| Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership Types for Safe Programming: Preventing Data Races and Deadlocks. Proc. Object-Oriented Programming Systems, Languages and Applications, Seattle, Washington, November 2002. |
....static checking. Let us consider some example invariants, loosely described. Uniqueness is often used to strongly delimit effects, e.g. to avoid the need for a lock [16] the invariant is that the object referenced from a unique field is not accessible from anywhere else. Boyapati et al. [5] use a form of ownership as basis for a locking discipline: if a thread holds a lock on a certain object, it need not lock objects owned by that object. Strong forms of ownership have been proposed for various kinds of modular reasoning [14, 15, 10, 7, 23, 20, 1] A typical ownership invariant is ....
....sharing owners encapsulated Reps . Phi V : Phi hj : i : Rh5 : Fig. 3. Refined confinement scheme; dotted arrows indicate disallowed references. We are not the first to treat sharing of ownership; see [19, 5, 8]. A closely related work is that by Boyapati, Liskov, and Shrira [6] which gives an ownership system that allows sharing and which is claimed to be strong enough to support modular reasoning. As they remark, most other proposals are either too permissive for sound local reasoning or too ....
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In ACM Symposium on Object Oriented Programming: Systems, Languages, and Applications (00PSLA), 2002.
....a program. The second approach to managing aliasing employs alias encapsulation, as exemplified by ownership types [15] Simply put, these impose a form of object level privacy by preventing objects (rather than just fields) from being accessed outside of their enclosing encapsulation boundaries [15, 32, 13, 8, 6, 14, 1]. Ownership types have been employed for reasoning about programs [14, 32] for alias management [15, 33] in program understanding [1] to eliminate data races [8] and deadlocks [6] from concurrent programs, and to enable safe lazy updates in object oriented databases [7] Existing attempts to ....
.... than just fields) from being accessed outside of their enclosing encapsulation boundaries [15, 32, 13, 8, 6, 14, 1] Ownership types have been employed for reasoning about programs [14, 32] for alias management [15, 33] in program understanding [1] to eliminate data races [8] and deadlocks [6] from concurrent programs, and to enable safe lazy updates in object oriented databases [7] Existing attempts to unify uniqueness and ownership typing [8, 1] unfortunately o#er little additional benefit from their combination, while perpetuating the abstraction problem. We believe that the key ....
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In OOPSLA, November 2002.
....plans to investigate this further, although recently Boyapati et.al. have made steps in this direction [8] Applications of ownership and uniqueness. Ownership typing has many applications: reasoning about programs [13, 19, 28] alias management [14, 29] eliminating data races [7] and deadlock [5], program understanding [1] and for safe lazy upgrades in object oriented databases [6] Other applications which could be explored for ownership typing, especially now that we have uniqueness, include enforcing protocols [15] memory management [20] and mobility. 10. CONCLUSIONS AND FUTURE ....
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In OOPSLA, November 2002.
....statically (except for casts) that objects are only accessed through appropriate high level capabilities. Parameterized Race Free Java (PRFJ) uses the concept of object ownership and uniqueness to develop a type system to guarantee that a program is free of data races [BR01] and deadlocks [BR02]. PRFJ was not designed to encapsulate owned objects. However, a variant of PRFJ supports a stronger notion of object encapsulation than AliasJava: owned objects are confined within the owner, its owned objects, and its inner classes [BR02] This variant is more restrictive than AliasJava: an ....
....that a program is free of data races [BR01] and deadlocks [BR02] PRFJ was not designed to encapsulate owned objects. However, a variant of PRFJ supports a stronger notion of object encapsulation than AliasJava: owned objects are confined within the owner, its owned objects, and its inner classes [BR02]. This variant is more restrictive than AliasJava: an object can delegate a capability to access its owned state to its other owned objects and to its inner classes, but not to trusted external classes and methods, even temporarily. Thus, iterators can only be implemented as inner classes of the ....
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership Types for Safe Programming: Preventing Data Races and Deadlocks. Proc. Object-Oriented Programming Systems, Languages and Applications, Seattle, Washington, November 2002.
....statically (except for casts) that objects are only accessed through appropriate high level capabilities. Parameterized Race Free Java (PRFJ) uses the concept of object ownership and uniqueness to develop a type system to guarantee that a program is free of data races [BR01] and deadlocks [BR02]. A variant of PRFJ supports a stronger notion of object encapsulation than AliasJava: owned objects are confined within the owner, its owned objects, and its inner classes [BR02] However, this variant is considerably more restrictive than AliasJava: an object can delegate a capability to access ....
.... and uniqueness to develop a type system to guarantee that a program is free of data races [BR01] and deadlocks [BR02] A variant of PRFJ supports a stronger notion of object encapsulation than AliasJava: owned objects are confined within the owner, its owned objects, and its inner classes [BR02]. However, this variant is considerably more restrictive than AliasJava: an object can delegate a capability to access its owned state to its other owned objects and to its inner classes, but not to trusted external classes and methods. Thus, iterators can only be implemented as inner classes of ....
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership Types for Safe Programming: Preventing Data Races and Deadlocks. Proc. Object-Oriented Programming Systems, Languages and Applications, Seattle, Washington, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA), November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA), November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA), November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership Types for Safe Programming: Preventing Data Races and Deadlocks. Proc. Object-Oriented Programming Systems, Languages and Applications, Seattle, Washington, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In OOPSLA 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 17th Conference on Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA'02), Seattle, Washington, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In ObjectOriented Programming, Systems, Languages, and Applications (OOPSLA), November 2002. 12
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: preventing data races and deadlocks. In Proceedings of the 17th ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications, pages 211--230. ACM Press, 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In ACM Symposium on Object Oriented Programming: Systems, Languages, and Applications (OOPSLA), 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA), pages 211--230, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the ACM Conference on Object-Oriented Programming, Systems, Languages, and Appplications (OOPSLA), November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the ACM Conference on Object-Oriented Programming, Systems, Languages, and Appplications (OOPSLA), November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 2002 ACM SIGPLAN Conference on Object-Oriented Programming Systems, Languages and Applications, OOPSLA 2002, volume 37, number 11 in SIGPLAN Notices, pages 211--230. ACM, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In ACM Conference on Object-Oriented Programming, Systems, Languages, and Applications, pages 211--230, Seattle, WA, November 2002. 342
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 17th ACM Conference on Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA'02), pages 211--230, Seattle, Washington, USA, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 2002 ACM SIGPLAN Conference on Object-Oriented Programming Systems, Languages and Applications, OOPSLA 2002, volume 37, number 11 in SIGPLAN Notices, pages 211--230. ACM, November 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: Preventing data races and deadlocks. In Proceedings of the 2002.
No context found.
Chandrasekhar Boyapati, Robert Lee, and Martin Rinard. Ownership types for safe programming: preventing data races and deadlocks. In Proceedings of the 17th ACM conference on Object-oriented programming, systems, languages, and applications, pages 211--230. ACM Press, 2002. ISBN 1-58113-471-1.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC