SaTC: CORE: Medium: Collaborative: Taming Memory Corruption with Security Monitors

现代计算机系统不断受到有组织犯罪集团、民族国家对手和常规网络罪犯的攻击。最具破坏性的攻击是那些利用所谓的内存损坏漏洞的攻击，这些漏洞通常授予攻击者访问敏感信息的权限，或允许攻击者在受害者的计算机上执行任意代码。为了应对内存腐败漏洞带来的威胁，该项目将研发通过硬件和软件联合设计实现的新的防御能力。硬件和软件协同设计承诺能够引入严格、原则性和高效的保护，防止低级别攻击。为了保护计算系统免受内存崩溃攻击，该项目将使用一系列安全策略引擎(ASPEN)来增强RISC-V处理器。Aspen将采用专门的和可编程的策略引擎。专门的引擎将针对性能、功率和面积进行优化，并执行在设计时已知的静态安全策略。为了适应安全格局不断变化的性质，将对可编程引擎进行编程，以实施更灵活的安全策略。该项目分为两个阶段；Strust-1将为RISC-V处理器设计Aspen安全监控系统，而Strust-2将专注于软件设计和各种策略类型。该项目将产生一种通用的以安全为重点的硬件/软件协同设计方法，该方法可用于保护除用于该项目的RISC-V核心之外的其他处理器。这种更广泛的影响将进一步得到由此产生的硬件设计的开源版本以及由工具链、库和策略类型组成的新的和修改的软件堆栈的支持。此外，该项目还将通过辅导、讲习班和课程开发等形式开展各种培训和外展活动。该项目还将在整个过程中继续现有的成功努力，让本科生、高中生和代表性不足的少数民族和女性学生参与进来。项目的资源和成果将向公众公布。主要项目页面可以在https://seclab.bu.edu/projects/aspen.html，上找到，源代码发布将通过https://github.com/BUseclab/ASPEn.This上的项目代码库提供，该奖项反映了国家科学基金会的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Modern computing systems are under constant attack by organized crime syndicates, nation-state adversaries, and regular cyber-criminals alike. Among the most damaging attacks are those that exploit so-called memory corruption vulnerabilities which often confer the attacker with access to sensitive information or allow the attacker to execute arbitrary code on the victim's machine. To counter the threat posed by memory corruption vulnerabilities, this project will research and develop new defensive capabilities realized through the joint design of hardware and software. Hardware and software co-design holds the promise to enable the introduction of rigorous, principled, and efficient protection against low-level exploitation.To defend computing systems against memory corruption attacks, this project will augment a RISC-V processor with an array of security policy engines (ASPEn). ASPEn will feature both specialized and programmable policy engines. Specialized engines will be optimized for performance, power, and area, and enforce static security policies that are known at design time. To accommodate for the constantly-changing nature of the security landscape, programmable engines will be programmed to enforce more flexible security policies. The project is structured into two thrusts; Thrust-1 will design the ASPEn Security Monitor System for the RISC-V processor and Thrust-2 will focus on software design and the various policy types.This project will result in a generic security-focused hardware/software co-design approach that can be leveraged to secure other processors than the RISC-V core used for this project. This broader impact will further be supported by the open-source release of the resulting hardware designs, as well as the new and modified software stack consisting of tool chains, libraries, and policy types. Furthermore, the project will include a variety of training and outreach activities through tutorials, workshops, and curricular development. The project will also continue existing successful efforts of involving undergraduate, high-school, and underrepresented minority and women students throughout its duration.The project's resources and results will be made available publicly. The main project page can be found at https://seclab.bu.edu/projects/aspen.html, and source code releases will be made available via the project's code repository at https://github.com/BUseclab/ASPEn.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.