Modeling and Investigating a Race Condition Detection Algorithm for Multithread Computational Systems

The paper presents the results of the comparative analysis of various methods for detecting and localizing race conditions in multithread systems with shared memory. The paper highlights the advantages of the static method, one of the effective implementations of which is an algorithm based on the a...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2019 XXI International Conference Complex Systems: Control and Modeling Problems (CSCMP) s. 356 - 359
Hlavní autoři: Barabanova, Polina, Zasov, Valery
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.09.2019
Témata:
algorithm
Analytical models
Computational modeling
Computer simulation
control flow graph
Flow graphs
flow synchronization
multithread programming
program package
program reliability
race conditions
Software
Software algorithms
Software reliability
Software testing
Static analysis
Synchronization
testing
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The paper presents the results of the comparative analysis of various methods for detecting and localizing race conditions in multithread systems with shared memory. The paper highlights the advantages of the static method, one of the effective implementations of which is an algorithm based on the analysis of a program presented as a control flow graph or one of its varieties-a joint flow execution graph. To solve applied problems, we carried out a software implementation of this algorithm, which served as the basis for developing a test program package for automated detection of race conditions in multithread systems. The reliability and efficiency of the program package is confirmed by the results of computer simulation illustrated using the Peterson algorithm.
DOI:10.1109/CSCMP45713.2019.8976855