UT professor Shravan Narayan, in collaboration with researchers from UC San Diego, Purdue, Google, Intel, Fastly, and Rivos, were finalists for the Intel Hardware Security Academic Award.
Their publication, titled Going Beyond the Limits of SFI: Flexible and Secure Hardware-Assisted In-Process Isolation with HFI, presents HFI, a new extension for in-process isolation to current processors.
Many applications such as browsers use isolation mechanisms in a variety of places, for example, to ensure attackers cannot target bugs in third-party libraries to compromise the browser. Hardware support for isolation would increase security and reduce runtime overheads of enforcing isolation in applications.
HFI supports isolation by making small changes to existing processors to support direct isolation of native code, as well as to speed up existing software-based fault isolation/sandboxing systems.
“HFI provides a simple and clean specification for CPUs to support isolation for application components. If CPUs adopt this extension, this would be a great opportunity to drive the security of the entire software ecosystem forward; HFI encourages developers to develop modular applications where bugs in one part of an application do not automatically translate into a bug in the entire application”
The Intel Hardware Security Academic Award is awarded to recognize advancements in “solutions, tools, and methodologies, which enhance the industry’s ability to deliver more secure and trustworthy foundational technologies,” according to the award’s website. The company typically awards only three published papers each year.
The award is not the first achievement for the publication. Researchers previously won a distinguished paper award at an ACM International Conference on Architectural Support for Programming Languages and Operating Systems conference held in Fall 2023. The update was also selected as one of the top computer architecture papers of the year in an IEEE Micro Top Picks in Computer Architecture.