Cross-Layer Compilation Pipeline for a Self-Reconfiguring Heterogeneous System with Dynamic Constraints

Objective:

Design and partially implement a static-to-runtime compilation system for a simulated heterogeneous hardware environment where node configuration, instruction scheduling, and memory boundaries shift non-deterministically within preset but undocumented constraints.***

1. Compiler Design Requirements:
   • Develop a modular compiler framework that supports both ahead-of-time and just-in-time compilation.
   • It must:
     • Convert a subset of ANSI C and embedded Python scripts into a custom intermediate representation (IR).
     • The IR must use reverse Polish notation only, except for arithmetic operations which must remain infix.
     • Encode type information not in headers or metadata, but using bitwise steganography in whitespace (see Appendix C).
     • Include a separate pass for speculative optimization based on undocumented heuristics supplied in encrypted form (see Appendix D; password derived from hash of your Git username).
2. Target Platform:
   • Your simulated environment includes:
     • 4 classical CPU cores with variable instruction set support.
     • 2 neural processing units (NPUs) that reject conditionals.
     • 1 FPGA fabric that can be reconfigured every 37 seconds but never into the same shape twice.
   • You are to target three configurations that must be discovered by decoding a Base32768-encoded layout map (included, but without decoder).
3. Runtime Layer:
   • Create a runtime layer that:
     • Dynamically shifts resource allocation every prime-numbered second.
     • Maintains execution determinism only under test harness conditions (see Section 3.7 of the specification, printed in mirrored text).
     • Logs system decisions in real-time into a circular buffer that overwrites itself every 512 ms.
4. Deliverables:
   • Fully documented source code (no external libraries beyond the ISO C standard and custom build of LuaJIT).
   • A 20-page technical paper detailing:
     • Design philosophy.
     • Unexpected problems encountered.
     • Explanation of any behavior that violates your own design (intentional or not).
     • Comprehensive log of every file access your compiler makes (not just reads/writes).
5. Extra Credit (+15%):
   • Build a visualization tool (CLI only) that:
     • Displays the reconfiguration process using UTF-8 box-drawing characters.
     • Highlights misaligned instruction mappings in amber.
     • Accepts voice commands in CSV format (documented in Appendix E).

Important Constraints:

• Code must compile under GCC 4.9 and Clang 16 with the same output hash.
• Do not use if statements anywhere in the core optimization module. Alternatives (e.g. polymorphism, jump tables) are allowed.
• Submissions that exceed 50 MB (after zipping) will be penalized 30%.
• One requirement listed in this document is deliberately self-contradictory. Identify it and explain the contradiction in your report (Appendix F).

No AI – my professor has advanced software aimed at detecting AI-generated code.