Engineering topics
Publications and source records attributed to Campbell, Roy H..
This paper describes tools that we built to support the construction of an object-oriented operating system in C++. The tools provide the automatic deletion of unwanted objects, first-class classes, dynamically loadable classes, and class-oriented debugging. As a consequence of our experience building Choices, we advocate these features as useful, simplifying and unifying many aspects of system programming.
Conventional operating systems provide little or no direct support for the services required for an efficient persistent object system implementation. We have built a persistent object scheme using a customization and extension of an object-oriented operating system called Choices. Choices includes a framework for the storage of persistent data that is suited to the construction of both conventional file system and persistent object system. In this paper we describe three areas in which persistent object support differs from file system support: storage organization, storage management, and typing. Persistent object systems must support various sizes of objects efficiently. Customizable containers, which are themselves persistent objects and can be nested, support a wide range of object sizes in Choices. Collections of persistent objects that are accessed as an aggregate and collections of light-weight persistent objects can be clustered in containers that are nested within containers for larger objects. Automated garbage collection schemes are added to storage management and have a major impact on persistent object applications. The Choices persistent object store provides extensible sets of persistent object types. The store contains not only the data for persistent objects but also the names of the classes to which they belong and the code for the operation of the classes. Besides presenting persistent object storage organization, storage management, and typing, this paper discusses how persistent objects are named and used within the Choices persistent data/file system framework.
In this paper we present a method for designing operating systems using object-oriented frameworks. A framework can be refined into subframeworks. Constraints specify the interactions between the subframeworks. We describe how we used object-oriented frameworks to design Choices, an object-oriented operating system.
Object-oriented design and programming has many software engineering advantages. Its application to large systems, however, has previously been constrained by performance concerns. The Choices operating system, which has over 75,000 lines of code, is object-oriented and programmed in C++. This paper is a case study of the performance of Choices.
The current design of the CHOICES persistent object implementation is summarized, and research in progress is outlined. CHOICES is implemented as an object-oriented system, and persistent objects appear to simplify and unify many functions of the system. It is demonstrated that persistent data can be accessed through an object-oriented file system model as efficiently as by an existing optimized commercial file system. The object-oriented file system can be specialized to provide an object store for persistent objects. The problems that arise in building an efficient persistent object scheme in a 32-bit virtual address space that only uses paging are described. Despite its limitations, the solution presented allows quite large numbers of objects to be active simultaneously, and permits sharing and efficient method calls.
The Choices family of operating systems exploits class hierarchies and object-oriented programming to facilitate the construction of customized operating systems for shared memory and networked multiprocessors. The software is being used in the Tapestry laboratory to study the performance of algorithms, mechanisms, and policies for parallel systems. Described here are the architectural design and class hierarchy of the Choices virtual memory management system. The software and hardware mechanisms and policies of a virtual memory system implement a memory hierarchy that exploits the trade-off between response times and storage capacities. In Choices, the notion of a memory hierarchy is captured by abstract classes. Concrete subclasses of those abstractions implement a virtual address space, segmentation, paging, physical memory management, secondary storage, and remote (that is, networked) storage. Captured in the notion of a memory hierarchy are classes that represent memory objects. These classes provide a storage mechanism that contains encapsulated data and have methods to read or write the memory object. Each of these classes provides specializations to represent the memory hierarchy.
This paper describes a multiprocessor operating system simulator that was developed by the authors in the Fall semester of 1987. The simulator was built in response to the need to provide students with an environment in which to build and test operating system concepts as part of the coursework of a third-year undergraduate operating systems course. Written in C++, the simulator uses the co-routine style task package that is distributed with the AT&T C++ Translator to provide a hierarchy of classes that represents a broad range of operating system software and hardware components. The class hierarchy closely follows that of the 'Choices' family of operating systems for loosely- and tightly-coupled multiprocessors. During an operating system course, these classes are refined and specialized by students in homework assignments to facilitate experimentation with different aspects of operating system design and policy decisions. The current implementation runs on the IBM RT PC under 4.3bsd UNIX.
The Software Automation, Generation and Administration (SAGA) project is investigating the design and construction of practical software engineering environments for developing and maintaining aerospace systems and applications software. The research includes the practical organization of the software lifecycle, configuration management, software requirements specifications, executable specifications, design methodologies, programming, verification, validation and testing, version control, maintenance, the reuse of software, software libraries, documentation, and automated management.
Research into software development is required to reduce its production cost and to improve its quality. Modern software systems, such as the embedded software required for NASA's space station initiative, stretch current software engineering techniques. The requirements to build large, reliable, and maintainable software systems increases with time. Much theoretical and practical research is in progress to improve software engineering techniques. One such technique is to build a software system or environment which directly supports the software engineering process, i.e., the SAGA project, comprising the research necessary to design and build a software development which automates the software engineering process. Progress under SAGA is described.
A key component of the 'SAGA' metatool system for software development support is an integrated modular environment that furnishes a uniform view of software components and can be adapted for different programming languages and development methodologies. Attention is presently given to the model/representation of the entities thus used in software development; programs are structured on the basis of modules whose relationships form layers of abstraction. Different views of the program under development are provided for different development phases. The implementation of the modular environment based on the model of a UNIX system is presented.
The SAGA system is a software environment that is designed to support most of the software development activities that occur in a software lifecycle. The system can be configured to support specific software development applications using given programming languages, tools, and methodologies. Meta-tools are provided to ease configuration. The SAGA system consists of a small number of software components that are adapted by the meta-tools into specific tools for use in the software development application. The modules are design so that the meta-tools can construct an environment which is both integrated and flexible. The SAGA project is documented in several papers which are presented.
ENCOMPASS is an example integrated software engineering environment being constructed by the SAGA project. ENCOMPASS supports the specification, design, construction and maintenance of efficient, validated, and verified programs in a modular programming language. The life cycle paradigm, schema of software configurations, and hierarchical library structure used by ENCOMPASS is presented. In ENCOMPASS, the software life cycle is viewed as a sequence of developments, each of which reuses components from the previous ones. Each development proceeds through the phases planning, requirements definition, validation, design, implementation, and system integration. The components in a software system are modeled as entities which have relationships between them. An entity may have different versions and different views of the same project are allowed. The simple entities supported by ENCOMPASS may be combined into modules which may be collected into projects. ENCOMPASS supports multiple programmers and projects using a hierarchical library system containing a workspace for each programmer; a project library for each project, and a global library common to all projects.
The concept of unstructured structure editing and ted, an editor for unstructured trees, is described. Ted is used to manipulate hierarchies of information in an unrestricted manner. The tool was implemented and applied to the problem of organizing formal proofs. As a proof management tool, it maintains the validity of a proof and its constituent lemmas independently from the methods used to validate the proof. It includes an adaptable interface which may be used to invoke theorem provers and other aids to proof construction. Using ted, a user may construct, maintain, and verify formal proofs using a variety of theorem provers, proof checkers, and formatters.