This project seeks to treat the GPU as a standard hosted target by porting the LLVM C library to run on the GPU and achieve truly general purpose GPU programming. We show how LLVM/Clang can be used to compile regular, freestanding C++ to target the GPU as well as introduce a novel mechanism to invoke generic host services from the GPU. This allows us to compile a generic program and run it completely on the GPU, as well as provide missing system services to users of standard offloading languages such as OpenMP, CUDA, and HIP.