This is an experiment in using MLIR to automatically select the best Tensor Processing Primitives for linear algebra.
This repository contains an out-of-tree MLIR dialect as well as an opt-like tool to operate on that dialect and a runner-like tool to execute and benchmark MLIR kernels.
It also contains the recipes to use LIBXSMM from inside MLIR and can be used by other tools to drive our passes.
There's work in progress inside IREE to use this work on their pipeline.
This repository was previously called tpp-sandbox.
If you have a checkout with the previous name, please follow these instructions to rename the remote locally.
# Clone
git clone https://github.com/llvm/llvm-project.git
# Please check-out the llvm version in: build_tools/llvm_version.txt
mkdir llvm-project/build
pushd llvm-project/build
# This is important for the next step
export CUSTOM_LLVM_ROOT=`pwd`
echo $CUSTOM_LLVM_ROOT
export PATH=$CUSTOM_LLVM_ROOT/bin:$PATH
# Build
cmake -G Ninja ../llvm \
-DLLVM_ENABLE_PROJECTS=mlir \
-DLLVM_BUILD_EXAMPLES=ON \
-DLLVM_INSTALL_UTILS=ON \
-DLLVM_TARGETS_TO_BUILD="X86;NVPTX;AMDGPU" \
-DCMAKE_BUILD_TYPE=RelWithDebInfo \
-DLLVM_ENABLE_ASSERTIONS=ON \
-DCMAKE_C_COMPILER=clang \
-DCMAKE_CXX_COMPILER=clang++ \
-DLLVM_USE_LINKER=lld
ninja
popdThis setup assumes that you have built LLVM and MLIR in $CUSTOM_LLVM_ROOT as above.
# Clone
git clone https://github.com/plaidml/tpp-mlir.git
mkdir tpp-mlir/build
pushd tpp-mlir/build
# Build & test
cmake -G Ninja .. \
-DCMAKE_BUILD_TYPE=RelWithDebInfo \
-DMLIR_DIR=$CUSTOM_LLVM_ROOT/lib/cmake/mlir \
-DLLVM_EXTERNAL_LIT=$CUSTOM_LLVM_ROOT/bin/llvm-lit
cmake --build . --target check-tpp-opt
popdTo build the documentation from the TableGen description of the dialect operations, run:
cmake --build . --target mlir-docThis dialect template is made available under the Apache License 2.0 with LLVM Exceptions. See the LICENSE.txt file for more details.
-
Nice link for conv: https://d2l.ai/chapter_convolutional-neural-networks/padding-and-strides.html
-
in IREE: Codegen/Common/ConvertToDestinationPassingStylePass.cpp
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent