %3 = memref.alloc(%arg0, %arg1) : memref<?x?xf32>
scf.yield %1 : memref<?x?xf32>
}
return %2 : memref<?x?xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc(%arg0, %arg0)
+// CHECK: %[[ALLOC0:.*]] = memref.alloc(%arg0, %arg0)
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.if
// CHECK: scf.yield %[[ALLOC0]]
-// CHECK: %[[ALLOC2:.*]] = alloc(%arg0, %arg1)
+// CHECK: %[[ALLOC2:.*]] = memref.alloc(%arg0, %arg1)
// CHECK-NEXT: scf.yield %[[ALLOC0]]
// CHECK: return %[[ALLOC1]]
@@ -443,7 +443,7 @@ func @nested_region_control_flow(
// CHECK-LABEL: func @inner_region_control_flow
func @inner_region_control_flow(%arg0 : index) -> memref<2x2xf32> {
- %0 = alloc() : memref<2x2xf32>
+ %0 = memref.alloc() : memref<2x2xf32>
%1 = test.region_if %0 : memref<2x2xf32> -> (memref<2x2xf32>) then {
^bb0(%arg1 : memref<2x2xf32>):
test.region_if_yield %arg1 : memref<2x2xf32>
@@ -457,7 +457,7 @@ func @inner_region_control_flow(%arg0 : index) -> memref<2x2xf32> {
return %1 : memref<2x2xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: %[[ALLOC1:.*]] = test.region_if
// CHECK-NEXT: ^bb0(%[[ALLOC2:.*]]:{{.*}}):
// CHECK-NEXT: test.region_if_yield %[[ALLOC2]]
@@ -479,20 +479,20 @@ func @loop_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
- %3 = alloc() : memref<2xf32>
+ %3 = memref.alloc() : memref<2xf32>
scf.yield %3 : memref<2xf32>
}
test.copy(%1, %res) : (memref<2xf32>, memref<2xf32>)
return
}
-// CHECK-NEXT: %[[ALLOCA:.*]] = alloca()
+// CHECK-NEXT: %[[ALLOCA:.*]] = memref.alloca()
// CHECK-NEXT: scf.for
-// CHECK: %[[ALLOC:.*]] = alloc()
+// CHECK: %[[ALLOC:.*]] = memref.alloc()
// -----
@@ -509,7 +509,7 @@ func @loop_nested_if_no_alloc(
%step: index,
%buf: memref<2xf32>,
%res: memref<2xf32>) {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
@@ -524,7 +524,7 @@ func @loop_nested_if_no_alloc(
return
}
-// CHECK: %[[ALLOCA0:.*]] = alloca()
+// CHECK: %[[ALLOCA0:.*]] = memref.alloca()
// CHECK-NEXT: %[[ALLOCA1:.*]] = scf.for {{.*}} iter_args(%[[IALLOCA:.*]] =
// CHECK: %[[ALLOCA2:.*]] = scf.if
// CHECK: scf.yield %[[ALLOCA0]]
@@ -544,12 +544,12 @@ func @loop_nested_if_alloc(
%ub: index,
%step: index,
%buf: memref<2xf32>) -> memref<2xf32> {
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
%1 = scf.for %i = %lb to %ub step %step
iter_args(%iterBuf = %buf) -> memref<2xf32> {
%2 = cmpi eq, %i, %ub : index
%3 = scf.if %2 -> (memref<2xf32>) {
- %4 = alloc() : memref<2xf32>
+ %4 = memref.alloc() : memref<2xf32>
scf.yield %4 : memref<2xf32>
} else {
scf.yield %0 : memref<2xf32>
@@ -559,10 +559,10 @@ func @loop_nested_if_alloc(
return %1 : memref<2xf32>
}
-// CHECK: %[[ALLOC0:.*]] = alloc()
+// CHECK: %[[ALLOC0:.*]] = memref.alloc()
// CHECK-NEXT: %[[ALLOC1:.*]] = scf.for {{.*}}
// CHECK: %[[ALLOC2:.*]] = scf.if
-// CHECK: %[[ALLOC3:.*]] = alloc()
+// CHECK: %[[ALLOC3:.*]] = memref.alloc()
// CHECK-NEXT: scf.yield %[[ALLOC3]]
// CHECK: scf.yield %[[ALLOC0]]
// CHECK: scf.yield %[[ALLOC2]]
@@ -575,12 +575,12 @@ func @loop_nested_if_alloc(
// CHECK-LABEL: func @large_buffer_allocation
func @large_buffer_allocation(%arg0: memref<2048xf32>) {
- %0 = alloc() : memref<2048xf32>
+ %0 = memref.alloc() : memref<2048xf32>
test.copy(%0, %arg0) : (memref<2048xf32>, memref<2048xf32>)
return
}
-// CHECK-NEXT: %[[ALLOC:.*]] = alloc()
+// CHECK-NEXT: %[[ALLOC:.*]] = memref.alloc()
// CHECK-NEXT: test.copy
// -----
@@ -591,11 +591,11 @@ func @large_buffer_allocation(%arg0: memref<2048xf32>) {
// CHECK-LABEL: func @indexElementType
func @indexElementType() {
- %0 = alloc() : memref<4xindex>
+ %0 = memref.alloc() : memref<4xindex>
return
}
-// DEFINDEX-NEXT: alloca()
-// BIGINDEX-NEXT: alloca()
-// LOWLIMIT-NEXT: alloc()
-// RANK-NEXT: alloca()
+// DEFINDEX-NEXT: memref.alloca()
+// BIGINDEX-NEXT: memref.alloca()
+// LOWLIMIT-NEXT: memref.alloc()
+// RANK-NEXT: memref.alloca()
// CHECK-NEXT: return
diff --git a/mlir/test/lib/Dialect/Affine/TestAffineDataCopy.cpp b/mlir/test/lib/Dialect/Affine/TestAffineDataCopy.cpp
index d05d61fdf164..48085573a2d9 100644
--- a/mlir/test/lib/Dialect/Affine/TestAffineDataCopy.cpp
+++ b/mlir/test/lib/Dialect/Affine/TestAffineDataCopy.cpp
@@ -13,6 +13,7 @@
#include "mlir/Analysis/Utils.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
#include "mlir/Transforms/LoopUtils.h"
@@ -31,6 +32,9 @@ struct TestAffineDataCopy
TestAffineDataCopy() = default;
TestAffineDataCopy(const TestAffineDataCopy &pass){};
+ void getDependentDialects(DialectRegistry ®istry) const override {
+ registry.insert<memref::MemRefDialect>();
+ }
void runOnFunction() override;
private:
diff --git a/mlir/test/lib/Dialect/Test/TestDialect.cpp b/mlir/test/lib/Dialect/Test/TestDialect.cpp
index 991094d3b0b0..6da619652c07 100644
--- a/mlir/test/lib/Dialect/Test/TestDialect.cpp
+++ b/mlir/test/lib/Dialect/Test/TestDialect.cpp
@@ -10,6 +10,7 @@
#include "TestAttributes.h"
#include "TestTypes.h"
#include "mlir/Dialect/DLTI/DLTI.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BuiltinOps.h"
#include "mlir/IR/DialectImplementation.h"
@@ -680,7 +681,7 @@ LogicalResult OpWithShapedTypeInferTypeInterfaceOp::inferReturnTypeComponents(
LogicalResult OpWithShapedTypeInferTypeInterfaceOp::reifyReturnTypeShapes(
OpBuilder &builder, llvm::SmallVectorImpl<Value> &shapes) {
shapes = SmallVector<Value, 1>{
- builder.createOrFold<DimOp>(getLoc(), getOperand(0), 0)};
+ builder.createOrFold<memref::DimOp>(getLoc(), getOperand(0), 0)};
return success();
}
diff --git a/mlir/test/lib/Dialect/Test/TestPatterns.cpp b/mlir/test/lib/Dialect/Test/TestPatterns.cpp
index 277f6ccdc283..53651de0235a 100644
--- a/mlir/test/lib/Dialect/Test/TestPatterns.cpp
+++ b/mlir/test/lib/Dialect/Test/TestPatterns.cpp
@@ -7,6 +7,7 @@
//===----------------------------------------------------------------------===//
#include "TestDialect.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/StandardOps/Transforms/FuncConversions.h"
#include "mlir/IR/Matchers.h"
@@ -136,6 +137,10 @@ static void reifyReturnShape(Operation *op) {
struct TestReturnTypeDriver
: public PassWrapper<TestReturnTypeDriver, FunctionPass> {
+ void getDependentDialects(DialectRegistry ®istry) const override {
