commit	7b76af21a5ce8cd6fdac24e25a032c8d212b9282	[log] [tgz]
author	MaheshRavishankar <1663364+MaheshRavishankar@users.noreply.github.com>	Sat Sep 11 08:48:00 2021 -0700
committer	GitHub <noreply@github.com>	Sat Sep 11 08:48:00 2021 -0700
tree	9ac2142e10b12d3d8125df73ad18acf37e421705
parent	475045245a9ac0f913e368bff671c86cb9d36af9 [diff]

Plumb through all `TiledOpInterface` ops. (#7006)

This changes all places that use TiledOpInterface to using the
interface directly and not op-specific pattern. The external models
that implement the TiledOpInterface are also registered during
startup time.

This also has some changes to how configurations are set. For
tensor.insert_slice the lowering results in no compute ops.
So this needs to be handled. (Some of this needs some cleanup after
use of workgroup marker is dropped).

This leads to another issue on GPU. tensor.insert_slice is lowered
to a copy on bufferization. If the configuration is determined before
bufferization, this copy is missed. For now, its OK to change the
bufferization to run before configuration selection, but a better solution
is needed here.
Note: This was initially pulled together for use with
tensor.insert_slice which implements the TiledOpInterface as an
external model. This runs into some upstream bugs, so the interface is
not attached to the op as it is supposed to be till those are fixed.

35 files changed

tree: 9ac2142e10b12d3d8125df73ad18acf37e421705

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.

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

GitHub issues: Feature requests, bugs, and other work tracking
IREE Discord server: Daily development discussions with the core team and collaborators
iree-discuss email list: Announcements, general and low-priority discussion

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.

Build Status

CI System	Build System	Platform	Architecture	Configuration / Component
Kokoro	Bazel	Linux	x86-64
Kokoro	CMake & Bazel	Linux	x86-64 (swiftshader)	Integrations
Kokoro	CMake & Bazel	Linux	x86-64 (turing)	Integrations
Kokoro	CMake	Linux	x86-64 (swiftshader)
Kokoro	CMake	Linux	x86-64 (swiftshader)	asan
Kokoro	CMake	Linux	x86-64 (turing)
Kokoro	CMake	Android	arm64-v8a	Runtime (build only)
Kokoro	CMake	Bare Metal	risc-v-32	Runtime
Kokoro	CMake	Linux	risc-v-64	Runtime
Buildkite	CMake	Android	arm64-v8a	Runtime
BuildKite	CMake	Android	arm64-v8a	Runtime Benchmarks
BuildKite	CMake	Linux	x86-64	Tracing + Standalone Runtime

Presentations and Talks

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.