L. Aceto and M. Hennessy, Termination, deadlock and divergence, Journal of the ACM 39(1), 1992, pp. 147-187.  Home/Search   Document Not in Database   Summary   ACM   TOC   Related Articles   Check

....any process with a port of type (Y; in Y ) can be connected to the (Y; out Y ) port of Protocol . No further checking is needed in order to establish the correctness of these connections. 6 Related Work The analysis of deadlock freedom in concurrency has been investigated by a number of authors [4, 6, 7, 9, 21, 23, 24]. The work of Roscoe and Daithi [23] is one of the most recent, and also the most relevant to our own. Their approach is to define a variant function which assigns a value to each state of a process. A network of processes, satisfying certain conditions, is deadlock free if for each process P that ....

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39:147--187, January 1992.

.... (1) oe j= oe j= H ) oe j= The reader familiar with the literature on Hennessy Milner logics [43] for prebisimulationlike relations will have noted that our notion of satisfaction for negative transition formulae is akin to that for formulae of the form [a]F given in, e.g. [60, 82, 83, 1, 8]. In those references, the new interpretation is necessary to obtain monotonicity of the satisfaction relation with respect to the appropriate notion of prebisimulation. In this study, our interpretation of negative premises will be crucial to obtain operations that are monotonic with respect to ....

.... terms will be given in this study by means of the standard recursion rule (3) In order to define the operational semantics of CREC( Sigma G ; PVar) we need, first of all, to extend the convergence predicate to CREC( Sigma G ; PVar) This can be done in standard fashion following, e.g. [45, 38, 8]. Definition 4.1 The convergence predicate # (abbreviated to # when the GSOS system G is clear from the context) is the least predicate over CREC( Sigma G ; PVar) that satisfies the following clauses: 1. f(P 1 ; P l ) # if (a) f (b) for every argument i of f , if f tests i then P ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187.

....modification of the system introduced in [BKO88] However in that reference only a sub language consisting of recursion free terms of ACP is considered. The logical characterization gives us an interesting and strong proof technique for proving properties of the behavioural preorder. Like in [AH92] we use it to prove that the preorder is finitary in the sense of [Abr91] However here we go a step further and prove the algebraicity, in the sense of [Hen88] of the preorder by using its modal characterization. This allows us to reduce the proof of the completeness of the proof system to only ....

....to only proving the completeness over finite or recursion free terms. This proof technique is interesting in itself as the algebraicity usually follows only as a corollary to a full abstractness result with respect to a denotational model defined in terms of an algebraic domain. See for instance [Hen88, AH92, HI93, Ing95] for this. The structure of what remains of the paper is as follows: In the next section we define the readiness preorder as an abstraction on the stepwise semantics given by labelled transition systems. We also give a bisimulation like characterization of the preorder. Section 3 is devoted to a ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39(1):147--187, January 1992.

....as E rec is complete wrt. ACT 10 (n) v by (1) and Thm. 2. 4: In the proof for the full abstractness of the model with respect to our representant for the behavioural preorder we will use some standard techniques which are used to prove similar full abstractness results in the literature [Hen88, AH92, HI93]. This gives us some guidelines about properties our behavioural interpretation of the preorder induced by the model should satisfy. This is formulated in the following lemma: Theorem 2.9 Assume that CCSL Theta CCSL satisfies the following conditions: 1. Finitariness: For all ct; cu 2 CCSL ....

....Definition 5.5 The sort of ct 2 CCSL is given by 1. Sort(cp) fc 2 Chanjcp Gamma ; chan(a) cg [ fSort(cu)j9:cp Gamma cug 2. Sort( x]t) fSort(t[v=x] jv 2 V alg 3. Sort(e;cq) Sort(cq) Note that, because of our restriction to finite renamings, Sort(ct) is finite for all ct [Abr91, AH92]. Definition 5.6 [Compact Projections] We define the n th projection of ct on CoTerms inductively as follows: 1. a) cp (b) cp f:ct jcp Gamma ctg jcp g 2. a) x]p) x] Omega (b) x]p) x]x : v 1 ; Delta Delta Delta ; v n 1 ) p[v 1 =x] Delta Delta ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39(1):147--187, January 1992.

