LoopUnrollPass

This project implements a custom LLVM compiler pass that performs loop unrolling optimizations at the Intermediate Representation (IR) level. The pass is built using LLVM's new PassManager infrastructure.

🛠 Features

• Identifies and unrolls simple loops in LLVM IR.
• Designed for experimentation and learning about compiler optimizations.
• Cleanly integrates with LLVM as a dynamically loadable pass.

📁 Project Structure

LoopUnrollPassProject/
├── CMakeLists.txt          # Build configuration using CMake
└── LoopUnrollPass.cpp      # Custom LLVM pass implementation

🔧 Build Instructions

Prerequisites

• LLVM and Clang installed (LLVM 12 or later recommended)
• CMake 3.10 or newer
• A C++ compiler (e.g., clang++ or g++)

Steps

git clone https://github.com/your-username/LoopUnrollPass.git
cd LoopUnrollPass
mkdir build
cd build
cmake ..
make

This will build the LoopUnrollPass as a shared object (.so or .dylib).

🧪 Running the Pass

You can apply the pass to an LLVM IR file:

opt -load-pass-plugin ./LoopUnrollPass.so -passes=loop-unroll -disable-output < input.ll

Note: Make sure the LLVM tools like opt and clang are in your PATH.

📚 References

• LLVM Writing an LLVM Pass
• LLVM Pass Infrastructure

🧑‍💻 Author

Damin Risho J V