Codegenerierung für Datenverarbeitung

What it is about

You learn how code generation for data-processing programs works: from program representations (source code, IR, bytecode) through optimizations to the generation of machine code and execution in virtual machines. In the end you will be able to design appropriate representations, translate them efficiently, and take optimizations as well as sandboxing measures into account.

What you will be able to do

• Implement an efficient execution environment for data-processing programs
• Design program representations for high-level languages, database queries, and machine code
• Translate program representations into performant machine code
• Apply optimizations (e.g. SSA, LLVM-IR)
• Consider sandboxing and JIT optimizations

What the module consists of

• VorlesungDelivery of content through lectures with examples and discussions; serves as a foundation and motivation for independent engagement
• ÜbungImplementation of the conveyed concepts to deepen understanding

Teaching method

• Vortragto introduce concepts, with concrete examples and discussions
• Übungenfor practical implementation and deepening through implementing the techniques