....several process description languages. In particular, when proving full abstraction results for denotational semantics based on the Scott Strachey approach [29] for CCS like languages, several preorders based on bisimulation have been considered. The interested reader is invited to consult, e.g. [13, 10, 4, 6, 17] for examples of full abstraction results with respect to bisimulation based preorders for variations on CCS. In this paper, we shall study one such bisimulation based preorder whose connections with domain theoretic models are by now well understood, viz. the prebisimulation preorder ....

....trees. Such an indirect definition often makes it quite hard to establish results about this relation, and a lot of research effort has been devoted to finding alternative, observation independent versions of the finitary bisimulation for different process description languages. cf. e.g. [10, 4, 6, 16] for examples of these results. A general, observation independent characterization of the finitary bisimulation for arbitrary transition systems has been given by Abramsky in [4] This characterization is couched in logical terms, and is an impressive byproduct of Abramsky s theory of domains ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187.

....any process with a port of type (Y; in Y ) can be connected to the (Y; out Y ) port of Protocol . No further checking is needed in order to establish the correctness of these connections. 6 Related Work The analysis of deadlock freedom in concurrency has been investigated by a number of authors [4, 6, 7, 9, 21, 23, 24]. The work of Roscoe and Daithi [23] is one of the most recent, and also the most relevant to our own. Their approach is to define a variant function which assigns a value to each state of a process. A network of processes, satisfying certain conditions, is deadlock free if for each process P that ....

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39:147--187, January 1992.

....is made about the quantity of faults, and conditional, given the maximal number of times they can occur successively. There is a number of directions that we plan to develop this work. We plan to study the use of other bisimulation like relations, like the partial [Wal90] the terminating [AH92] and the context dependent [Lar87] bisimilarities for fault tolerance. In the presence of faulty transitions, a convergent process may diverge and the one which terminates (successfully) may deadlock. We plan to relate our theory with modal specifications [LT88] which constrain possible ....

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of ACM, 39(1):147--187, 1992.

....is the fact that it allows for the abstraction from divergence, i.e. infinite sequences of internal computations, in process behaviours. Semantic theories for processes based on the bisimulation idea which take divergence into account have also been considered in the literature, see, e.g. [Wal90, AH92]. In those studies, bisimulation equivalence is extended to a bisimulation preorder, usually referred to as prebisimulation. Intuitively if a process p is smaller than a process q with respect to the bisimulation preorder, then the two processes are bisimilar, but p may diverge more often than q. ....

....have presented more abstract formal descriptions for it, using different theoretical tools to analyse in depth the nature of the concept. Examples of such alternative descriptions are characterizations of bisimulation by means of modal logics, Sti87, Mil89] sound and complete axiomatizations, [Hen81, Wal90, AH92], and fully abstract denotational models, Hen81, Abr91, AH92] The denotational models are usually given in terms of Sigma domains, i.e. algebraic cpos endowed with a continuous Sigma algebra structure. The existence of a fully abstract model in terms of a Sigma domain has the consequence ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39(1):147--187, January 1992.

....transitions and or negated predicates occurs. We will sketch this in the next paragraph. In the literature we see more and more that operational rules in the style of Plotkin are decorated with extra predicates on states to express matters like (un)successful termination, convergence, divergence [1], enabledness [7] maximal delay, side conditions [20] etc. Baeten and Verhoef give many examples of this kind of decorated transition rules in their paper on the path format [5] thereby showing that there is a need for a general format describing such decorated rules. Another phenomenon that we ....

....0 by closed terms t and t 0 . Since the number of s occurring in (t) is one greater than the number of s occurring in t we are done. The other rules are dealt with just as simply. Termination and convergence The next example is an operational semantics originating from Aceto and Hennessy [1]. It is an operational semantics of a CCS like process algebra extended with a successful termination predicate and a convergence predicate. Their approach is to first inductively define both predicates and then define the transition relation using one of the predicates. In this semantics they use ....

[Article contains additional citation context not shown here]

L. Aceto, M. Hennessy, Termination, deadlock and divergence, JACM, 39(1):147--187, January 1992.

....repeated inputs. In general, the choice of input or output can be made independently at each step. The structure of ASProc D allows a rich variety of input output behaviour to be specified. 6 Related Work The analysis of deadlock freedom in concurrency has been investigated by a number of authors [4, 7, 8, 10, 22, 24, 25]. The work of Roscoe and Daithi [24] is one of the most recent, and also the most relevant to our own. Their approach is to define a variant function which assigns a value to each state of a process. A network of processes, satisfying certain conditions, is deadlock free if for each process P that ....

