Details
| Original language | English |
|---|---|
| Title of host publication | Computer Safety, Reliability, and Security - 41st International Conference, SAFECOMP 2022, Proceedings |
| Subtitle of host publication | SAFECOMP 2022 - Proceedings |
| Editors | Mario Trapp, Francesca Saglietti, Marc Spisländer, Friedemann Bitsch |
| Pages | 252-266 |
| Number of pages | 15 |
| ISBN (electronic) | 978-3-031-14835-4 |
| Publication status | Published - 25 Aug 2022 |
| Event | 41st SAFECOMP: International Conference on Computer Safety, Reliability, and Security, 2022 - Munic, Germany Duration: 6 Sept 2022 → 9 Sept 2022 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 13414 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (electronic) | 1611-3349 |
Abstract
Fault-injection (FI) campaigns provide an in-depth resilience analysis of safety-critical systems in the presence of transient hardware faults. However, FI campaigns require many independent injection experiments and, combined, long run times, especially if we aim for a high coverage of the fault space. Besides reducing the number of pilot injections (e.g., with def-use pruning) in the first place, we can also speed up the overall campaign by speeding up individual experiments. From our experiments, we see that the timeout failure class is especially important here: Although timeouts account only for 8% (QSort) of the injections, they require 32% of the campaign run time. In this paper, we analyze and discuss the nature of timeouts as a failure class, and reason about the general design of dynamic timeout detectors. Based on those insights, we propose ACTOR, a method to identify and abort stuck experiments early by performing autocorrelation on the branch-target history. Applied to seven MiBench benchmarks, we can reduce the number of executed post-injection instructions by up to 30%, which translates into an end-to-end saving of 27%. Thereby, the absolute classification error of experiments as timeouts was always less than 0.5%.
ASJC Scopus subject areas
- Mathematics(all)
- Theoretical Computer Science
- Computer Science(all)
- General Computer Science
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Computer Safety, Reliability, and Security - 41st International Conference, SAFECOMP 2022, Proceedings: SAFECOMP 2022 - Proceedings. ed. / Mario Trapp; Francesca Saglietti; Marc Spisländer; Friedemann Bitsch. 2022. p. 252-266 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13414 LNCS).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - ACTOR
T2 - 41st SAFECOMP: International Conference on Computer Safety, Reliability, and Security, 2022
AU - Thomas, Tim-Marek
AU - Dietrich, Christian
AU - Pusz, Oskar
AU - Lohmann, Daniel
PY - 2022/8/25
Y1 - 2022/8/25
N2 - Fault-injection (FI) campaigns provide an in-depth resilience analysis of safety-critical systems in the presence of transient hardware faults. However, FI campaigns require many independent injection experiments and, combined, long run times, especially if we aim for a high coverage of the fault space. Besides reducing the number of pilot injections (e.g., with def-use pruning) in the first place, we can also speed up the overall campaign by speeding up individual experiments. From our experiments, we see that the timeout failure class is especially important here: Although timeouts account only for 8% (QSort) of the injections, they require 32% of the campaign run time. In this paper, we analyze and discuss the nature of timeouts as a failure class, and reason about the general design of dynamic timeout detectors. Based on those insights, we propose ACTOR, a method to identify and abort stuck experiments early by performing autocorrelation on the branch-target history. Applied to seven MiBench benchmarks, we can reduce the number of executed post-injection instructions by up to 30%, which translates into an end-to-end saving of 27%. Thereby, the absolute classification error of experiments as timeouts was always less than 0.5%.
AB - Fault-injection (FI) campaigns provide an in-depth resilience analysis of safety-critical systems in the presence of transient hardware faults. However, FI campaigns require many independent injection experiments and, combined, long run times, especially if we aim for a high coverage of the fault space. Besides reducing the number of pilot injections (e.g., with def-use pruning) in the first place, we can also speed up the overall campaign by speeding up individual experiments. From our experiments, we see that the timeout failure class is especially important here: Although timeouts account only for 8% (QSort) of the injections, they require 32% of the campaign run time. In this paper, we analyze and discuss the nature of timeouts as a failure class, and reason about the general design of dynamic timeout detectors. Based on those insights, we propose ACTOR, a method to identify and abort stuck experiments early by performing autocorrelation on the branch-target history. Applied to seven MiBench benchmarks, we can reduce the number of executed post-injection instructions by up to 30%, which translates into an end-to-end saving of 27%. Thereby, the absolute classification error of experiments as timeouts was always less than 0.5%.
UR - http://www.scopus.com/inward/record.url?scp=85137998003&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-14835-4_17
DO - 10.1007/978-3-031-14835-4_17
M3 - Conference contribution
SN - 978-3-031-14834-7
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 252
EP - 266
BT - Computer Safety, Reliability, and Security - 41st International Conference, SAFECOMP 2022, Proceedings
A2 - Trapp, Mario
A2 - Saglietti, Francesca
A2 - Spisländer, Marc
A2 - Bitsch, Friedemann
Y2 - 6 September 2022 through 9 September 2022
ER -