+ registry.insert<memref::MemRefDialect>();
+ }
+
void runOnFunction() override {
if (getFunction().getName() == "testCreateFunctions") {
std::vector<Operation *> ops;
diff --git a/mlir/test/lib/Transforms/TestConvVectorization.cpp b/mlir/test/lib/Transforms/TestConvVectorization.cpp
index 089b0c27a0e0..cda3542235da 100644
--- a/mlir/test/lib/Transforms/TestConvVectorization.cpp
+++ b/mlir/test/lib/Transforms/TestConvVectorization.cpp
@@ -37,6 +37,7 @@ class TestConvVectorization
void getDependentDialects(DialectRegistry ®istry) const override {
registry.insert<VectorDialect>();
registry.insert<linalg::LinalgDialect>();
+ registry.insert<memref::MemRefDialect>();
registry.insert<scf::SCFDialect>();
registry.insert<AffineDialect>();
registry.insert<StandardOpsDialect>();
diff --git a/mlir/test/lib/Transforms/TestGpuMemoryPromotion.cpp b/mlir/test/lib/Transforms/TestGpuMemoryPromotion.cpp
index ac53d97a3d3a..e583452f187e 100644
--- a/mlir/test/lib/Transforms/TestGpuMemoryPromotion.cpp
+++ b/mlir/test/lib/Transforms/TestGpuMemoryPromotion.cpp
@@ -14,6 +14,7 @@
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/GPU/GPUDialect.h"
#include "mlir/Dialect/GPU/MemoryPromotion.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/SPIRV/IR/SPIRVDialect.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
@@ -31,7 +32,8 @@ class TestGpuMemoryPromotionPass
: public PassWrapper<TestGpuMemoryPromotionPass,
OperationPass<gpu::GPUFuncOp>> {
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, StandardOpsDialect, scf::SCFDialect>();
+ registry.insert<AffineDialect, memref::MemRefDialect, StandardOpsDialect,
+ scf::SCFDialect>();
}
void runOnOperation() override {
diff --git a/mlir/test/lib/Transforms/TestGpuRewrite.cpp b/mlir/test/lib/Transforms/TestGpuRewrite.cpp
index 56512386f314..44ffd38bdc23 100644
--- a/mlir/test/lib/Transforms/TestGpuRewrite.cpp
+++ b/mlir/test/lib/Transforms/TestGpuRewrite.cpp
@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "mlir/Dialect/GPU/Passes.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Pass/Pass.h"
#include "mlir/Transforms/GreedyPatternRewriteDriver.h"
@@ -21,7 +22,7 @@ namespace {
struct TestGpuRewritePass
: public PassWrapper<TestGpuRewritePass, OperationPass<ModuleOp>> {
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<StandardOpsDialect>();
+ registry.insert<StandardOpsDialect, memref::MemRefDialect>();
}
void runOnOperation() override {
OwningRewritePatternList patterns;
diff --git a/mlir/test/lib/Transforms/TestLinalgCodegenStrategy.cpp b/mlir/test/lib/Transforms/TestLinalgCodegenStrategy.cpp
index 4fafec36e122..11a22bde9aac 100644
--- a/mlir/test/lib/Transforms/TestLinalgCodegenStrategy.cpp
+++ b/mlir/test/lib/Transforms/TestLinalgCodegenStrategy.cpp
@@ -36,6 +36,7 @@ struct TestLinalgCodegenStrategy
registry.insert<AffineDialect,
gpu::GPUDialect,
linalg::LinalgDialect,
+ memref::MemRefDialect,
scf::SCFDialect,
StandardOpsDialect,
vector::VectorDialect>();
diff --git a/mlir/test/lib/Transforms/TestLinalgFusionTransforms.cpp b/mlir/test/lib/Transforms/TestLinalgFusionTransforms.cpp
index 921f177d5d1c..1efc565a1c60 100644
--- a/mlir/test/lib/Transforms/TestLinalgFusionTransforms.cpp
+++ b/mlir/test/lib/Transforms/TestLinalgFusionTransforms.cpp
@@ -102,8 +102,8 @@ struct TestLinalgFusionTransforms
TestLinalgFusionTransforms(const TestLinalgFusionTransforms &pass) {}
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, linalg::LinalgDialect, scf::SCFDialect,
- StandardOpsDialect>();
+ registry.insert<AffineDialect, linalg::LinalgDialect, memref::MemRefDialect,
+ scf::SCFDialect, StandardOpsDialect>();
}
void runOnFunction() override {
@@ -211,7 +211,8 @@ struct TestLinalgTileAndFuseSequencePass
llvm::cl::ZeroOrMore, llvm::cl::MiscFlags::CommaSeparated};
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, linalg::LinalgDialect, scf::SCFDialect>();
+ registry.insert<AffineDialect, linalg::LinalgDialect, memref::MemRefDialect,
+ scf::SCFDialect>();
}
void runOnFunction() override {
diff --git a/mlir/test/lib/Transforms/TestLinalgTransforms.cpp b/mlir/test/lib/Transforms/TestLinalgTransforms.cpp
index ccdcf2b9bdf8..6cc390f54ca2 100644
--- a/mlir/test/lib/Transforms/TestLinalgTransforms.cpp
+++ b/mlir/test/lib/Transforms/TestLinalgTransforms.cpp
@@ -35,6 +35,7 @@ struct TestLinalgTransforms
void getDependentDialects(DialectRegistry ®istry) const override {
// clang-format off
registry.insert<AffineDialect,
+ memref::MemRefDialect,
scf::SCFDialect,
StandardOpsDialect,
vector::VectorDialect,
@@ -261,22 +262,22 @@ static void fillL1TilingAndMatmulToVectorPatterns(
//===----------------------------------------------------------------------===//
// Allocation call back
-static Optional<Value> allocCallBackFn(OpBuilder &b, SubViewOp subView,
+static Optional<Value> allocCallBackFn(OpBuilder &b, memref::SubViewOp subView,
ArrayRef<Value> boundingSubViewSize,
OperationFolder *folder) {
SmallVector<int64_t, 4> shape(boundingSubViewSize.size(), -1);
return b
- .create<AllocOp>(subView.getLoc(),
- MemRefType::get(shape,
- subView.getType().getElementType(),
- /*affineMapComposition =*/{}, 3),
- boundingSubViewSize)
+ .create<memref::AllocOp>(
+ subView.getLoc(),
+ MemRefType::get(shape, subView.getType().getElementType(),
+ /*affineMapComposition =*/{}, 3),
+ boundingSubViewSize)
.getResult();
}
// Deallocation callback
static LogicalResult deallocCallBackFn(OpBuilder &b, Value buffer) {
- b.create<DeallocOp>(buffer.getLoc(), buffer);
+ b.create<memref::DeallocOp>(buffer.getLoc(), buffer);
return success();
}
diff --git a/mlir/test/lib/Transforms/TestMemRefStrideCalculation.cpp b/mlir/test/lib/Transforms/TestMemRefStrideCalculation.cpp
index 3d7625973126..8b5c507b3794 100644
--- a/mlir/test/lib/Transforms/TestMemRefStrideCalculation.cpp
+++ b/mlir/test/lib/Transforms/TestMemRefStrideCalculation.cpp
@@ -6,6 +6,7 @@
//
//===----------------------------------------------------------------------===//
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/IR/BuiltinTypes.h"
#include "mlir/Pass/Pass.h"
@@ -23,7 +24,7 @@ struct TestMemRefStrideCalculation
/// Traverse AllocOp and compute strides of each MemRefType independently.
