Google Git
Sign in
opensecura / 3p / openxla / iree / c3fae2f7443908e91a00eb65a4f330ad1f71e63f
commitc3fae2f7443908e91a00eb65a4f330ad1f71e63f[log] [tgz]
authorMax191 <44243577+Max191@users.noreply.github.com>Thu Oct 24 08:36:54 2024 -0700
committerGitHub <noreply@github.com>Thu Oct 24 11:36:54 2024 -0400
tree0e5c15c8e3d9833158f02cdad8cc4d80c8c0b6c3
parent4d20b82812951fd971f930f43167c11e46da1c25 [diff]
[LLVMGPU] Use forall workgroup distribution in TileAndFuse pipeline (#18565)

This switches the TileAndFuse pipeline to use scf.forall distribution.
Using scf.forall distribution also requires some changes to the pass
ordering in the TileAndFuse pipeline, which is also handled by this PR:
1. The main difference is that PackToIntrinsics happens before workgroup
distribution. Otherwise, collapse_shape ops can end up at the end of the
workgroup forall, and an extra buffer is created.
2. Pack decomposition is now staged, with packs/unpacks at the function
boundaries being decomposed early before workgroup decomposition, and
the rest being decomposed after reduction tiling as before. This
prevents unpacks being fused into the workgroup forall and causing the
same problem as in (1).
3. `ConcretizeMmaShapes` now runs before workgroup tiling as well,
so the resulting collapse_shape on the multi_mma op result can be
propagated to the function boundary before any tiling. This is also to
avoid the same problem as in (1).

The lowering configs on the MMA path have also changed, since they now
need to account for inner tile sizes of packing.

depends on https://github.com/iree-org/iree/pull/18852

Signed-off-by: Max Dawkins <max.dawkins@gmail.com>
5 files changed
tree: 0e5c15c8e3d9833158f02cdad8cc4d80c8c0b6c3
  1. .github/
  2. build_tools/
  3. compiler/
  4. docs/
  5. experimental/
  6. integrations/
  7. lib/
  8. llvm-external-projects/
  9. runtime/
  10. samples/
  11. tests/
  12. third_party/
  13. tools/
  14. .bazel_to_cmake.cfg.py
  15. .bazelignore
  16. .bazelrc
  17. .bazelversion
  18. .clang-format
  19. .git-blame-ignore-revs
  20. .gitattributes
  21. .gitignore
  22. .gitmodules
  23. .pre-commit-config.yaml
  24. .yamllint.yml
  25. AUTHORS
  26. BUILD.bazel
  27. CITATION.cff
  28. CMakeLists.txt
  29. configure_bazel.py
  30. CONTRIBUTING.md
  31. LICENSE
  32. MAINTAINERS.md
  33. README.md
  34. RELEASING.md
  35. 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.

IREE Discord Status pre-commit OpenSSF Best Practices

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

Release status

PackageRelease status
GitHub release (stable)GitHub Release
GitHub release (nightly)GitHub Release
Python iree-compilerPyPI version
Python iree-runtimePyPI version

Build status

CI PkgCI

Host platformBuild status
LinuxCI - Linux x64 clang
CI - Linux arm64 clang
macOSCI - macOS x64 clang
WindowsCI - Windows x64 MSVC

For the full list of workflows see https://iree.dev/developers/general/github-actions/.

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

Community meeting recordings: IREE YouTube channel

  • 2021-06-09: IREE Runtime Design Tech Talk (recording and slides)
  • 2020-08-20: IREE CodeGen: MLIR Open Design Meeting Presentation (recording and slides)
  • 2020-03-18: Interactive HAL IR Walkthrough (recording)
  • 2020-01-31: End-to-end MLIR Workflow in IREE: MLIR Open Design Meeting Presentation (recording and slides)

License

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