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39:147--187, January 1992.

....with axiomatising equivalences. Partial orders over processes have been introduced for several reasons to match a notion of effective testing [Mil81] to allow for divergent [Wal88] and under specified [CS90] processes and to allow a CPO based denotational semantics [HP80,Hen81, Hen88,AH88,Abr91] Walker gives in [Wal88] a complete system using a variant of the unique fixed point rule of [Mil84] The CPO models admit infinitary rules such as induction. We recall a definition from [Hen88] for term models which requires a definition of the syntactic approximants to a recursive ....

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Technical Report 6/88, Sussex University, 1988.

....we get the orginal ACP back as an SRM specification of the general theory. As a side effect, this introduces a new constant ; into ACP. As future work, we leave the exact determination of the interaction of all the special constants: ffi (inaction, BeK84] nil (CCS termination [Mil89] [AcH92]) ffl (termi Process Algebra with Feedback 22 nation option [Vra91] silent step [Mil89] Delta (divergence [AcH92] chaos [BrHR84] 0 (zero process [BaB90] X Y = Y X (X Y ) Z = X (Y Z) X X = X (X Y ) Delta Z = X Delta Z Y Delta Z (X Delta Y ) Delta Z = ....

....; into ACP. As future work, we leave the exact determination of the interaction of all the special constants: ffi (inaction, BeK84] nil (CCS termination [Mil89] AcH92] ffl (termi Process Algebra with Feedback 22 nation option [Vra91] silent step [Mil89] Delta (divergence [AcH92]) chaos [BrHR84] 0 (zero process [BaB90] X Y = Y X (X Y ) Z = X (Y Z) X X = X (X Y ) Delta Z = X Delta Z Y Delta Z (X Delta Y ) Delta Z = X Delta (Y Delta Z) X ; X a ffi = a ; Delta X = ffi Delta X = ffi X Delta ; X Delta ffi a j b = fl(a; b) if ....

L. Aceto and M. Hennessy, Termination, deadlock and divergence, Journal of the ACM 39(1), 1992, pp. 147-187.

....then terminate unsuccessfully. Baeten and Van Glabbeek [6] give an operational semantics with a termination predicate that is more in accordance with our intuition about the empty process. We will return to this issue in another context where we treat an operational semantics of Aceto and Hennessy [1] in which a termination predicate occurs; see section 2. As a last example of semantics that are better understood with than without predicates we will mention the clear operational semantics using predicates that Klusener [18] gives for various real time process algebras. He also gives ....

.... divergence, weak bisimulation, a zero object, side conditions, functions, real time, discrete time, sequencing, negative premises, negative conclusions, and priorities (or a combination of these notions) The first example concerns an operational semantics that originates from Aceto and Hennessy [1]. It is an operational semantics of a CCS like process algebra extended with a successful termination predicate and a convergence predicate. Their approach is to first inductively define the termination predicate p , which is a postfix denoted predicate. In table 2 we enumerate the rules as they ....

[Article contains additional citation context not shown here]

L. Aceto, M. Hennessy, Termination, deadlock and divergence, JACM 39(1):147--187, 1992.

No context found.

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187. 12

....can often be expressed positively using predicates [27] However, in the literature we see more and more that SOS definitions are decorated with predicates and or negative premises. For example, predicates are used to express matters like (un)successful termination, convergence, divergence [10], enabledness [41] maximal delay, and side conditions [165] Negative premises are used to describe, e.g. deadlock detection [137] sequencing [55] priorities [24, 65] probabilistic behaviour [139] urgency [58] and various real [136] and discrete time [23, 127, 223] settings. Since ....

....or not. The reader familiar with the literature on Hennessy Milner logics (cf. Def. 2.7) for prebisimulation like relations (cf. Def. 7. 2 to follow) may have noticed that the notion of satisfaction for negative premises discussed above is akin to that for formulae of the form [a] given in, e.g. [1, 10, 154, 207, 70 209]. In those references, the new interpretation is necessary to obtain monotonicity of the satisfaction relation with respect to the appropriate notion of prebisimulation. The intuitionistic interpretation of negative premises given in [11, 12] is crucial to obtain operations that are monotonic with ....

[Article contains additional citation context not shown here]

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187.

....can often be expressed positively using pred icates [27] However, in the literature we see more and more that SOS definitions are decorated with predicates and or negative premises. For example, predicates are used to express matters like (un)successful termination, convergence, divergence [10], enabledhess [43] maximal delay, and side conditions [171] Negative premises are used to describe, e.g. deadlock detection [141] sequencing [58] priorities [24, 69] probabilistic behaviour [143] urgency [61] and various real [140] and discrete time [23, 131, 232] settings. Since ....

....or not. The reader familiar with the literature on Hennessy Milner logics (cf. Def. 2.7) for prebisimulatiomlike relations (cf. Def. 7. 2 to follow) may have noticed that the notion of satisfaction for negative premises discussed above is akin to that for formulae of the form [a]T given in, e.g. [1, 10, 160, 216, 218]. In those references, the new interpretation is necessary to obtain mono tonicity of the satisfaction relation with respect to the appropriate notion of prebisimulation. The intuitionistic interpretation of negative premises given in [11, 12] is crucial to obtain operations that are monotonic ....

[Article contains additional citation context not shown here]

L. ACETO AND M. HENNESSY, Termination, deadlock and divergence, J. Assoc. Cornput. Mach., 39 (1992), pp. 147-187.

....Moreover, negative premises can often be expressed positively using predicates [9] However, in the literature SOS de nitions are often decorated with predicates and or negative premises. For example, predicates are used to express matters like (un)successful termination, convergence, divergence [3], enabledness [14] maximal delay, and side conditions [42] Negative premises are used to describe, e.g. deadlock detection [38] sequencing [17] priorities [7, 21] probabilistic behaviour [39] urgency [19] and various real [37] and discrete time [6, 35] settings. Since predicates and ....

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147-187.

....appropriateness of denotational models for several process description languages. In particular, when proving full abstraction results for denotational semantics based on the Scott Strachey approach for CCS like languages, several preorders based on bisimulation have been considered; see, e.g. [6, 3, 4]. In this paper, we shall study one such bisimulationbased preorder whose connections with domain theoretic models are by now well understood, viz. the prebisimulation preorder . investigated in, e.g. 6, 3] Intuitively, p . q holds of processes p and q if p and q can simulate each other s ....

....q iff for every t 2 ST(Act) t . p implies t . q : 3) This is an immediate consequence of the fact that t . p iff t . p, for every finite synchronization tree t and process p. A binary relation over processes that enjoys property (3) is usually called finitary or finitely approximable [6, 4]. 6 Lemma 3.3 The preorder Before proving this lemma, we introduce some intermediate definitions and results. For every process p 2 Proc, finite action set A and non negative integer n, we define a synchronization tree p as follows: A;0) Delta = Omega (A;n 1) Delta X n a : ....

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187.

....Moreover, negative premises can often be expressed positively using predicates [9] However, in the literature SOS definitions are often decorated with predicates and or negative premises. For example, predicates are used to express matters like (un)successful termination, convergence, divergence [3], enabledness [14] maximal delay, and side conditions [42] Negative premises are used to describe, e.g. deadlock detection [38] sequencing [17] priorities [7, 21] probabilistic behaviour [39] urgency [19] and various real [37] and discrete time [6, 35] settings. Since predicates and ....

L. Aceto and M. Hennessy, Termination, deadlock and divergence, J. Assoc. Comput. Mach., 39 (1992), pp. 147--187.

No context found.

L. Aceto and M. Hennessy, Termination, deadlock and divergence, Journal of the ACM 39(1), 1992, pp. 147-187.

No context found.

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Technical Report 6/88, Sussex University, 1988.

No context found.

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Technical Report 6/88, Sussex University, 1988.

No context found.

L. Aceto and M. Hennessy. Termination, deadlock and divergence. Journal of the ACM, 39(1):147--187, January 1992.

No context found.

L. Aceto and M. Hennessy. Termination, Deadlock and Divergence. In Mathematical Foundations of Programming Semantics, LNCS 442, pages 301#318. Springer Verlag, 1989.

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