void TestMemRefStrideCalculation::runOnFunction() {
llvm::outs() << "Testing: " << getFunction().getName() << "\n";
- getFunction().walk([&](AllocOp allocOp) {
+ getFunction().walk([&](memref::AllocOp allocOp) {
auto memrefType = allocOp.getResult().getType().cast<MemRefType>();
int64_t offset;
SmallVector<int64_t, 4> strides;
diff --git a/mlir/test/lib/Transforms/TestSparsification.cpp b/mlir/test/lib/Transforms/TestSparsification.cpp
index 72c8f7b7e826..a76b8664f76c 100644
--- a/mlir/test/lib/Transforms/TestSparsification.cpp
+++ b/mlir/test/lib/Transforms/TestSparsification.cpp
@@ -50,8 +50,8 @@ struct TestSparsification
/// Registers all dialects required by testing.
void getDependentDialects(DialectRegistry ®istry) const override {
- registry
- .insert<scf::SCFDialect, vector::VectorDialect, LLVM::LLVMDialect>();
+ registry.insert<memref::MemRefDialect, scf::SCFDialect,
+ vector::VectorDialect, LLVM::LLVMDialect>();
}
/// Returns parallelization strategy given on command line.
@@ -134,8 +134,7 @@ namespace test {
void registerTestSparsification() {
PassRegistration<TestSparsification> sparsificationPass(
- "test-sparsification",
- "Test automatic generation of sparse tensor code");
+ "test-sparsification", "Test automatic generation of sparse tensor code");
}
} // namespace test
diff --git a/mlir/test/lib/Transforms/TestVectorTransforms.cpp b/mlir/test/lib/Transforms/TestVectorTransforms.cpp
index d45235043536..f11ee1347f57 100644
--- a/mlir/test/lib/Transforms/TestVectorTransforms.cpp
+++ b/mlir/test/lib/Transforms/TestVectorTransforms.cpp
@@ -11,6 +11,7 @@
#include "mlir/Analysis/SliceAnalysis.h"
#include "mlir/Dialect/Affine/IR/AffineOps.h"
#include "mlir/Dialect/Linalg/IR/LinalgOps.h"
+#include "mlir/Dialect/MemRef/IR/MemRef.h"
#include "mlir/Dialect/SCF/SCF.h"
#include "mlir/Dialect/StandardOps/IR/Ops.h"
#include "mlir/Dialect/Vector/VectorOps.h"
@@ -268,7 +269,7 @@ struct TestVectorToLoopPatterns
type.getNumElements() % multiplicity != 0)
return mlir::WalkResult::advance();
auto filterAlloc = [](Operation *op) {
- if (isa<ConstantOp, AllocOp, CallOp>(op))
+ if (isa<ConstantOp, memref::AllocOp, CallOp>(op))
return false;
return true;
};
@@ -335,7 +336,8 @@ struct TestVectorTransferFullPartialSplitPatterns
const TestVectorTransferFullPartialSplitPatterns &pass) {}
void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<AffineDialect, linalg::LinalgDialect, scf::SCFDialect>();
+ registry.insert<AffineDialect, linalg::LinalgDialect, memref::MemRefDialect,
+ scf::SCFDialect>();
}
Option<bool> useLinalgOps{
@@ -363,6 +365,9 @@ struct TestVectorTransferOpt
struct TestVectorTransferLoweringPatterns
: public PassWrapper<TestVectorTransferLoweringPatterns, FunctionPass> {
+ void getDependentDialects(DialectRegistry ®istry) const override {
+ registry.insert<memref::MemRefDialect>();
+ }
void runOnFunction() override {
OwningRewritePatternList patterns;
populateVectorTransferLoweringPatterns(patterns, &getContext());
diff --git a/mlir/test/mlir-cpu-runner/async-value.mlir b/mlir/test/mlir-cpu-runner/async-value.mlir
index 1460875f8ab8..e58f9e06c1ae 100644
--- a/mlir/test/mlir-cpu-runner/async-value.mlir
+++ b/mlir/test/mlir-cpu-runner/async-value.mlir
@@ -44,13 +44,13 @@ func @main() {
// Memref allocated inside async.execute region.
// ------------------------------------------------------------------------ //
%token2, %result2 = async.execute[%token0] -> !async.value<memref<f32>> {
- %5 = alloc() : memref<f32>
+ %5 = memref.alloc() : memref<f32>
%c0 = constant 0.25 : f32
- store %c0, %5[]: memref<f32>
+ memref.store %c0, %5[]: memref<f32>
async.yield %5 : memref<f32>
}
%6 = async.await %result2 : !async.value<memref<f32>>
- %7 = memref_cast %6 : memref<f32> to memref<*xf32>
+ %7 = memref.cast %6 : memref<f32> to memref<*xf32>
// CHECK: Unranked Memref
// CHECK-SAME: rank = 0 offset = 0 sizes = [] strides = []
@@ -61,9 +61,9 @@ func @main() {
// Memref passed as async.execute operand.
