[all-commits] [llvm/llvm-project] 6c3c5f: [mlir][memref] Improve type inference for rank-red...

Matthias Springer via All-commits all-commits at lists.llvm.org
Tue Jul 5 07:49:29 PDT 2022


  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 6c3c5f8069d97e635b1887a6f9ac410391b89fae
      https://github.com/llvm/llvm-project/commit/6c3c5f8069d97e635b1887a6f9ac410391b89fae
  Author: Matthias Springer <springerm at google.com>
  Date:   2022-07-05 (Tue, 05 Jul 2022)

  Changed paths:
    M mlir/include/mlir/Dialect/MemRef/IR/MemRefOps.td
    M mlir/lib/Dialect/Bufferization/Transforms/AllocTensorElimination.cpp
    M mlir/lib/Dialect/MemRef/IR/MemRefOps.cpp
    M mlir/lib/Dialect/MemRef/Transforms/MultiBuffer.cpp
    M mlir/lib/Dialect/Tensor/Transforms/BufferizableOpInterfaceImpl.cpp
    M mlir/lib/Dialect/Vector/Transforms/VectorTransferOpTransforms.cpp
    M mlir/test/Dialect/Tensor/bufferize.mlir
    M mlir/test/Dialect/Tensor/one-shot-bufferize.mlir
    M mlir/unittests/Dialect/MemRef/InferShapeTest.cpp

  Log Message:
  -----------
  [mlir][memref] Improve type inference for rank-reducing subviews

The result shape of a rank-reducing subview cannot be inferred in the general case. Just the result rank is not enough. The only thing that we can infer is the layout map.

This change also improves the bufferization patterns of tensor.extract_slice and tensor.insert_slice to fully support rank-reducing operations.

Differential Revision: https://reviews.llvm.org/D129144




More information about the All-commits mailing list