[all-commits] [llvm/llvm-project] 5106a8: [MLIR][Shape] Lower `shape_of` to `dynamic_tensor_...

[frgossen via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 5106a8b8f8d0d3dd6c3fc0554f05402d8d9177ef
      https://github.com/llvm/llvm-project/commit/5106a8b8f8d0d3dd6c3fc0554f05402d8d9177ef
  Author: Frederik Gossen <frgossen at google.com>
  Date:   2020-09-09 (Wed, 09 Sep 2020)

  Changed paths:
    M mlir/include/mlir/Dialect/StandardOps/IR/Ops.td
    M mlir/lib/Conversion/ShapeToStandard/ShapeToStandard.cpp
    M mlir/lib/Dialect/StandardOps/IR/Ops.cpp
    M mlir/test/Conversion/ShapeToStandard/shape-to-standard.mlir

  Log Message:
  -----------
  [MLIR][Shape] Lower `shape_of` to `dynamic_tensor_from_elements`

Take advantage of the new `dynamic_tensor_from_elements` operation in `std`.
Instead of stack-allocated memory, we can now lower directly to a single `std`
operation.

Differential Revision: https://reviews.llvm.org/D86935