// ------------------------------------------------------------------------ //
%token3 = async.execute(%result2 as %unwrapped : !async.value<memref<f32>>) {
- %8 = load %unwrapped[]: memref<f32>
+ %8 = memref.load %unwrapped[]: memref<f32>
%9 = addf %8, %8 : f32
- store %9, %unwrapped[]: memref<f32>
+ memref.store %9, %unwrapped[]: memref<f32>
async.yield
}
async.await %token3 : !async.token
@@ -73,7 +73,7 @@ func @main() {
// CHECK-NEXT: [0.5]
call @print_memref_f32(%7): (memref<*xf32>) -> ()
- dealloc %6 : memref<f32>
+ memref.dealloc %6 : memref<f32>
return
}
diff --git a/mlir/test/mlir-cpu-runner/async.mlir b/mlir/test/mlir-cpu-runner/async.mlir
index ce30741e1015..30967928fc62 100644
--- a/mlir/test/mlir-cpu-runner/async.mlir
+++ b/mlir/test/mlir-cpu-runner/async.mlir
@@ -24,23 +24,23 @@ func @main() {
%c3 = constant 3.0 : f32
%c4 = constant 4.0 : f32
- %A = alloc() : memref<4xf32>
+ %A = memref.alloc() : memref<4xf32>
linalg.fill(%A, %c0) : memref<4xf32>, f32
// CHECK: [0, 0, 0, 0]
- %U = memref_cast %A : memref<4xf32> to memref<*xf32>
+ %U = memref.cast %A : memref<4xf32> to memref<*xf32>
call @print_memref_f32(%U): (memref<*xf32>) -> ()
// CHECK: Current thread id: [[MAIN:.*]]
// CHECK: [1, 0, 0, 0]
- store %c1, %A[%i0]: memref<4xf32>
+ memref.store %c1, %A[%i0]: memref<4xf32>
call @mlirAsyncRuntimePrintCurrentThreadId(): () -> ()
call @print_memref_f32(%U): (memref<*xf32>) -> ()
%outer = async.execute {
// CHECK: Current thread id: [[THREAD0:.*]]
// CHECK: [1, 2, 0, 0]
- store %c2, %A[%i1]: memref<4xf32>
+ memref.store %c2, %A[%i1]: memref<4xf32>
call @mlirAsyncRuntimePrintCurrentThreadId(): () -> ()
call @print_memref_f32(%U): (memref<*xf32>) -> ()
@@ -54,7 +54,7 @@ func @main() {
%inner = async.execute [%noop] {
// CHECK: Current thread id: [[THREAD2:.*]]
// CHECK: [1, 2, 3, 0]
- store %c3, %A[%i2]: memref<4xf32>
+ memref.store %c3, %A[%i2]: memref<4xf32>
call @mlirAsyncRuntimePrintCurrentThreadId(): () -> ()
call @print_memref_f32(%U): (memref<*xf32>) -> ()
@@ -64,7 +64,7 @@ func @main() {
// CHECK: Current thread id: [[THREAD3:.*]]
// CHECK: [1, 2, 3, 4]
- store %c4, %A[%i3]: memref<4xf32>
+ memref.store %c4, %A[%i3]: memref<4xf32>
call @mlirAsyncRuntimePrintCurrentThreadId(): () -> ()
call @print_memref_f32(%U): (memref<*xf32>) -> ()
@@ -77,7 +77,7 @@ func @main() {
call @mlirAsyncRuntimePrintCurrentThreadId(): () -> ()
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<4xf32>
+ memref.dealloc %A : memref<4xf32>
return
}
diff --git a/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir b/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir
index 155c3e858420..dd644ddbe322 100644
--- a/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir
+++ b/mlir/test/mlir-cpu-runner/bare_ptr_call_conv.mlir
@@ -12,14 +12,14 @@ func @simple_add1_add2_test(%arg0: memref<2xf32>, %arg1: memref<2xf32>) {
%cst = constant 1.000000e+00 : f32
%cst_0 = constant 2.000000e+00 : f32
scf.for %arg2 = %c0 to %c2 step %c1 {
- %0 = load %arg0[%arg2] : memref<2xf32>
+ %0 = memref.load %arg0[%arg2] : memref<2xf32>
%1 = addf %0, %cst : f32
- store %1, %arg0[%arg2] : memref<2xf32>
+ memref.store %1, %arg0[%arg2] : memref<2xf32>
// CHECK: 2, 2
- %2 = load %arg1[%arg2] : memref<2xf32>
+ %2 = memref.load %arg1[%arg2] : memref<2xf32>
%3 = addf %1, %cst_0 : f32
- store %3, %arg1[%arg2] : memref<2xf32>
+ memref.store %3, %arg1[%arg2] : memref<2xf32>
// CHECK-NEXT: 4, 4
}
return
@@ -39,30 +39,30 @@ func @main()
%c1 = constant 1 : index
%cst = constant 1.000000e+00 : f32
%cst_0 = constant 2.000000e+00 : f32
- %a = alloc() : memref<2xf32>
- %b = alloc() : memref<2xf32>
+ %a = memref.alloc() : memref<2xf32>
+ %b = memref.alloc() : memref<2xf32>
scf.for %i = %c0 to %c2 step %c1 {
- store %cst, %a[%i] : memref<2xf32>
- store %cst, %b[%i] : memref<2xf32>
+ memref.store %cst, %a[%i] : memref<2xf32>
+ memref.store %cst, %b[%i] : memref<2xf32>
}
call @simple_add1_add2_test(%a, %b) : (memref<2xf32>, memref<2xf32>) -> ()
- %l0 = load %a[%c0] : memref<2xf32>
+ %l0 = memref.load %a[%c0] : memref<2xf32>
call @printF32(%l0) : (f32) -> ()
call @printComma() : () -> ()
- %l1 = load %a[%c1] : memref<2xf32>
+ %l1 = memref.load %a[%c1] : memref<2xf32>
call @printF32(%l1) : (f32) -> ()
call @printNewline() : () -> ()
- %l2 = load %b[%c0] : memref<2xf32>
+ %l2 = memref.load %b[%c0] : memref<2xf32>
call @printF32(%l2) : (f32) -> ()
call @printComma() : () -> ()
- %l3 = load %b[%c1] : memref<2xf32>
+ %l3 = memref.load %b[%c1] : memref<2xf32>
call @printF32(%l3) : (f32) -> ()
call @printNewline() : () -> ()
- dealloc %a : memref<2xf32>
- dealloc %b : memref<2xf32>
+ memref.dealloc %a : memref<2xf32>
+ memref.dealloc %b : memref<2xf32>
return
}
diff --git a/mlir/test/mlir-cpu-runner/global_memref.mlir b/mlir/test/mlir-cpu-runner/global_memref.mlir
index 5dbd75a62a50..1d2a2e02dfb2 100644
--- a/mlir/test/mlir-cpu-runner/global_memref.mlir
+++ b/mlir/test/mlir-cpu-runner/global_memref.mlir
@@ -4,10 +4,10 @@ func private @print_memref_f32(memref<*xf32>) attributes { llvm.emit_c_interface
func private @print_memref_i32(memref<*xi32>) attributes { llvm.emit_c_interface }
func private @printNewline() -> ()
-global_memref "private" @gv0 : memref<4xf32> = dense<[0.0, 1.0, 2.0, 3.0]>
+memref.global "private" @gv0 : memref<4xf32> = dense<[0.0, 1.0, 2.0, 3.0]>
func @test1DMemref() {
- %0 = get_global_memref @gv0 : memref<4xf32>
- %U = memref_cast %0 : memref<4xf32> to memref<*xf32>
+ %0 = memref.get_global @gv0 : memref<4xf32>
+ %U = memref.cast %0 : memref<4xf32> to memref<*xf32>
// CHECK: rank = 1
// CHECK: offset = 0
// CHECK: sizes = [4]
@@ -21,8 +21,8 @@ func @test1DMemref() {
%c2 = constant 2 : index
%fp0 = constant 4.0 : f32
%fp1 = constant 5.0 : f32
- store %fp0, %0[%c0] : memref<4xf32>
- store %fp1, %0[%c2] : memref<4xf32>
