Verifying Temporal Properties of Systems

122 Pages·1992·3.178 MB·English

by  Julian Charles Bradfield (auth.)

Progress in Theoretical Computer Science
Editor
Ronald V. Book, University of California
Editorial Board
Erwin Engeler, ETH Zentrum, Zurich, Switzerland
Gerard Huet, INRIA, Le Chesnay, France
Jean-Pierre Jouannaud, Universite de Paris-Sud, Orsay, France
Robin Milner, University of Edinburgh, Edinburgh, Scotland
Maurice Nivat, Universite de Paris VII, Paris, France
Martin Wirsing, Universitat Passau, Passau, Germany
Julian Charles Bradfield
Verifying Temporal
Properties of
Systems
Birkhauser
Boston • Basel • Berlin
Julian Charles Bradfield
Department of Computer Science
University of Edinburgh
The King's Building
Mayfield Road
Edinburgh
The United Kingdom
EH93JZ
Printed on acid-free paper.
© Julian Charles Bradfield 1992.
Softcover reprint of the hardcover 1st edition 1992
Copyright is not claimed for works of U.S. Govermnent employees.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior permission of the copyright owner.
Permission to photocopy for internal or personal use, or the internal or personal use of specific clients,
is granted by Birkhiiuser Boston for libraries and other users registered with the Copyright Clearance
Center (CCC) , provided that the base fee of $0.00 per copy, plus $0.20 per page is paid directly to
CCC, 21 Congress Street, Salem, MA 01970, U.S.A. Special requests should be addressed directly to
Birkhiiuser Boston, 675 Massachusetts Avenue, Cambridge, MA 02139, U.S.A.
ISBN-13: 978-1-4684-6821-2 e-ISBN-13: 978-1-4684-6819-9
DOl: 10.1007/978-1-4684-6819-9
Camera ready text prepared by the Author.
Contents
Contents .............................................................. v
Preface .............................................................. vii
Acknowledgements .................................................. viii
Chapter 1. Introduction ................................................ 1
1.1 Infinite state model-checking ...................................... 1
1.2 Background ...................................................... 2
1.2.1 Imperative programs and Hoare logic ......................... 2
1.2.2 Dynamic logic ................................................ 5
1.2.3 Modal and temporal logic ..................................... 7
1.2.4 Modal mu-calculus ........................................... 9
1.2.5 Model checking ............................................... 9
1.3 Local model-checking and infinite systems ....................... 10
1.4 Synopsis ........................................................ 11
Chapter 2. Program Logics and the Mu-Calculus ...................... 14
2.1 Semantics of temporal logics ..................................... 14
2.2 The propositional modal mu-calculus ............................ 16
Chapter 3. The Tableau System ....................................... 30
3.1 Intuition behind the tableau system .............................. 31
3.2 Definition of the tableau system ................................. 33
3.3 Simple examples ................................................. 37
3.4 Soundness of the tableau system ................................. 40
3.5 Completeness of the tableau system .............................. 42
3.6 Variations on the theme ......................................... 44
3.7 The tableau system and Hoare logic ............................. 46
Chapter 4. Applications to Nets ....................................... 51
4.1 Petri nets ....................................................... 51
4.1.1 Basic definitions ............................................. 51
4.1.2 Properties and classes of nets ................................ 53
4.1.3 Nets in systems modelling ................................... 57
4.1.4 Building a reader-writer system by shared places ............ 59
vi CONTENTS
4.1.5 Transition-oriented design ................................... 62
4.2 Basic application to nets ......................................... 64
4.2.1 A simple example ........................................... 65
4.2.2 A slot machine .............................................. 67
4.3 Using schematic tableaux ........................................ 70
4.3.1 A simple parametrized safety proof. ......................... 71
4.3.2 A parametrized liveness proof. ............................... 74
4.4 Using limited reach ability analysis-the coverability graph ....... 77
4.5 Some remarks on compositionality. .............................. 82
Chapter 5. The Complexity of Mu-Formulae on Nets .................. 85
5.1 Beyond semi-linearity. ........................................... 86
5.2 Undecidability of the model-checking problem .................... 89
5.3 Ascending the arithmetical hierarchy. ............................ 91
5.4 Beyond the arithmetical hierarchy. .............................. 96
5.4.1 Proof by inductive definitions ................................ 97
5.4.2 Proof by partially-ordered quantification ..................... 98
Chapter 6. Conclusions and Further Work ............................ 101
6.1 Incorporating reasoning ........................................ 101
6.2 Decidability of model-checking .................................. 102
6.3 Proving success ................................................ 102
References ........................................................... 105
List of Notations ..................................................... 110
Index ................................................................ 112
Preface
This monograph aims to provide a powerful general-purpose proof tech
nique for the verification of systems, whether finite or infinite. It extends
the idea of finite local model-checking, which was introduced by Stirling and
Walker: rather than traversing the entire state space of a model, as is done
for model-checking in the sense of Emerson, Clarke et ai. (checking whether
a (finite) model satisfies a formula), local model-checking asks whether a
particular state satisfies a formula, and only explores the nearby states far
enough to answer that question. The technique used was a tableau method,
constructing a tableau according to the formula and the local structure of
the model. This tableau technique is here generalized to the infinite case by
considering sets of states, rather than single states; because the logic used,
the propositional modal mu-calculus, separates simple modal and boolean
connectives from powerful fix-point operators (which make the logic more
