Introduction to Code Generation
You'll learn: How compilers turn your program's intermediate representation into actual machine instructions or bytecode that a computer can execute.
What Is Code Generation?
Code generation is the final translation step in a compiler. After parsing your source code, building an abstract syntax tree, and converting it into an intermediate representation (IR), the compiler must produce target code — the actual instructions that will run on hardware or a virtual machine.
Think of it like translating a recipe (your IR) into specific cooking actions. The IR says "combine ingredients," but code generation produces the exact movements: "pick up bowl with left hand, pour 2 cups flour, stir clockwise 10 times."
From IR to Executable Code
The code generator takes structured IR — whether that's three-address code, SSA form, or another representation you've learned — and maps it to:
- Machine code: Native CPU instructions (x86, ARM, etc.)
- Bytecode: Virtual machine instructions (like Java bytecode or Python bytecode)
The generator must handle:
- Instruction selection: Choosing which target instructions implement each IR operation
- Register allocation: Deciding which variables live in fast CPU registers vs slower memory
- Instruction ordering: Arranging operations to follow target platform rules
For example, an IR instruction like t1 = a + b might become the x86 assembly ADD EAX, EBX or the JVM bytecode iadd, depending on your target platform.
Why It Matters
Code generation bridges the gap between high-level program meaning (captured in IR) and low-level execution. It's where compiler optimizations meet hardware reality, and where your abstract computations become concrete instructions.
Key Takeaway: Code generation transforms intermediate representation into executable target instructions, translating platform-independent program structure into platform-specific machine code or bytecode.