+ memref.store %fp0, %0[%c0] : memref<4xf32>
+ memref.store %fp1, %0[%c2] : memref<4xf32>
// CHECK: rank = 1
// CHECK: offset = 0
// CHECK: sizes = [4]
@@ -33,10 +33,10 @@ func @test1DMemref() {
return
}
-global_memref constant @gv1 : memref<3x2xi32> = dense<[[0, 1],[2, 3],[4, 5]]>
+memref.global constant @gv1 : memref<3x2xi32> = dense<[[0, 1],[2, 3],[4, 5]]>
func @testConstantMemref() {
- %0 = get_global_memref @gv1 : memref<3x2xi32>
- %U = memref_cast %0 : memref<3x2xi32> to memref<*xi32>
+ %0 = memref.get_global @gv1 : memref<3x2xi32>
+ %U = memref.cast %0 : memref<3x2xi32> to memref<*xi32>
// CHECK: rank = 2
// CHECK: offset = 0
// CHECK: sizes = [3, 2]
@@ -49,10 +49,10 @@ func @testConstantMemref() {
return
}
-global_memref "private" @gv2 : memref<4x2xf32> = dense<[[0.0, 1.0], [2.0, 3.0], [4.0, 5.0], [6.0, 7.0]]>
+memref.global "private" @gv2 : memref<4x2xf32> = dense<[[0.0, 1.0], [2.0, 3.0], [4.0, 5.0], [6.0, 7.0]]>
func @test2DMemref() {
- %0 = get_global_memref @gv2 : memref<4x2xf32>
- %U = memref_cast %0 : memref<4x2xf32> to memref<*xf32>
+ %0 = memref.get_global @gv2 : memref<4x2xf32>
+ %U = memref.cast %0 : memref<4x2xf32> to memref<*xf32>
// CHECK: rank = 2
// CHECK: offset = 0
// CHECK: sizes = [4, 2]
@@ -68,7 +68,7 @@ func @test2DMemref() {
%c0 = constant 0 : index
%c1 = constant 1 : index
%fp10 = constant 10.0 : f32
- store %fp10, %0[%c0, %c1] : memref<4x2xf32>
+ memref.store %fp10, %0[%c0, %c1] : memref<4x2xf32>
// CHECK: rank = 2
// CHECK: offset = 0
// CHECK: sizes = [4, 2]
@@ -82,10 +82,10 @@ func @test2DMemref() {
return
}
-global_memref @gv3 : memref<i32> = dense<11>
+memref.global @gv3 : memref<i32> = dense<11>
func @testScalarMemref() {
- %0 = get_global_memref @gv3 : memref<i32>
- %U = memref_cast %0 : memref<i32> to memref<*xi32>
+ %0 = memref.get_global @gv3 : memref<i32>
+ %U = memref.cast %0 : memref<i32> to memref<*xi32>
// CHECK: rank = 0
// CHECK: offset = 0
// CHECK: sizes = []
diff --git a/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir b/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir
index 62a252988c96..b7442d5f5789 100644
--- a/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir
+++ b/mlir/test/mlir-cpu-runner/memref_reinterpret_cast.mlir
@@ -10,17 +10,17 @@ func @main() -> () {
%c1 = constant 1 : index
// Initialize input.
- %input = alloc() : memref<2x3xf32>
- %dim_x = dim %input, %c0 : memref<2x3xf32>
- %dim_y = dim %input, %c1 : memref<2x3xf32>
+ %input = memref.alloc() : memref<2x3xf32>
+ %dim_x = memref.dim %input, %c0 : memref<2x3xf32>
+ %dim_y = memref.dim %input, %c1 : memref<2x3xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%dim_x, %dim_y) step (%c1, %c1) {
%prod = muli %i, %dim_y : index
%val = addi %prod, %j : index
%val_i64 = index_cast %val : index to i64
%val_f32 = sitofp %val_i64 : i64 to f32
- store %val_f32, %input[%i, %j] : memref<2x3xf32>
+ memref.store %val_f32, %input[%i, %j] : memref<2x3xf32>
}
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
call @print_memref_f32(%unranked_input) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [2, 3] strides = [3, 1]
// CHECK-NEXT: [0, 1, 2]
@@ -35,11 +35,11 @@ func @main() -> () {
}
func @cast_ranked_memref_to_static_shape(%input : memref<2x3xf32>) {
- %output = memref_reinterpret_cast %input to
+ %output = memref.reinterpret_cast %input to
offset: [0], sizes: [6, 1], strides: [1, 1]
: memref<2x3xf32> to memref<6x1xf32>
- %unranked_output = memref_cast %output
+ %unranked_output = memref.cast %output
: memref<6x1xf32> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [6, 1] strides = [1, 1] data =
@@ -56,11 +56,11 @@ func @cast_ranked_memref_to_dynamic_shape(%input : memref<2x3xf32>) {
%c0 = constant 0 : index
%c1 = constant 1 : index
%c6 = constant 6 : index
- %output = memref_reinterpret_cast %input to
+ %output = memref.reinterpret_cast %input to
offset: [%c0], sizes: [%c1, %c6], strides: [%c6, %c1]
: memref<2x3xf32> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %unranked_output = memref_cast %output
+ %unranked_output = memref.cast %output
: memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [1, 6] strides = [6, 1] data =
@@ -69,12 +69,12 @@ func @cast_ranked_memref_to_dynamic_shape(%input : memref<2x3xf32>) {
}
func @cast_unranked_memref_to_static_shape(%input : memref<2x3xf32>) {
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
- %output = memref_reinterpret_cast %unranked_input to
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
+ %output = memref.reinterpret_cast %unranked_input to
offset: [0], sizes: [6, 1], strides: [1, 1]
: memref<*xf32> to memref<6x1xf32>
- %unranked_output = memref_cast %output
+ %unranked_output = memref.cast %output
: memref<6x1xf32> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [6, 1] strides = [1, 1] data =
@@ -88,15 +88,15 @@ func @cast_unranked_memref_to_static_shape(%input : memref<2x3xf32>) {
}
func @cast_unranked_memref_to_dynamic_shape(%input : memref<2x3xf32>) {
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c6 = constant 6 : index
- %output = memref_reinterpret_cast %unranked_input to
+ %output = memref.reinterpret_cast %unranked_input to
offset: [%c0], sizes: [%c1, %c6], strides: [%c6, %c1]
: memref<*xf32> to memref<?x?xf32, offset: ?, strides: [?, ?]>
- %unranked_output = memref_cast %output
+ %unranked_output = memref.cast %output
: memref<?x?xf32, offset: ?, strides: [?, ?]> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [1, 6] strides = [6, 1] data =
diff --git a/mlir/test/mlir-cpu-runner/memref_reshape.mlir b/mlir/test/mlir-cpu-runner/memref_reshape.mlir
index aa031802350c..f6d01d9e6af9 100644
--- a/mlir/test/mlir-cpu-runner/memref_reshape.mlir
+++ b/mlir/test/mlir-cpu-runner/memref_reshape.mlir
@@ -11,28 +11,28 @@ func @main() -> () {
%c1 = constant 1 : index
// Initialize input.
