An Automatically Verified Prototype of the Android Permissions System

In a previous work we presented formal specifications of idealized formulations of the permission model of Android in the Coq proof assistant. This formal development is about 23 KLOC of Coq code, including proofs. In this work the Coq model is encoded in { l o g } (‘setlog’)—a satisfiability solver...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Journal of automated reasoning Ročník 67; číslo 2; s. 17
Hlavní autoři: Cristiá, Maximiliano, De Luca, Guido, Luna, Carlos
Médium: Journal Article
Jazyk:angličtina
Vydáno: Dordrecht Springer Netherlands 01.06.2023
Springer Nature B.V
Témata:
ISSN:0168-7433, 1573-0670
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In a previous work we presented formal specifications of idealized formulations of the permission model of Android in the Coq proof assistant. This formal development is about 23 KLOC of Coq code, including proofs. In this work the Coq model is encoded in { l o g } (‘setlog’)—a satisfiability solver and a constraint logic programming language— which is then used to automatically discharge most of the proofs performed in Coq. We show how the Coq model is encoded in { l o g } and how automated proofs are performed. The resulting { l o g } model is an automatically verified executable prototype of the Android permissions system. Detailed data on the empirical evaluation resulting after executing all the proofs in { l o g } is provided. The integration of Coq and { l o g } as to provide a framework featuring automated proof and prototype generation is discussed.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0168-7433
1573-0670
DOI:10.1007/s10817-023-09666-2