D. Schmidt, Data- ow analysis is model checking of abstract interpretations, in: Symp. on Princ. of Prog. Lang., ACM Press, New York, NY, 1998, pp. 38-48.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....Neil D. Jones and Eric Van Wyk [6] Email: xeno diku.dk, Fax: 45) 35 32 14 01. c 2002 Published by Elsevier Science B. V. lead to stronger optimization algorithms [4] Other works have investigated temporal logic as a means to express program analyses and to derive analysis algorithms [12,13]. However declarative methods for specifying compiler optimizations have received relatively little attention. One approach to reason about program transformations for imperative programs by Kozen and Patron [5] demonstrates how an extension of Kleene algebra, Kleene algebra with tests (KAT) can ....

Schmidt, D.A. Data- ow analysis is model checking of abstract interpretations In Proc. of 25th ACM Symposium on Principles of Programming Languages, ACM, 1998.

....data ow analysis and it should be possible (although we have not formally checked this) to formalize the monovariant data ow analysis using our JVM model with an accordingly modi ed notion of well typing. Finally, note that we formalize model checking not algorithmically, as done in [16], but declaratively. Consequently our formalization of model checking is independent of the implemented model checking algorithm. In our work with Posegga [1] for instance, we implemented the backends both for symbolic and for explicit state model checkers. As our correctness theorem here is a ....

D. A. Schmidt. Data ow analysis is model checking of abstract interpretations. In POPL'98, pages 38-48. ACM Press 1998.

.... 10, 11] In the recent years there has been a convergence of ideas in static analysis and model checking: di erent kinds of program analysis can be performed by xed point computations and these computations can either be carried out by specialized algorithms or by general purpose model checkers [14, 15]. Whereas static analysis techniques have a longstanding history, the application of model checking to static analysis problems is more recent. The idea of using model checking for bytecode veri cation was originally suggested by Posegga and Vogt [9] They carried out a few small examples by hand ....

D. Schmidt. Data ow analysis is model checking of abstract interpretations. In Conference record of POPL '98, pages 38-48, San Diego, 1998. ACM Press.

....code, but correctness proofs were out of its scope (and would have been impractically complex in a denotational framework, witness [14] 1. 3 Model checking and program analysis This unsatisfactory situation has improved with the advent of model checking approaches to program analysis [4, 20, 22, 23, 25, 21]. Work by Ste en and Schmidt [22, 23] showed that temporal logic is well suited to describing data dependencies and other program properties exploited in classical compiler optimizations. In particular, work by Knoop, Ste en and R uthing [12] showed that new insights could be gained from using ....

....its scope (and would have been impractically complex in a denotational framework, witness [14] 1. 3 Model checking and program analysis This unsatisfactory situation has improved with the advent of model checking approaches to program analysis [4, 20, 22, 23, 25, 21] Work by Ste en and Schmidt [22, 23] showed that temporal logic is well suited to describing data dependencies and other program properties exploited in classical compiler optimizations. In particular, work by Knoop, Ste en and R uthing [12] showed that new insights could be gained from using temporal logic, enabling new and ....

D.A. Schmidt. Data- ow analysis is model checking of abstract interpretations. Proc. 25th ACM POPL, January 1998, 11 pages.

....methods have N 1, very few have N = 2, and N 2 is unheard of. Hence, the extra cost of polyvariant veri cation is entirely acceptable in practice. 6 Model checking of abstract interpretations It is folk lore that data ow analyses can be viewed as model checking of abstract interpretations [28]. Since a large part of bytecode veri cation is obviously an abstract interpretation (of a defensive JVM at the type level) it is natural to look at the remaining parts from a model checking perspective. Posegga and Vogt [22] were the rst to do so. They outline an algorithm that takes the ....

D. A. Schmidt. Data ow analysis is model checking of abstract interpretations. In POPL'98, pages 38-48. ACM Press, 1998.

....by re ning a domain to achieve completeness with respect to an arbitrary function, therefore including most domain re nements known in the literature. Our results are applied in the systematic re nement of data ow analysis as the model checking of abstract interpretations, as formulated in [20]. Related works The body of research on the connection among model checking, abstract interpretation and data ow analysis is huge, even though this eld still represents a major challenge in formal methods. In particular the problem of completeness in abstract model checking has been studied by ....

....sound and complete data ow analysis algorithms as complete model checking of abstract interpretations. Once again, completeness is achieved by removing spurious counterexamples. The idea of data ow analysis as model checking goes back to Ste en s pioneering work [22] More recently Schmidt [20, 21] showed that data ow analysis problems can be formulated as model checking problems of a trace based abstract interpretation of operational semantics. Intuitively, the concrete semantics of a program is a tree whose paths represent execution traces and whose nodes display the program s changing ....

[Article contains additional citation context not shown here]

D. A. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the 25th ACM SIGPLAN-SIGACT symposium on Principles of programming languages (POPL '98 ), pages 38-48. ACM Press, New York, 1998.

.... and uses the ESC Java checker to refute inconsistent annotations until convergence [15] Boolean programs can be viewed as abstract interpretations of the underlying program [8] The connections between model checking, data ow analysis and abstract interpretation have been explored before [28] [9] The model checker Bebop is based on earlier work on interprocedural data ow analysis [29, 26] Automatic iterative re nement based on error paths rst appeared in [20] and more recently in [6] Both e orts deal with nite state systems. An alternative approach to static validation of ....

D. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the Twenty Fifth Annual Symposium on Principles of Programming Languages, pages 38-48. ACM Press, 1998.

....for concurrent garbage collection. 1 Introduction Abstract model checking is a methodology for verifying state tarnsition systems. In abstract model checking, we construct an abstract system of the target system by an abstract map, and then apply traditional model checking to the abstract system [6, 10, 9, 7]. While the number of states is in nite or intractable, the number of abstract states can be reduced to nite or tractable provided that the abstract map is appropriate[2, 5] In this paper, we consider the case that states of the system have link structures. A link structure consists cells which ....

....for concurrent garbage collection. One is the so called on the y garbage collector formulated by Dijkstra et al. 3] The other is the snapshot garbage collector by Yuasa [12] modi ed to t our model. 4 R[0] XXXXX Xz R[1] R[2] registers C[7] J J J J J C[8] C[9] C[10] cells Figure 1: registers and cells 4.1.1 Cells and Registers The heap for cells is formally de ned as a function (or an array) from cell indices to cells. We do not specify the set of cell indices at the moment. We use C to denote the heap, and C[i] to denote the i th cell. C : heap = ....

David A. Schmidt. Data- ow analysis is model checking of abstract interpretations. In POPL'98: The 25th ACM SIGPLAN-SIGACT Symposium on Principles of Programming Languages, pages 38-48, 1998.

.... of re nement, including trace containment, have been studied before [Mil71] AL88] Kur94] Boolean programs can be viewed as abstract interpretations of the underlying program [CC77] The connections between model checking, data ow analysis and abstract interpretation have been explored before [Sch98] CC00] The model checking problem for pushdown automata has been studied before [SB92] BEM97] FWW97] The CFL reachability framework from [RHS95] builds on earlier work in interprocedural data ow analysis from [KS92] and [SP81] However, our use of CFL reachability as a model checking ....

D.A. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the Twenty Fifth Annual Symposium on Principles of Programming Languages, pages 38-48. ACM Press, 1998.

....programs, these approaches abstract away data, and concentrate only on control. As a result spurious paths can arise in these models due to information loss about data correlations. The connections between model checking, data ow analysis and abstract interpretation have been explored before [Sch98] CC00] The RHS algorithm [RHS95,RHS96] builds on earlier work in interprocedural data ow analysis from [KS92] and [SP81] We have shown how this algorithm can be generalized to work as a model checking procedure for boolean programs. Also, our choice of hybrid representation of the state space ....

D.A. Schmidt. Data ow analysis is model checking of abstract interpretation. In POPL 98: Principles of Programming Languages, pages 38-48. ACM Press, 1998.

.... of re nement, including trace containment, have been studied before [Mil71] AL88] Kur94] Boolean programs can be viewed as abstract interpretations of the underlying program [CC77] The connections between model checking, data ow analysis and abstract interpretation have been explored before [Sch98] CC00] The model checking problem for pushdown automata has been studied before [SB92] BEM97] FWW97] The CFL reachability framework from [RHS95] builds on earlier work in interprocedural data ow analysis from [KS92] and [SP81] However, our use of CFL reachability as a model checking ....

D.A. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the Twenty Fifth Annual Symposium on Principles of Programming Languages, pages 38-48. ACM Press, 1998.

....must respect the labels: c R a c implies that a a . Thus, properties that hold true for all paths starting at a also hold true for all paths starting at c. This supports sound veri cation of LTL and ACTL coded properties of temporal logic on the abstract Kripke structure [4, 12, 34, 37]. A left total simulation ensures that every state in the concrete structure can be modelled in the abstract structure. Right totality ensures that there are no super uous abstract states. Simulations play a crucial role in equivalence proofs of interpreters [20] Each execution step of ....

....that follow the present work are: i) the extraction of properties from an abstract Kripke structure and (ii) the application of properties to transformation of the corresponding concrete Kripke structure. Topic (i) can be best understood as validating temporal logic properties by model checking [34 37]; Topic (ii) is surprisingly underdeveloped, although Lacey, Jones, Van Wyk, and Frederiksen [18] and Cousot and Cousot [11] give recent proposals. Acknowledgements: I thank Neil Jones for many years of inspiring work. The current paper evolved from research presented at the Workshop on Re ....

D.A. Schmidt. Data- ow analysis is model checking of abstract interpretations. In Proc. 25th ACM Symp. on Principles of Prog. Languages. ACM Press, 1998.

No context found.

D. Schmidt, Data- ow analysis is model checking of abstract interpretations, in: Symp. on Princ. of Prog. Lang., ACM Press, New York, NY, 1998, pp. 38-48.

No context found.

D.A. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the Twenty Fifth Annual Symposium on Principles of Programming Languages, pages 38-48. ACM Press, 1998.

No context found.

David A. Schmidt. Data ow analysis is model checking of abstract interpretations. In Conference Record of the 25th ACM SIGPLAN-SIGACT Symposium on Principles of Programming Languages, San Diego CA, January 1998.

No context found.

D.A. Schmidt. Data ow analysis is model checking of abstract interpretation. In Proceedings of the Twenty Fifth Annual Symposium on Principles of Programming Languages, pages 38-48. ACM Press, 1998.

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