Clang-Chimera is a framework to easily develop operators to automatically modify C/C++ codes through code patterns identification.
See the guide!
In order to build clang-chimera from source you will need an installation of LLVM/Clang 3.9.1 compiled with the following cmake flags:
-DLLVM_ENABLE_CXX1Y=true
-DLLVM_ENABLE_RTTI=ON
-DLLVM_ENABLE_FFI=ON
Clang-chimera will expect LLVM/Clang binaries to be installed in /usr/bin.
Use this section to understand how to install LLVM/Clang 3.9.1 correctly. Note that you will need these packages to be installed:
- cmake
- git
- ninja-build
- build-essential (gcc, g++ and their libraries)
- zlib
- ffi
- edit
- ncurses
- boost
Considering Ubuntu 18.04 as development platform, use the following command to install the previous dependencies:
# apt-get install cmake ninja-build build-essential zlib1g-dev libffi-dev libedit-dev libncurses5-dev libboost-dev
Now, you can download LLVM/Clang 3.9.1 from source and build it.
$ git clone -b release_39 https://github.com/llvm-mirror/llvm.git llvm
$ cd llvm/tools
$ git clone -b release_39 https://github.com/llvm-mirror/clang.git clang
$ mkdir ../build && cd ../build
$ CC=gcc CXX=g++ cmake -DLLVM_ENABLE_CXX1Y=true -DLLVM_ENABLE_RTTI=ON -DLLVM_TARGET_ARCH="X86" -DLLVM_TARGETS_TO_BUILD="X86" -DLLVM_ENABLE_FFI=ON -DCMAKE_INSTALL_PREFIX=/usr -DCMAKE_BUILD_TYPE=Release -G Ninja ..
$ ninja && sudo ninja install
When the installation process ends, try to run:
$ llvm-config --version
$ clang --version
First of all you need to solve a LLVM bug, involving the JIT-compiler: the file /usr/lib/cmake/llvm/LLVM-Config.cmake tries to find the JIT-compiler library with the name jit, but it fails because the actual name of this component is mcjit. So it is necessary to edit LLVM-Config.cmake in this way:
# sed -i 's/list(APPEND link_components "jit")/list(APPEND link_components "mcjit")/g' /usr/lib/cmake/llvm/LLVM-Config.cmake
Before building Clang-Chimera, you need to configure it with CMake. You need to pass to CMake configurator the list of LLVM's available components. You can run the script run_cmake or configure Clang-Chimera manually.
Now you can build Clang-Chimera source. It is recomended to use Ninja as building tool, since it takes advantage of the multicore parallelism by default.
$ cd ~/clang-chimera
$ ./run_cmake
$ cd build
$ ninja
$ sudo ninja install
At the end of the process you will find clang-chimera in /usr/bin. Try run:
$ clang-chimera -version
If you don't want to build LLVM/Clang and Clang-Chimera from scratch, a ready-to-use solution is provided through a Docker Container. Please refer to IIDEAA Docker repository for further details.
Github is for social coding: if you want to write code, I encourage contributions through pull requests from forks of this repository.