| commit | af387d39d2dd553d03943c6a698cc15b6a8fc483 | [log] [tgz] |
|---|---|---|
| author | Han-Chung Wang <hanhan0912@gmail.com> | Mon Feb 05 09:55:03 2024 -0800 |
| committer | GitHub <noreply@github.com> | Mon Feb 05 09:55:03 2024 -0800 |
| tree | b0cbe6dd1508c8c1214eb98849077b69d435c3ac | |
| parent | dd5f707446cb0033bf7c96daa5997ac4e8aaaf7c [diff] |
Bump LLVM to llvm/llvm-project@730f498c961f (#16286) - Revert llvm/llvm-project@2800448 locally because it crashes many the RISC-V compilation. After raising it to the author, they reverted it upstream. - Revert llvm/llvm-project@66347e51 locally because it causes regression. It is already reverted in upstream. - The cuda matmul regression appears in the LLVM NVPTX backend. We are generating the same LLVM bitcode before [converting to ISA](https://github.com/openxla/iree/blob/861815c/compiler/plugins/target/CUDA/CUDATarget.cpp#L600); CUDA perf is not our priority; we don't have folks super familiar with that layer and it is quite involved fixing there. So we are fine taking the hit for now and revisit later. (maybe later it'd be fixed in upstream..) Additional fixes: - Apply fixes for https://github.com/llvm/llvm-project/commit/ce7cc723b9d51ad9c741bbaeecb5e008b2b81338 - Apply fixes for https://github.com/llvm/llvm-project/commit/b91bba89edfb25d011e1f2366cda5dec605c87f6 --------- Co-authored-by: Diego Caballero <diegocaballero@google.com> Co-authored-by: Lei Zhang <antiagainst@gmail.com>
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 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!
See our website for more information.
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.