Google Git
Sign in
opensecura / 3p / openxla / iree / c27ed41cb17e8e8ce282cf9b838b0af1dd48c540
commitc27ed41cb17e8e8ce282cf9b838b0af1dd48c540[log] [tgz]
authorMax191 <44243577+Max191@users.noreply.github.com>Thu Jan 11 14:47:16 2024 -0800
committerGitHub <noreply@github.com>Thu Jan 11 17:47:16 2024 -0500
tree7570defef42773de6d8afcdbf321c5a27a13bfbd
parent17e9529e8340b007a61d62ca80a31afc2829ec0c [diff]
[GlobalOpt][CPU] Move to using indexing maps for data tiling encodings instead of named op enums (#15984)

This PR adds a `user_indexing_maps` attribute to `linalg_ext.encoding`, and
uses this attribute in MaterializeEncoding in place of the case-by-case
enums for matmul and batch_matmul. This will enable data tiling on
transposed matmul cases like `linalg.matmul_transpose_a`, and is a step
towards data-tiling of `linalg.generic` contraction ops.

In SetEncoding, the `user_indexing_maps` attribute is set, containing the
indexing maps of the LHS, RHS, and RESULT of the op to be data-tiled.
The case-by-case checks are removed by this PR, and transposed
`linalg::ContractionOpInterface` ops are allowed to get encodings. The
`MATMUL` and `BATCH_MATMUL` user encodings are kept for now, but will
eventually be removed.

In MaterializeEncoding, the `user_indexing_maps` are used to infer the
contraction dimensions (M, N, K, Batch) of the inputs, and a
`tensor.pack` op is created with appropriate `inner_dims_pos` and
`outer_dims_perm` to transpose and pack the input into the canonical
`linalg.mmt4d` input shapes.
8 files changed
tree: 7570defef42773de6d8afcdbf321c5a27a13bfbd
  1. .devcontainer/
  2. .github/
  3. build_tools/
  4. compiler/
  5. docs/
  6. experimental/
  7. integrations/
  8. lib/
  9. llvm-external-projects/
  10. runtime/
  11. samples/
  12. tests/
  13. third_party/
  14. tools/
  15. .bazel_to_cmake.cfg.py
  16. .bazelignore
  17. .bazelrc
  18. .bazelversion
  19. .clang-format
  20. .dockerignore
  21. .git-blame-ignore-revs
  22. .gitignore
  23. .gitmodules
  24. .yamllint.yml
  25. AUTHORS
  26. BUILD.bazel
  27. CITATION.cff
  28. CMakeLists.txt
  29. configure_bazel.py
  30. CONTRIBUTING.md
  31. LICENSE
  32. README.md
  33. WORKSPACE
README.md

IREE: Intermediate Representation Execution Environment

IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler and runtime that lowers Machine Learning (ML) models to a unified IR that scales up to meet the needs of the datacenter and down to satisfy the constraints and special considerations of mobile and edge deployments.

See our website for project details, user guides, and instructions on building from source.

CI Status

Project Status

IREE is still in its early phase. We have settled down on the overarching infrastructure and are actively improving various software components as well as project logistics. It is still quite far from ready for everyday use and is made available without any support at the moment. With that said, we welcome any kind of feedback on any communication channels!

Communication Channels

Related Project Channels

  • MLIR topic within LLVM Discourse: IREE is enabled by and heavily relies on MLIR. IREE sometimes is referred to in certain MLIR discussions. Useful if you are also interested in MLIR evolution.

Architecture Overview

IREE Architecture IREE Architecture

See our website for more information.

Presentations and Talks

License

IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.