[all-commits] [llvm/llvm-project] d8fc27: [mlir][sparse] improved sparse runtime support lib...

[Aart Bik via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: d8fc27301d18f0935ba99ead7ac61aa6a53f16e4
      https://github.com/llvm/llvm-project/commit/d8fc27301d18f0935ba99ead7ac61aa6a53f16e4
  Author: Aart Bik <ajcbik at google.com>
  Date:   2021-01-16 (Sat, 16 Jan 2021)

  Changed paths:
    M mlir/include/mlir/ExecutionEngine/CRunnerUtils.h
    M mlir/integration_test/CMakeLists.txt
    A mlir/integration_test/Sparse/CPU/frostt-example.mlir
    M mlir/integration_test/Sparse/CPU/matrix-market-example.mlir
    A mlir/integration_test/data/test.tns
    M mlir/lib/ExecutionEngine/SparseUtils.cpp

  Log Message:
  -----------
  [mlir][sparse] improved sparse runtime support library

Added the ability to read (an extended version of) the FROSTT
file format, so that we can now read in sparse tensors of arbitrary
rank. Generalized the API to deal with more than two dimensions.

Also added the ability to sort the indices of sparse tensors
lexicographically. This is an important step towards supporting
auto gen of initialization code, since sparse storage formats
are easier to initialize if the indices are sorted. Since most
external formats don't enforce such properties, it is convenient
to have this ability in our runtime support library.

Lastly, the re-entrant problem of the original implementation
is fixed by passing an opaque object around (rather than having
a single static variable, ugh!).

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D94852