Search papers, labs, and topics across Lattice.
RepoOMP introduces a hybrid framework that enhances OpenMP parallelization of hotspots in mature code repositories by effectively managing dependency awareness. By constructing a Multi-granularity Attributes Performance graph (MAP) and utilizing a Structured Transformation Context (STC), RepoOMP successfully routes hotspots between deterministic rules and an LLM agent, leading to significant performance improvements. Evaluations show that RepoOMP achieves average speedups of up to 8.96x on specific benchmarks, with median speedup of 2.25x across real-world hotspots, demonstrating its effectiveness in optimizing parallelization without overwhelming the model with irrelevant data.
RepoOMP achieves up to 8.96x speedup in parallelizing hotspots while reducing agent-side token costs by nearly 68%.
OpenMP parallelization of hotspots in mature repositories remains difficult because loop safety and optimization payoff often depend on non-local evidence. Rule-based tools under-parallelize when legality is not locally provable, while agent-based approaches become unstable when retrieval misses decisive dependencies or includes irrelevant code. We present RepoOMP, a hybrid framework that recovers parallelization-relevant evidence before generation. RepoOMP builds a Multi-granularity Attributes Performance graph (MAP), routes hotspots between deterministic rules and an LLM agent, and constructs a Structured Transformation Context (STC) that exposes dependency facts without flooding the model with unrelated repository text. We evaluate RepoOMP on 951 profiled hotspots from NPB, BOTS, FFmpeg, NCNN, and GROMACS. Under compilation, workload-specific checks, and positive speedup, 372 hotspots are accepted, including 330 real-world repository hotspots. RepoOMP achieves average speedups of $8.23\times$ on NPB and $8.96\times$ on BOTS. For the nine detailed real-world kernels used in matched-backbone and robustness analyses, RepoOMP reaches a cross-backbone mean of $5.25\times$, improves speedup by 18--28\%, and reduces agent-side token cost by 47--68\% relative to the unstructured Claude Code baseline. Across 330 accepted real-world hotspots, median speedup is $2.25\times$. Overall, RepoOMP provides an evidence-guided workflow for hotspot parallelization in repository settings. The open-source repository is available at https://github.com/Qlalq/RepoOMP_Simplified.
