. Many high-level language compilers generate C code and then invoke a C compiler to do code generation, register allocation, stack management, and low-level optimization. To date, most of these compilers link the resulting code against a conservative mark-sweep garbage collector in order to reclaim unused memory. We introduce a new collector, MCC, based on mostly-copying collection, and characterize the conditions that favor such a collector over a mark-sweep collector. In particular we demonstrate that mostly-copying collection outperforms conservative mark-sweep under the same conditions that accurate copying collection outperforms accurate mark-sweep: Specifically, MCC meets or exceeds the performance of a mature mark-sweep collector when allocation rates are high, and physical memory is large relative to the live data. 1 High Level Overview Languages such as C [1] and C++ [27] are GC-unfriendly because they allow programs to violate the key premise of tracing garbage collection--...

Object-based (i.e. classless) models are very effective for elucidating requirements from users, and they support exploratory programming and rapid prototyping. On the other hand, class-based models are preferred to perform design and implementation, providing descriptive power and some types of error checking. We consider the evolution of object-based models into class-based production systems. One of the most difficult problems of this transition is the change from explicit description of individuals to implicit description of class instances. Reorganization support aims to make the system structures and properties evident and enforceable. Structural properties are useful to organize the code regardless of its meaning, and automated support can be enlisted to identify potential structures and properties, leaving the programmer with choice of alternative workspace organizations. We analyze the organization and management of classless objects, regarding the goals of redundancy elimina...

This dissertation pushes the boundary between textual and visual expression in a new way. We focus on the issue of writing specifications using a visual notation, and we describe two visual languages for this purpose. These languages provide users with the ability to formally specify security properties of a file system. Relative to previous text-based approaches to security specification, our use of a visual notation yields specifications that are more clear and concise. Moreover, since we formally define the semantics of our languages, every well-formed picture also has a precise meaning, and each can be processed by a computer. Our thesis is that visual languages can be practical and useful, even on a large scale, if they are targeted to sufficiently restricted domains of interest. We focus on two different aspects of the security domain. First, we use the instance language to specify security configurations, that is, fixed access relationships between users and files on a file sys...

Reusable objects components are usually organized in libraries. Improving the reusability of such libraries passes by recognizing that there are two parties to component reuse, with di erent (sometimes con icting) interests: reuse providers, who privilege optimality of implementation (e.g. to minimize redundancy, or to maximize domain content), and reusers, who privilege identi cation of assets (by optimal domain representation). These con icting optimality criteria can be mirrored by having separate hierarchies for types (interfaces) and classes (implementation code). In this paper, we propose a \biological metaphor " for object-oriented hierarchies, borrow some taxonomic numerical techniques for hierarchy construction, and apply them to organize classes and types separate and optimally. In particular, we present a mapping mechanism between types and classes; introduce a set of criteria to minimize the number of classes needed to implement a given set of types; use cladistic algorithms to construct minimal hierarchies of either types or classes; and combine these strands into an iterative type/class co-design process. The application of the approach is illustrated by its equivalence in several object-oriented languages. To provide necessary background, we include a brief introduction to biological classi cation and to cladistics, specially techniques for hierarchy deduction and minimality criteria. c 1997 John Wiley & Sons 1.

ui n'est plus ambigu + @ @ @ @ @ @ @         a + > > > > > > > > ~ ~ ~ ~ ~ ~ ~ ~ b c ## + > > > > > > > > ~ ~ ~ ~ ~ ~ ~ + @ @ @ @ @ @ @ @         c a b Probleme: definir la syntaxe de facon que toutes les ambigu tes syntaxiques soient resolubles simplement par l'analyse syntaxique. Ce n'est pas les cas des vecteurs et fonctions FORTRAN, comme on verra plus avant. Semantique denotationnelle DEA'96 Roberto Di Cosmo 4 Semantique . semantique: Une fois bien defini quels sont les programmes syntaxiquement correctes, on doit pouvoir dire de facon precise et univoque ce que chaque programme "fait" a l'execution (ce qu'il calcule, ou comment il se conduit), y compris si l'execution donne lieu a des erreurs: cela est indispensable pour -- l'utilisateur -- l'implementateur

n'est plus ambigu + @ @ @ @ @ @ @         a + > > > > > > > > ~ ~ ~ ~ ~ ~ ~ ~ b c ## + > > > > > > > > ~ ~ ~ ~ ~ ~ ~ + @ @ @ @ @ @ @ @         c a b Probleme: definir la syntaxe de facon que toutes les ambigu tes syntaxiques soient resolubles simplement par l'analyse syntaxique. Ce n'est pas les cas des vecteurs et fonctions FORTRAN, comme on verra plus avant. Semantique denotationnelle DEA'96 Roberto Di Cosmo 4 Semantique . semantique: Une fois bien defini quels sont les programmes syntaxiquement correctes, on doit pouvoir dire de facon precise et univoque ce que chaque programme "fait" a l'execution (ce qu'il calcule, ou comment il se conduit), y compris si l'execution donne lieu a des erreurs: cela est indispensable pour -- l'utilisateu

Object-based (i.e. classless) models are very effective for elucidating requirements from users, and they support exploratory programming and rapid prototyping. On the other hand, class-based models are preferred to perform design and implementation, providing descriptive power and some types of error checking. We consider the evolution of object-based models into class-based production systems. One of the most difficult problems of this transition is the change from explicit description of individuals to implicit description of class instances. Reorganization support aims to make the system structures and properties evident and enforceable. Structural properties are useful to organize the code regardless of its meaning, and automated support can be enlisted to identify potential structures and properties, leaving the programmer with choice of alternative workspace organizations. We analyze the organization and management of classless objects, regarding the goals of redundancy elimina...