- %input = alloc() : memref<2x3xf32>
- %dim_x = dim %input, %c0 : memref<2x3xf32>
- %dim_y = dim %input, %c1 : memref<2x3xf32>
+ %input = memref.alloc() : memref<2x3xf32>
+ %dim_x = memref.dim %input, %c0 : memref<2x3xf32>
+ %dim_y = memref.dim %input, %c1 : memref<2x3xf32>
scf.parallel (%i, %j) = (%c0, %c0) to (%dim_x, %dim_y) step (%c1, %c1) {
%prod = muli %i, %dim_y : index
%val = addi %prod, %j : index
%val_i64 = index_cast %val : index to i64
%val_f32 = sitofp %val_i64 : i64 to f32
- store %val_f32, %input[%i, %j] : memref<2x3xf32>
+ memref.store %val_f32, %input[%i, %j] : memref<2x3xf32>
}
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
call @print_memref_f32(%unranked_input) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [2, 3] strides = [3, 1]
// CHECK-NEXT: [0, 1, 2]
// CHECK-NEXT: [3, 4, 5]
// Initialize shape.
- %shape = alloc() : memref<2xindex>
+ %shape = memref.alloc() : memref<2xindex>
%c2 = constant 2 : index
%c3 = constant 3 : index
- store %c3, %shape[%c0] : memref<2xindex>
- store %c2, %shape[%c1] : memref<2xindex>
+ memref.store %c3, %shape[%c0] : memref<2xindex>
+ memref.store %c2, %shape[%c1] : memref<2xindex>
// Test cases.
call @reshape_ranked_memref_to_ranked(%input, %shape)
@@ -48,10 +48,10 @@ func @main() -> () {
func @reshape_ranked_memref_to_ranked(%input : memref<2x3xf32>,
%shape : memref<2xindex>) {
- %output = memref_reshape %input(%shape)
+ %output = memref.reshape %input(%shape)
: (memref<2x3xf32>, memref<2xindex>) -> memref<?x?xf32>
- %unranked_output = memref_cast %output : memref<?x?xf32> to memref<*xf32>
+ %unranked_output = memref.cast %output : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [3, 2] strides = [2, 1] data =
// CHECK: [0, 1],
@@ -62,11 +62,11 @@ func @reshape_ranked_memref_to_ranked(%input : memref<2x3xf32>,
func @reshape_unranked_memref_to_ranked(%input : memref<2x3xf32>,
%shape : memref<2xindex>) {
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
- %output = memref_reshape %input(%shape)
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
+ %output = memref.reshape %input(%shape)
: (memref<2x3xf32>, memref<2xindex>) -> memref<?x?xf32>
- %unranked_output = memref_cast %output : memref<?x?xf32> to memref<*xf32>
+ %unranked_output = memref.cast %output : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%unranked_output) : (memref<*xf32>) -> ()
// CHECK: rank = 2 offset = 0 sizes = [3, 2] strides = [2, 1] data =
// CHECK: [0, 1],
@@ -77,8 +77,8 @@ func @reshape_unranked_memref_to_ranked(%input : memref<2x3xf32>,
func @reshape_ranked_memref_to_unranked(%input : memref<2x3xf32>,
%shape : memref<2xindex>) {
- %dyn_size_shape = memref_cast %shape : memref<2xindex> to memref<?xindex>
- %output = memref_reshape %input(%dyn_size_shape)
+ %dyn_size_shape = memref.cast %shape : memref<2xindex> to memref<?xindex>
+ %output = memref.reshape %input(%dyn_size_shape)
: (memref<2x3xf32>, memref<?xindex>) -> memref<*xf32>
call @print_memref_f32(%output) : (memref<*xf32>) -> ()
@@ -91,9 +91,9 @@ func @reshape_ranked_memref_to_unranked(%input : memref<2x3xf32>,
func @reshape_unranked_memref_to_unranked(%input : memref<2x3xf32>,
%shape : memref<2xindex>) {
- %unranked_input = memref_cast %input : memref<2x3xf32> to memref<*xf32>
- %dyn_size_shape = memref_cast %shape : memref<2xindex> to memref<?xindex>
- %output = memref_reshape %input(%dyn_size_shape)
+ %unranked_input = memref.cast %input : memref<2x3xf32> to memref<*xf32>
+ %dyn_size_shape = memref.cast %shape : memref<2xindex> to memref<?xindex>
+ %output = memref.reshape %input(%dyn_size_shape)
: (memref<2x3xf32>, memref<?xindex>) -> memref<*xf32>
call @print_memref_f32(%output) : (memref<*xf32>) -> ()
diff --git a/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir b/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir
index 5653d77ffede..c0da813375b1 100644
--- a/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir
+++ b/mlir/test/mlir-cpu-runner/sgemm_naive_codegen.mlir
@@ -1,9 +1,9 @@
// 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
func @main() {
- %A = alloc() : memref<16x16xf32>
- %B = alloc() : memref<16x16xf32>
- %C = alloc() : memref<16x16xf32>
+ %A = memref.alloc() : memref<16x16xf32>
+ %B = memref.alloc() : memref<16x16xf32>
+ %C = memref.alloc() : memref<16x16xf32>
%cf1 = constant 1.00000e+00 : f32
@@ -27,9 +27,9 @@ func @main() {
%c1 = constant 1 : index
%c2 = constant 2 : index
- %M = dim %C, %c0 : memref<16x16xf32>
- %N = dim %C, %c1 : memref<16x16xf32>
- %K = dim %A, %c1 : memref<16x16xf32>
+ %M = memref.dim %C, %c0 : memref<16x16xf32>
+ %N = memref.dim %C, %c1 : memref<16x16xf32>
+ %K = memref.dim %A, %c1 : memref<16x16xf32>
%f1 = muli %M, %N : index
%f2 = muli %f1, %K : index
@@ -50,7 +50,7 @@ func @sgemm_naive(%arg0: memref<16x16xf32>, %arg1: memref<16x16xf32>, %arg2: mem
%c0 = constant 0 : index
affine.for %arg3 = 0 to 16 {
affine.for %arg4 = 0 to 16 {
- %m = alloc() : memref<1xf32>
+ %m = memref.alloc() : memref<1xf32>
%v = affine.load %arg2[%arg3, %arg4] : memref<16x16xf32>
affine.store %v, %m[%c0] : memref<1xf32>
affine.for %arg5 = 0 to 16 {
@@ -63,7 +63,7 @@ func @sgemm_naive(%arg0: memref<16x16xf32>, %arg1: memref<16x16xf32>, %arg2: mem
}
%s = affine.load %m[%c0] : memref<1xf32>
affine.store %s, %arg2[%arg3, %arg4] : memref<16x16xf32>
- dealloc %m : memref<1xf32>
+ memref.dealloc %m : memref<1xf32>
}
}
return
diff --git a/mlir/test/mlir-cpu-runner/unranked_memref.mlir b/mlir/test/mlir-cpu-runner/unranked_memref.mlir
index af19f6952ece..72fe0271afb1 100644
--- a/mlir/test/mlir-cpu-runner/unranked_memref.mlir
+++ b/mlir/test/mlir-cpu-runner/unranked_memref.mlir
@@ -39,35 +39,35 @@
// CHECK-SAME: strides = [3, 1]
// CHECK-COUNT-4: [1, 1, 1]
func @main() -> () {
- %A = alloc() : memref<10x3xf32, 0>
+ %A = memref.alloc() : memref<10x3xf32, 0>
%f2 = constant 2.00000e+00 : f32
%f5 = constant 5.00000e+00 : f32
%f10 = constant 10.00000e+00 : f32
- %V = memref_cast %A : memref<10x3xf32, 0> to memref<?x?xf32>
+ %V = memref.cast %A : memref<10x3xf32, 0> to memref<?x?xf32>
linalg.fill(%V, %f10) : memref<?x?xf32, 0>, f32
- %U = memref_cast %A : memref<10x3xf32, 0> to memref<*xf32>
+ %U = memref.cast %A : memref<10x3xf32, 0> to memref<*xf32>
call @print_memref_f32(%U) : (memref<*xf32>) -> ()
- %V2 = memref_cast %U : memref<*xf32> to memref<?x?xf32>
+ %V2 = memref.cast %U : memref<*xf32> to memref<?x?xf32>
linalg.fill(%V2, %f5) : memref<?x?xf32, 0>, f32
- %U2 = memref_cast %V2 : memref<?x?xf32, 0> to memref<*xf32>
+ %U2 = memref.cast %V2 : memref<?x?xf32, 0> to memref<*xf32>
call @print_memref_f32(%U2) : (memref<*xf32>) -> ()
- %V3 = memref_cast %V2 : memref<?x?xf32> to memref<*xf32>
- %V4 = memref_cast %V3 : memref<*xf32> to memref<?x?xf32>
+ %V3 = memref.cast %V2 : memref<?x?xf32> to memref<*xf32>
+ %V4 = memref.cast %V3 : memref<*xf32> to memref<?x?xf32>
linalg.fill(%V4, %f2) : memref<?x?xf32, 0>, f32
- %U3 = memref_cast %V2 : memref<?x?xf32> to memref<*xf32>
+ %U3 = memref.cast %V2 : memref<?x?xf32> to memref<*xf32>
call @print_memref_f32(%U3) : (memref<*xf32>) -> ()
// 122 is ASCII for 'z'.
