Abstract
A temporal verification method which is based upon partial order semantics of traces [31] is presented. The semantic model used here can express the distributed nature of a program. E.g., properties such as serializability of database transactions, layering of a program, snapshots or the parallel execution of program segments.
| Original language | English |
|---|---|
| Title of host publication | Automata, Languages and Programming - l7th International Colloquium, Proceedings |
| Editors | Michael S. Paterson |
| Publisher | Springer Verlag |
| Pages | 553-571 |
| Number of pages | 19 |
| ISBN (Print) | 9783540528265 |
| DOIs | |
| State | Published - 1990 |
| Externally published | Yes |
| Event | 17th International Colloquium on Automata, Languages and Programming, 1990 - Warwick, United Kingdom Duration: 16 Jul 1990 → 20 Jul 1990 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 443 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | 17th International Colloquium on Automata, Languages and Programming, 1990 |
|---|---|
| Country/Territory | United Kingdom |
| City | Warwick |
| Period | 16/07/90 → 20/07/90 |
Bibliographical note
Publisher Copyright:© Springer-Verlag Berlin Heidelberg 1990.
Funding
Partial order semantics is recently becoming accepted as an intuitive model for representing distributed programs executions (see [33,43,22,39]). Among its many presentations, trace semantics [31] seems to be of a particular elegance. While it is generally recognized that partial order semantics provides a significantly more faithful representation of concurrency than, say, interleaving semantics, the development of corresponding logics for specifying and verifying properties of programs, as they appear in the partial order semantics, has been slower. It is only recently [18,40,37] that several versions of temporal logics over partial orders have been defined and illustrated. In this paper we continue the development of the temporal logic ISTL* [18] (Interleaving Set Temporal Logic), by presenting for it complete proof rules for proving properties of the form AG~ (in all traces ~) and EF~o (every run contains an observation, on which eventually ~). These two rules cover many of the safety and liveness properties one may wish to prove for a concurrent program. Let us compare the expressive power of logics based on interleaving semantics with logics similar to the one presented here, that are based on partial order trace semantics [31] or similarly on POMSETS [39,13]. Interleaving semantics assign to a concurrent program a large set of observations (interleaving sequences) that represent all the possible executions of the program when linearized along a single time axis. Trace theory introduces more structure into this large set by defining an equivalence relation between the observations, such that two observations are considered equivalent if they differ only by the order they execute independent (concurrent) actions. We refer to the equivalence classes of this relation as runs or computations (interleaving sets in the terminology of [18]). A temporal logic over interleaving semantics specifies a property of the program by requiring that all observations satisfy a particular sequence predicate, that usually constrains the order in *This research was supported in part by the European Community ESPRIT Basic Research Action project 3096 (SPEC).
| Funders | Funder number |
|---|---|
| European Community ESPRIT Basic Research Action |