expressive than many other temporal logics), it is possible to give a rela
tively straightforward set of rules for constructing a tableau. Much of the
subtlety is removed from the tableau itself, and put into a relation on the
state space defined by the tableau-the success of the tableau then depends
on the well-foundedness of this relation.
The generalized tableau technique is exhibited on Petri nets, and various
standard notions from net theory are shown to playa part in the use of the
technique on nets-in particular, the invariant calculus has a major role.
The requirement for a finite presentation of tableaux for infinite systems
raises the question of the expressive power of the mu-calculus. This is
studied in some detail, and it is shown that on reasonably powerful models
of computation, such as Petri nets, the mu-calculus can express properties
that are not merely undecidable, but not even arithmetical.
This monograph is based on my doctoral dissertation, examined in May
1991 at the University of Edinburgh; some of the material has been pub
lished in [Bra9l], [BrS90] and [BrS91].
Acknowledgements
Most of this work was done while I was a research student in the De
partment of Computer Science at the University of Edinburgh, supported
by an award from the U.K. Science and Engineering Research Council; the
remainder was done while I was a research assistant on the SERC grant
'Mathematically Proven Safety Systems' (GR/F 38808) at Edinburgh.
I thank especially my thesis supervisor, Colin Stirling, without whose
guidance, encouragement and criticism this work would not have appeared.
My thesis examiners, Eike Best and Robin Milner, suggested some im
provements; Robin Milner suggested producing this monograph.
I have benefited from conversations with many people in Edinburgh,
particularly Mads Dam and Glenn Bruns.
I thank Glynn Winskel, who introduced me to temporal logic and model
checking, and suggested that I go to Edinburgh.
Finally, I thank Perdita Stevens for moral support throughout.
Chapter 1
Introduction
1.1 Infinite state model-checking.
This monograph is concerned with the verification of infinite systems. 'Ver
ification' has connotations of algorithmic checking, and is chosen for that
reason, for the topic is the combination of two areas which have hitherto
been considered separately.
Verification in its widest sense has been a major research topic since the
beginnings of computer science. For some time, effort was directed towards
proving properties of programs by means of logical reasoning, with the
meaning of programs given either by logic also-whether predicate logic, as
with Floyd, de Bakker, Park et al., or temporal logic, as with Manna and
Pnueli-or by a denotational or operational semantics.
About ten years ago, a new approach was begun by Clarke, Emerson,
Sifakis and others. This approach is termed 'model-checking', since the idea
is to consider some system as a model for some logic, and check whether the
model satisfies a given formula of the logic expressing some desirable prop
erty. The distinctive feature is checking: rather than performing proofs, one
has an algorithm which takes the model and formula as input and returns
a yes/no answer. Clarke et al. developed algorithms for the logic CTL (de
scribed later), and their algorithms have been implemented by themselves
and others (including an implementation in purely functional ML by this
author), and have produced useful results.
Also ten years ago, Pratt and Kozen introduced a temporal logic called
the 'modal mu-calculus'. This logic combines standard modal logic with
least and greatest fix-point operators to produce a remarkably expressive
logic. Although it looks 'modal', in that no mention is made of paths, the
fix-point operators allow the expression of very complex 'temporal' proper
ties, that is, properties involving paths. The modal mu-calculus subsumes
many other temporal logics, and so its study is especially useful. Amongst
other research, the model-checking idea was transferred to the modal mu
calculus by Emerson and Lei.
2 INTRODUCTION
However, these model-checking algorithms all proceed by an exhaustive
traversal of the state space of the model. Therefore, they are inherently
incapable of considering infinite systems. Moreover, this exhaustive traver
sal may well be unnecessary--some properties depend only on very small
parts of a system.
On the other hand, infinite systems and potentially infinite systems,
which become more common as interest develops in concurrent and dis
tributed systems, are of course amenable to the logical attack mentioned
first. The problem here is that even when a complete proof system is avail
able, an effective proof system is usually not. The quest for effective proof
techniques has produced much interesting research (particularly on Petri
nets), but it is a fact of life that there are no general effective techniques
for any but a very small class of problems.
My purpose here is to bring together the ideas of finite, algorithmic,
model-checking and the ability to perform proofs. This is made easier by
using the propositional modal mu-calculus, since its connectives fall into
two classes. The first class comprises the modal and boolean operators,
which are very simple in nature and can be 'checked'. The second com
prises the fix-point operators: these, especially the least fix-point, introduce
complexity which may require subtle techniques to analyse.
1.2 Background.
1.2.1 Imperative programs and Hoare logic.
The formal proof of program correctness began in the late 1960s. Floyd
[Fl067) proposed a method of assigning meanings to imperative languages
by annotating each point in the control flow with a proposition, in some
logic such as predicate calculus, which should hold there, and he gave
techniques for doing this. Floyd's work was formalized and developed by
Manna, both alone and with Pnueli [MaP69], to turn program properties
into questions of satisfiability or validity in first order logic. This was
extended by Park [Par70] who considered second order logic and the use
of fix-point induction to prove properties. This work considered recursive
program schemes, i.e. programs with what amounts to mutually recursively
defined functions. By restricting consideration to straightforward programs
with loops, but no functions or procedures, Hoare [Hoa69] produced his cel
ebrated system of axioms for partial correctness. The language studied by
Hoare was an abstraction of the common contemporary languages such as
Fortran and Algol, and in its simplest form is thus: