Automatic Generation of Parallel Unit Tests
-
Name:
Konferenzartikel
-
Author:
Jochen Schimmel
Korbinian Molitorisz
Ali Jannesari
Walter F. Tichy -
Zusammenfassung
Multithreaded software is subject to data races.Currently available data race detectors report such errors to thedeveloper, but consume large amounts of time and memory manyapproaches are not applicable for large software projects. Unittests containing fractions of the program lead to better results.We propose AutoRT, an approach to automatically generateparallel unit tests as target for data race detectors. AutoRT uses the Single Static Multiple Dynamic (SSMD) analysis pattern toreduce complexity and can therefore be used efficiently even inlarge software projects. We evaluate AutoRT using MicrosoftCHESS and show that with SSMD all 110 data races containedin our sample programs can be located.
-
Year:
2013
- Links:
Titel Vorname Nachname |
---|
Dr. Ing. Ali Jannesari |
Dr. Ing. Korbinian Molitorisz |
Prof. em. Dr. Walter F. Tichy |
Dipl.-Inform. Jochen Schimmel |
Titel |
---|
SRG Entwurfsmustergestützte Anwendungsparallelisierung |
Forschungsprojekt AParT - Entwurfsmustergestütze Anwendungsparallelisierung |
Bibtex
@inproceedings{,
author={Jochen Schimmel, Korbinian Molitorisz, Ali Jannesari, Walter F. Tichy},
title={Automatic Generation of Parallel Unit Tests},
year=2013,
month=May,
booktitle={8th IEEE/ACM International Workshop on Automation of Software Test (AST 2013)},
publisher={IEEE},
url={https://ps.ipd.kit.edu/downloads/},
abstract={Multithreaded software is subject to data races.Currently available data race detectors report such errors to thedeveloper, but consume large amounts of time and memory manyapproaches are not applicable for large software projects. Unittests containing fractions of the program lead to better results.We propose AutoRT, an approach to automatically generateparallel unit tests as target for data race detectors. AutoRT usesthe Single Static Multiple Dynamic (SSMD) analysis pattern toreduce complexity and can therefore be used efficiently even inlarge software projects. We evaluate AutoRT using MicrosoftCHESS and show that with SSMD all 110 data races containedin our sample programs can be located.},
pages={40-46},
}