LLVM Compiler Infrastructure

LLVM (Low Level Virtual Machine) is a free and open-source compiler infrastructure project that provides a collection of modular and reusable compiler and toolchain technologies. Originally developed by Chris Lattner as a research project at the University of Illinois at Urbana-Champaign (UIUC) starting in 2000, under the guidance of advisor Vikram Adve, LLVM has grown into one of the most important compiler projects in the world. Apple became the primary sponsor in 2005 (Lattner joined Apple), and LLVM is now maintained by the LLVM Foundation with major contributions from Apple, Google, ARM, NVIDIA, AMD, Intel, and Mozilla. Key features: LLVM IR (Intermediate Representation): a machine-independent, typed, RISC-like instruction set that serves as the common intermediary between frontends and backends. Frontends (like Clang) compile source code to LLVM IR; optimization passes transform the IR; backends generate target machine code (x86, ARM, RISC-V, WebAssembly, MIPS, PowerPC). Three-phase compiler design: frontend (parsing, semantic analysis, IR generation) to optimizer (IR-to-IR transformation passes) to backend (code generation). This modular design allows different languages to share optimizations and code generators. Clang: the C, C++, Objective-C, and Objective-C++ frontend. Known for fast compilation, excellent diagnostics (clear error messages with fix-it hints), and compatibility with GCC. Clang powers Xcode, Android NDK, and many IDEs. Optimization passes: over 100 optimization passes including dead code elimination, constant propagation, loop unrolling, inlining, vectorization, and interprocedural optimization. Passes are managed by the Pass Manager. LLVM Core: the optimizer and code generator supporting static compilation and JIT (Just-In-Time) compilation. LLDB: debugger. LLD: linker. MLIR: Multi-Level IR for building reusable and extensible compilers. C++. Apache-2.0 (with LLVM exceptions).