P2Patch: Exploit-Guided Vulnerability Repair

📄 arXiv: 2610.20245v1 📥 PDF

作者: Mohsen Iranmanesh, Sina Moradi Sabet, Sina Marefat, Ali Javidi Ghasr, Maxime Newman Nereyabagabo, Kousha Amouzesh, Jaivir Malhi, Vinci Velasco, Alex Chung, Mohammad A. Tayebi

分类: cs.CR, cs.AI, cs.SE

发布日期: 2026-10-14

备注: 25 pages, 6 figures, 13 tables. M. Iranmanesh and S. Moradi Sabet contributed equally. Code: https://github.com/P2Patch/P2patch

💡 一句话要点

提出P2Patch以解决软件漏洞修复中的验证不足问题

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 自动化修复 漏洞检测 对抗性测试 程序安全 软件工程 机器学习

📋 核心要点

  1. 现有的自动化程序修复方法在验证补丁有效性时，通常依赖于构建补丁时的静态证据，容易忽视变种漏洞的存在。
  2. P2Patch通过生成对抗的方式，将每个补丁视为新的攻击目标，确保补丁能够抵御所有已知的漏洞证明。
  3. 实验结果显示，P2Patch在修复覆盖率上显著优于现有方法，提升了从78.1%到84.7%的修复效果。

📝 摘要（中文）

随着大型语言模型被广泛应用于自动化程序修复，现有系统通常仅根据构建补丁时使用的证据进行验证，导致补丁可能在特定情况下看似成功，但仍然存在变种漏洞。本文提出P2Patch，一个生成对抗修复系统，将每个补丁视为新的攻击目标。通过构建漏洞证明和进行补丁修复，P2Patch确保最终补丁能够阻止在修复过程中积累的所有漏洞证明。实验结果表明，P2Patch在101个C/C++和Java的CVE案例中，成功发现41%的新漏洞，并显著提高了修复覆盖率。

🔬 方法详解

问题定义：本文旨在解决现有自动化程序修复方法在验证补丁有效性时的不足，尤其是它们通常只依据构建补丁时的证据，导致补丁可能未能阻止变种漏洞的攻击。

核心思路：P2Patch的核心思想是将每个补丁视为新的攻击目标，通过生成对抗的方式，确保补丁能够抵御所有已知的漏洞证明，从而提高修复的全面性和有效性。

技术框架：P2Patch的整体架构包括三个主要模块：漏洞发现模块、补丁修复模块和攻击验证模块。首先，漏洞发现模块生成标准化的漏洞报告，接着补丁修复模块进行程序修复，最后攻击验证模块对修复后的程序进行攻击测试。

关键创新：P2Patch的主要创新在于引入了对抗性测试机制，通过独立的攻击验证，确保补丁能够有效阻止所有已知的漏洞证明，这与传统方法的单一验证方式形成了鲜明对比。

关键设计：在技术细节上，P2Patch采用了特定的损失函数来评估补丁的有效性，并设计了多层次的网络结构以支持生成对抗训练，从而提高了漏洞修复的准确性和全面性。

🖼️ 关键图片

fig_0
fig_1

📊 实验亮点

实验结果显示，P2Patch在101个C/C++和Java的CVE案例中，成功发现41%的新漏洞，显著提高了修复覆盖率，从78.1%提升至84.7%，并且在相同的验证条件下，认证的攻击路径数量也超过了现有的最先进修复系统。

🎯 应用场景

P2Patch的研究成果在软件安全领域具有广泛的应用潜力，尤其是在自动化漏洞修复和安全审计方面。通过提高补丁的有效性，P2Patch能够帮助开发者更好地保护软件系统免受潜在攻击，提升软件的安全性和可靠性。未来，该方法还可以扩展到其他编程语言和系统中，进一步推动自动化修复技术的发展。

📄 摘要（原文）

Large language models are increasingly deployed as autonomous agents for automated program repair, patching software vulnerabilities through multi-step reasoning and tool use. Yet these systems typically validate a patch against the same evidence used to construct it -- a static alert, sanitizer trace, published exploit, or test suite. A patch that blocks that one piece of evidence can therefore look successful while remaining vulnerable to the same vulnerability with some variation. We present P2Patch, a generative adversarial repair system that treats each patch as a new attack target. Starting from a normalized vulnerability finding, an exploitation agent constructs a reproducing proof of vulnerability (PoV), a patching agent repairs the program, and the exploitation agent then attacks the patched program for a distinct same-vulnerability exploit. Every successful exploit becomes a permanent repair requirement, and the final patch must block all PoVs accumulated during the patching process, with construction-time PoVs kept strictly separate from held-out evaluation oracles. Across 101 C/C++ and Java CVE subjects, this adversarial loop surfaces a distinct confirmed exploit in 41% of runs after the patch has already blocked its original construction-time PoV. Against single-pass repair from the same finding using Claude Code, P2Patch raises held-out fix-closed coverage from 78.1% to 84.7% and beyond-upstream coverage from 42.1% to 62.5%. Under the same held-out oracles, it also certifies more attack paths as blocked than recent state-of-the-art agentic repair systems on overlapping subjects. These results show that adversarial testing and independent security oracles expose repair gaps that one-shot validation misses.