[all-commits] [llvm/llvm-project] a3672a: [mlir][sparse] avoid unnecessary tmp COO buffer an...

[PeimingLiu via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: a3672add76051ed27486eb163a8d0efbdf76f898
      https://github.com/llvm/llvm-project/commit/a3672add76051ed27486eb163a8d0efbdf76f898
  Author: Peiming Liu <peiming at google.com>
  Date:   2022-12-16 (Fri, 16 Dec 2022)

  Changed paths:
    M mlir/include/mlir/Dialect/SparseTensor/IR/SparseTensorAttrDefs.td
    M mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp
    M mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
    M mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorRewriting.cpp
    M mlir/test/Dialect/SparseTensor/sparse_concat_codegen.mlir

  Log Message:
  -----------
  [mlir][sparse] avoid unnecessary tmp COO buffer and convert when lowering ConcatentateOp.

When concat along dim 0, and all inputs/outputs are ordered with identity dimension ordering,
the concatenated coordinates will be yield in lexOrder, thus no need to sort.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D140228