Distributed Data Structures in the 4P Paradigm

. Distributed data structures are those that are shared by parallel processes. They provide a flexible and useful mechanism for parallel programming. This paper shows how distributed data structures can be programmed in a simple paradigm called 4P (short for procedures with preconditions as parallel programs). In the 4P paradigm, processes can specify different modes of access to the required parts of shared data structures. This provides the advantages that data structures are declared as in sequential programs, and copying is not necessary to achieve mutual exclusion. 1. Introduction In the traditional client/server model of languages like Ada 1 [1], a server task encapsulates a data structure [2]. Clients can access the data structure through the operations (entries, in the terminology of Ada) provided by the server; the semantics will allow only one client to be served at a time. One of the advantages of this model is that it uses the shared memory paradigm which programmers are ..