Google Git
Sign in
opensecura / 3p / openxla / iree / 3e9602e6eb19ee10ec1c64c645ab33a98b4f286f
commit3e9602e6eb19ee10ec1c64c645ab33a98b4f286f[log] [tgz]
authorBen Vanik <ben.vanik@gmail.com>Mon Oct 24 18:39:36 2022 -0700
committerGitHub <noreply@github.com>Mon Oct 24 18:39:36 2022 -0700
treef35b9115013f473edfe15c6da7b9ca766e27da93
parentc1348da05116f84fa0921b8a18512dda982aeb5f [diff]
parent2d79c208ba90b658ae84edb5fa79bb27fc83bb67 [diff]
Adding HAL profiling API and RenderDoc support for Vulkan. (#10893)

TLDR: configure with `-DIREE_ENABLE_RENDERDOC_PROFILING=ON`, pass the
`--device_profiling_mode=queue` flag to the IREE tools, and launch the
tools from the RenderDoc UI in order to get a capture (or use
`renderdoccmd capture`):

![image](https://user-images.githubusercontent.com/75337/197648585-b34bd661-cfd1-4fbb-a6f9-2b73bec81b6a.png)

Things are set up to allow for other profiling modes in the future but
how best to integrate those is TBD. We can figure out how to scale this
with other tooling and on other backends but the rough shape of the API
should be compatible with the various backend APIs we target
(D3D/Metal/CUDA/Vulkan/perf/etc). Note that because RenderDoc will also
capture D3D the cmake flag is generic but both the Vulkan and D3D HAL
implementations will need to load it themselves (no real code worth
sharing as D3D naturally only needs the Windows API query path).

Docs have notes that I've verified on Windows. Someone looking to use
this on Android will need to figure that out and can add what they find.

Fixes #45. Forty five. Wow.
tree: f35b9115013f473edfe15c6da7b9ca766e27da93
  1. .github/
  2. benchmarks/
  3. build_tools/
  4. compiler/
  5. docs/
  6. experimental/
  7. integrations/
  8. llvm-external-projects/
  9. runtime/
  10. samples/
  11. tests/
  12. third_party/
  13. tools/
  14. .bazelignore
  15. .bazelrc
  16. .bazelversion
  17. .clang-format
  18. .gitignore
  19. .gitmodules
  20. .pylintrc
  21. .style.yapf
  22. .yamllint.yml
  23. AUTHORS
  24. BUILD.bazel
  25. CITATION.cff
  26. CMakeLists.txt
  27. configure_bazel.py
  28. CONTRIBUTING.md
  29. LICENSE
  30. README.md
  31. 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

  • 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.