HintPilot: LLM-based Compiler Hint Synthesis for Code Optimization

TL;DR

HintPilot leverages large language models to synthesize compiler hints that guide optimizations, achieving significant speedups while maintaining correctness.

Contribution

It introduces a retrieval-augmented, profiling-guided method for synthesizing compiler hints using LLMs, bridging reasoning and traditional compiler infrastructure.

Findings

Up to 6.88x speedup over -Ofast on benchmarks

Preserves program correctness with synthesized hints

Employs retrieval-augmented synthesis and iterative refinement

Abstract

Code optimization remains a core objective in software development, yet modern compilers struggle to navigate the enormous optimization spaces. While recent research has looked into employing large language models (LLMs) to optimize source code directly, these techniques can introduce semantic errors and miss fine-grained compiler-level optimization opportunities. We present HintPilot, which bridges LLM-based reasoning with traditional compiler infrastructures via synthesizing compiler hints, annotations that steer compiler behavior. HintPilot employs retrieval-augmented synthesis over compiler documentation and applies profiling-guided iterative refinement to synthesize semantics-preserving and effective hints. Upon PolyBench and HumanEval-CPP benchmarks, HintPilot achieves up to 6.88x geometric mean speedup over -Ofast while preserving program correctness.