CWEEP: A Lexical Static Analysis Framework for CWE Early Prevention
2026-07-31 • Cryptography and Security
Cryptography and Security
AI summaryⓘ
The authors present CWEEP, a tool that helps find security problems in hardware design code without needing detailed security rules. It can spot the exact place in the code where issues might be and even suggests how to fix them. The tool works early in the development process and was tested on different faulty hardware designs. It showed better accuracy in warnings compared to an older tool, correctly identifying issues about 61% of the time.
hardware securityregister transfer level (RTL)static analysissecurity verificationSoC (System on Chip)bug detectioncode repairhardware designvulnerability detection
Authors
Bryan Kwan, Benjamin Tan
Abstract
As the hardware layer becomes a focus point for attackers, the need for improved hardware security verification techniques is more important than ever. State-of-the-art security verification techniques require significant manual effort from individuals with security expertise. Furthermore, there is no standard method to locate where the fault lies within the register transfer level (RTL) code. This paper presents CWEEP, a static analysis framework for detecting security weaknesses in RTL. CWEEP does not require a detailed security specification, so it can be used in the early stages of RTL development while properties are still under construction. Furthermore, CWEEP can identify the exact location in the RTL where the potential vulnerability resides and supports automatic code repair suggestions when applicable. Using datasets from the literature, we evaluate the performance of CWEEP on a set of two SoC designs with manually inserted bugs and on a large language model generated dataset, consisting of 3874 buggy modules. We find that CWEEP issues a correct warning up to 60.8% of the time. In contrast, the tool from a previous work issued a correct warning 17.5% of the time for the same dataset.