%i8_z = constant 122 : i8
- %I8 = alloc() : memref<i8>
- store %i8_z, %I8[]: memref<i8>
- %U4 = memref_cast %I8 : memref<i8> to memref<*xi8>
+ %I8 = memref.alloc() : memref<i8>
+ memref.store %i8_z, %I8[]: memref<i8>
+ %U4 = memref.cast %I8 : memref<i8> to memref<*xi8>
call @print_memref_i8(%U4) : (memref<*xi8>) -> ()
- dealloc %A : memref<10x3xf32, 0>
+ memref.dealloc %A : memref<10x3xf32, 0>
call @return_var_memref_caller() : () -> ()
call @return_two_var_memref_caller() : () -> ()
@@ -79,7 +79,7 @@ func private @print_memref_i8(memref<*xi8>) attributes { llvm.emit_c_interface }
func private @print_memref_f32(memref<*xf32>) attributes { llvm.emit_c_interface }
func @return_two_var_memref_caller() {
- %0 = alloca() : memref<4x3xf32>
+ %0 = memref.alloca() : memref<4x3xf32>
%c0f32 = constant 1.0 : f32
linalg.fill(%0, %c0f32) : memref<4x3xf32>, f32
%1:2 = call @return_two_var_memref(%0) : (memref<4x3xf32>) -> (memref<*xf32>, memref<*xf32>)
@@ -89,12 +89,12 @@ func @return_two_var_memref_caller() {
}
func @return_two_var_memref(%arg0: memref<4x3xf32>) -> (memref<*xf32>, memref<*xf32>) {
- %0 = memref_cast %arg0 : memref<4x3xf32> to memref<*xf32>
+ %0 = memref.cast %arg0 : memref<4x3xf32> to memref<*xf32>
return %0, %0 : memref<*xf32>, memref<*xf32>
}
func @return_var_memref_caller() {
- %0 = alloca() : memref<4x3xf32>
+ %0 = memref.alloca() : memref<4x3xf32>
%c0f32 = constant 1.0 : f32
linalg.fill(%0, %c0f32) : memref<4x3xf32>, f32
%1 = call @return_var_memref(%0) : (memref<4x3xf32>) -> memref<*xf32>
@@ -103,7 +103,7 @@ func @return_var_memref_caller() {
}
func @return_var_memref(%arg0: memref<4x3xf32>) -> memref<*xf32> {
- %0 = memref_cast %arg0: memref<4x3xf32> to memref<*xf32>
+ %0 = memref.cast %arg0: memref<4x3xf32> to memref<*xf32>
return %0 : memref<*xf32>
}
@@ -111,17 +111,17 @@ func private @printU64(index) -> ()
func private @printNewline() -> ()
func @dim_op_of_unranked() {
- %ranked = alloc() : memref<4x3xf32>
- %unranked = memref_cast %ranked: memref<4x3xf32> to memref<*xf32>
+ %ranked = memref.alloc() : memref<4x3xf32>
+ %unranked = memref.cast %ranked: memref<4x3xf32> to memref<*xf32>
%c0 = constant 0 : index
- %dim_0 = dim %unranked, %c0 : memref<*xf32>
+ %dim_0 = memref.dim %unranked, %c0 : memref<*xf32>
call @printU64(%dim_0) : (index) -> ()
call @printNewline() : () -> ()
// CHECK: 4
%c1 = constant 1 : index
- %dim_1 = dim %unranked, %c1 : memref<*xf32>
+ %dim_1 = memref.dim %unranked, %c1 : memref<*xf32>
call @printU64(%dim_1) : (index) -> ()
call @printNewline() : () -> ()
// CHECK: 3
diff --git a/mlir/test/mlir-cpu-runner/utils.mlir b/mlir/test/mlir-cpu-runner/utils.mlir
index b823ad4557ea..4a2eee086f9d 100644
--- a/mlir/test/mlir-cpu-runner/utils.mlir
+++ b/mlir/test/mlir-cpu-runner/utils.mlir
@@ -5,11 +5,11 @@
func @print_0d() {
%f = constant 2.00000e+00 : f32
- %A = alloc() : memref<f32>
- store %f, %A[]: memref<f32>
- %U = memref_cast %A : memref<f32> to memref<*xf32>
+ %A = memref.alloc() : memref<f32>
+ memref.store %f, %A[]: memref<f32>
+ %U = memref.cast %A : memref<f32> to memref<*xf32>
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<f32>
+ memref.dealloc %A : memref<f32>
return
}
// PRINT-0D: Unranked Memref base@ = {{.*}} rank = 0 offset = 0 sizes = [] strides = [] data =
@@ -17,12 +17,12 @@ func @print_0d() {
func @print_1d() {
%f = constant 2.00000e+00 : f32
- %A = alloc() : memref<16xf32>
- %B = memref_cast %A: memref<16xf32> to memref<?xf32>
+ %A = memref.alloc() : memref<16xf32>
+ %B = memref.cast %A: memref<16xf32> to memref<?xf32>
linalg.fill(%B, %f) : memref<?xf32>, f32
- %U = memref_cast %B : memref<?xf32> to memref<*xf32>
+ %U = memref.cast %B : memref<?xf32> to memref<*xf32>
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<16xf32>
+ memref.dealloc %A : memref<16xf32>
return
}
// PRINT-1D: Unranked Memref base@ = {{.*}} rank = 1 offset = 0 sizes = [16] strides = [1] data =
@@ -31,15 +31,15 @@ func @print_1d() {
func @print_3d() {
%f = constant 2.00000e+00 : f32
%f4 = constant 4.00000e+00 : f32
- %A = alloc() : memref<3x4x5xf32>
- %B = memref_cast %A: memref<3x4x5xf32> to memref<?x?x?xf32>
+ %A = memref.alloc() : memref<3x4x5xf32>
+ %B = memref.cast %A: memref<3x4x5xf32> to memref<?x?x?xf32>
linalg.fill(%B, %f) : memref<?x?x?xf32>, f32
%c2 = constant 2 : index
- store %f4, %B[%c2, %c2, %c2]: memref<?x?x?xf32>
- %U = memref_cast %B : memref<?x?x?xf32> to memref<*xf32>
