FVDCLS: Functional Verification of RISCV Based Dual-Core Lockstep Feature Using Fault Injection Mechanism

This paper presents the framework to verify RISCV based Dual-Core Lockstep (DCLS), a feature that enhances the fault tolerance in processors. The DCLS feature, when two identical processors run in parallel, effectively detects the error and rolls back the core operations to a previously known state. The purpose-built fault injection mechanism is developed using universal verification methodology (UVM) and contains two modules: i) The fault injector (FI), ii) The virtual comparator (VC). The FI systematically injects the error in one of the core's internal states by flipping one or multiple bits. While the VC stores the internal states of both the RISCV cores on each clock cycle and reports the anomaly, if detected. The SweRV EH1 (an open-source RISCV core) is customized to add the DCLS functionality and later Google RISC-V DV test suite is used to validate the proposed FVDCLS. In total, nine different verification scenarios are developed consisting of transient and permanent errors using FI module. For all these test scenarios the VC compared hundred different internal state signals/registers. The result shows that the FVDCLS successfully detected and reported all the mismatches. Moreover, the FVDCLS incorporates an error debugging module that generates the summary of conducted test particularly, the type of fault injected, the number of errors detected, and the injected fault coverage. The proposed FVDCLS is portable and configurable; therefore, can be seamlessly integrated with any processor (with DCLS feature) under test.

Authors

Publication Details

Published
2024-10-06
DOI
https://doi.org/10.1109/vlsi-soc62099.2024.10767783
Citations
2
Primary Topic
Smart Grid Security and Resilience
Type
article
Field-Weighted Citation Impact
0.66
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

FVDCLS: Functional Verification of RISCV Based Dual-Core Lockstep Feature Using Fault Injection Mechanism

Haroon Waris, Sajid Baloch, Muhammad Kashif Minhas
2 citations
Smart Grid Security and Resilience
0.66
article

FVDCLS: Functional Verification of RISCV Based Dual-Core Lockstep Feature Using Fault Injection Mechanism

Haroon Waris, Sajid Baloch, Muhammad Kashif Minhas
article en
2 citations

Abstract

This paper presents the framework to verify RISCV based Dual-Core Lockstep (DCLS), a feature that enhances the fault tolerance in processors. The DCLS feature, when two identical processors run in parallel, effectively detects the error and rolls back the core operations to a previously known state. The purpose-built fault injection mechanism is developed using universal verification methodology (UVM) and contains two modules: i) The fault injector (FI), ii) The virtual comparator (VC). The FI systematically injects the error in one of the core's internal states by flipping one or multiple bits. While the VC stores the internal states of both the RISCV cores on each clock cycle and reports the anomaly, if detected. The SweRV EH1 (an open-source RISCV core) is customized to add the DCLS functionality and later Google RISC-V DV test suite is used to validate the proposed FVDCLS. In total, nine different verification scenarios are developed consisting of transient and permanent errors using FI module. For all these test scenarios the VC compared hundred different internal state signals/registers. The result shows that the FVDCLS successfully detected and reported all the mismatches. Moreover, the FVDCLS incorporates an error debugging module that generates the summary of conducted test particularly, the type of fault injected, the number of errors detected, and the injected fault coverage. The proposed FVDCLS is portable and configurable; therefore, can be seamlessly integrated with any processor (with DCLS feature) under test.

Openalex Percentile: Top 28%
Smart Grid Security and Resilience
0.66
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.