Aakash Tyagi
Aakash Tyagi is a researcher listed as affiliated with Texas A&M University, USA, as indicated in the author byline of the USENIX Security 2022 paper on TheHuzz.
Notable Work
TheHuzz (USENIX Security 2022)
Aakash Tyagi is a co-author of:
TheHuzz: Instruction Fuzzing of Processors Using Golden-Reference Models for Finding Software-Exploitable Vulnerabilities
Rahul Kande, Addison Crump, Garrett Persyn, Patrick Jauernig, Ahmad-Reza Sadeghi, Aakash Tyagi, and Jeyavijayan Rajendran
31st USENIX Security Symposium (USENIX Security 22), Boston, MA, August 2022, pages 3219–3236.
The paper introduces TheHuzz, a hardware fuzzer that:
- Operates on commonly used hardware description languages (HDLs) such as Verilog and VHDL, addressing a key limitation of prior hardware fuzzers.
- Analyzes intrinsic behaviors of hardware designs and defines coverage metrics that model behaviors such as signal transitions and floating wires.
- Generates assembly-level instructions to drive the desired coverage values, thereby uncovering hardware bugs that are exploitable from software.
Evaluation and Results (from the paper)
- Evaluated on four popular open-source processors.
- Achieved 1.98× the speed of the industry-standard random regression approach and 3.33× the speed of the state-of-the-art hardware fuzzer DifuzzRTL.
- Detected 11 bugs across the four processors, of which 8 were new bugs, with demonstrated exploits.
- Outperformed the industrial formal verification tool Cadence JasperGold, which had previously detected 48% of bugs in a crowd-sourced hardware CTF where manual simulation detected 61%.
Collaborators
From the paper's author list and affiliations:
- Texas A&M University, USA: Rahul Kande, Addison Crump, Garrett Persyn, Aakash Tyagi, Jeyavijayan Rajendran
- Technische Universität Darmstadt, Germany: Patrick Jauernig, Ahmad-Reza Sadeghi
Contact email listed for Tyagi in the paper: tyagi@tamu.edu.