+ memref.store %f4, %B[%c2, %c2, %c2]: memref<?x?x?xf32>
+ %U = memref.cast %B : memref<?x?x?xf32> to memref<*xf32>
call @print_memref_f32(%U): (memref<*xf32>) -> ()
- dealloc %A : memref<3x4x5xf32>
+ memref.dealloc %A : memref<3x4x5xf32>
return
}
// PRINT-3D: Unranked Memref base@ = {{.*}} rank = 3 offset = 0 sizes = [3, 4, 5] strides = [20, 5, 1] data =
@@ -57,13 +57,13 @@ func @vector_splat_2d() {
%c0 = constant 0 : index
%f10 = constant 10.0 : f32
%vf10 = splat %f10: !vector_type_C
- %C = alloc() : !matrix_type_CC
- store %vf10, %C[%c0, %c0]: !matrix_type_CC
+ %C = memref.alloc() : !matrix_type_CC
+ memref.store %vf10, %C[%c0, %c0]: !matrix_type_CC
- %CC = memref_cast %C: !matrix_type_CC to memref<?x?x!vector_type_C>
+ %CC = memref.cast %C: !matrix_type_CC to memref<?x?x!vector_type_C>
call @print_memref_vector_4x4xf32(%CC): (memref<?x?x!vector_type_C>) -> ()
- dealloc %C : !matrix_type_CC
+ memref.dealloc %C : !matrix_type_CC
return
}
diff --git a/mlir/test/mlir-opt/commandline.mlir b/mlir/test/mlir-opt/commandline.mlir
index d08b5a931013..508a22063339 100644
--- a/mlir/test/mlir-opt/commandline.mlir
+++ b/mlir/test/mlir-opt/commandline.mlir
@@ -13,6 +13,7 @@
// CHECK-NEXT: llvm
// CHECK-NEXT: llvm_arm_sve
// CHECK-NEXT: math
+// CHECK-NEXT: memref
// CHECK-NEXT: nvvm
// CHECK-NEXT: omp
// CHECK-NEXT: pdl
diff --git a/mlir/test/mlir-reduce/multiple-function.mlir b/mlir/test/mlir-reduce/multiple-function.mlir
index c668271571f1..64609c1abcee 100644
--- a/mlir/test/mlir-reduce/multiple-function.mlir
+++ b/mlir/test/mlir-reduce/multiple-function.mlir
@@ -1,7 +1,7 @@
// UNSUPPORTED: system-windows
// RUN: mlir-reduce %s -test %S/failure-test.sh -pass-test function-reducer | FileCheck %s
-// This input should be reduced by the pass pipeline so that only
-// the @simple5 function remains as this is the shortest function
+// This input should be reduced by the pass pipeline so that only
+// the @simple5 function remains as this is the shortest function
// containing the interesting behavior.
// CHECK-NOT: func @simple1() {
@@ -26,7 +26,7 @@ func @simple4(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
"test.crashOp"(%1, %arg2) : (memref<2xf32>, memref<2xf32>) -> ()
diff --git a/mlir/test/mlir-reduce/simple-test.mlir b/mlir/test/mlir-reduce/simple-test.mlir
index 4d274e1a39fe..2f5a2c0fd077 100644
--- a/mlir/test/mlir-reduce/simple-test.mlir
+++ b/mlir/test/mlir-reduce/simple-test.mlir
@@ -6,7 +6,7 @@ func @simple1(%arg0: i1, %arg1: memref<2xf32>, %arg2: memref<2xf32>) {
^bb1:
br ^bb3(%arg1 : memref<2xf32>)
^bb2:
- %0 = alloc() : memref<2xf32>
+ %0 = memref.alloc() : memref<2xf32>
br ^bb3(%0 : memref<2xf32>)
^bb3(%1: memref<2xf32>):
return
diff --git a/mlir/unittests/ExecutionEngine/Invoke.cpp b/mlir/unittests/ExecutionEngine/Invoke.cpp
index f970cc221985..853451a13d2b 100644
--- a/mlir/unittests/ExecutionEngine/Invoke.cpp
+++ b/mlir/unittests/ExecutionEngine/Invoke.cpp
@@ -104,7 +104,7 @@ TEST(NativeMemRefJit, ZeroRankMemref) {
std::string moduleStr = R"mlir(
func @zero_ranked(%arg0 : memref<f32>) attributes { llvm.emit_c_interface } {
%cst42 = constant 42.0 : f32
- store %cst42, %arg0[] : memref<f32>
+ memref.store %cst42, %arg0[] : memref<f32>
return
}
)mlir";
@@ -139,7 +139,7 @@ TEST(NativeMemRefJit, RankOneMemref) {
func @one_ranked(%arg0 : memref<?xf32>) attributes { llvm.emit_c_interface } {
%cst42 = constant 42.0 : f32
%cst5 = constant 5 : index
- store %cst42, %arg0[%cst5] : memref<?xf32>
+ memref.store %cst42, %arg0[%cst5] : memref<?xf32>
return
}
)mlir";
@@ -192,8 +192,8 @@ TEST(NativeMemRefJit, BasicMemref) {
%x = constant 2 : index
%y = constant 1 : index
%cst42 = constant 42.0 : f32
- store %cst42, %arg0[%y, %x] : memref<?x?xf32>
- store %cst42, %arg1[%x, %y] : memref<?x?xf32>
+ memref.store %cst42, %arg0[%y, %x] : memref<?x?xf32>
+ memref.store %cst42, %arg1[%x, %y] : memref<?x?xf32>
return
}
)mlir";
@@ -234,7 +234,7 @@ TEST(NativeMemRefJit, JITCallback) {
std::string moduleStr = R"mlir(
func private @callback(%arg0: memref<?x?xf32>, %coefficient: i32) attributes { llvm.emit_c_interface }
func @caller_for_callback(%arg0: memref<?x?xf32>, %coefficient: i32) attributes { llvm.emit_c_interface } {
- %unranked = memref_cast %arg0: memref<?x?xf32> to memref<*xf32>
+ %unranked = memref.cast %arg0: memref<?x?xf32> to memref<*xf32>
call @callback(%arg0, %coefficient) : (memref<?x?xf32>, i32) -> ()
return
}
More information about the Mlir-commits
mailing list