Dynamic program behavior under paging

Current processors employ aggressive prediction mechanisms to improve performance and reduce power. It is increasingly important to understand and quantify a program's dynamic behavior to effectively design next-generation prediction mechanisms. In this paper, we develop algorithms and mechanisms inspired by DNA discovery tools to analyze and quantify program dynamic behavior in terms of regularities and patterns. We describe our PatternFinder tool and analyze its results to summarize most important branch and data address pattern behaviors for a set of program traces and SPEC CPU 2006 benchmarks. M aking the common case fast is a design principle that has been used in microprocessor design for decades. This principle applies when determining how to spend resources, since the performance impact on making some occurrence faster is higher if the occurrence is frequent. Although quantifying frequent behavior in an application's dynamic execution behavior is trivial in cases such as observing the frequency of each type of instruction, it is very challenging to summarize dynamic data reference behavior (1). As a finding overlapping or approximate patterns. In this paper, inspired by DNA discovery tools (5, 6), we adopt and revise the methods motivated by suffix trees (5) in order to develop comprehensive pattern discovery tools targeted for computer architecture. Suffix trees have several advantages over SEQUITUR in designing such a tool, which we discuss in Section 3. In this paper, we present a novel pattern analysis tool, the PatternFinder, which is designed to discover and quantify patterns, and the results produced by the tool that quantify data address and branch patterns in dynamic program behavior. PatternFinder is an offline analysis tool that can summarize input sequences using patterns, quantify spatial and temporal regularities in input sequences and enumerate hot patterns based on specified scoring criteria (e.g., coverage). It allows user customizable search with a number of input parameters. PatternFinder can be used to gain insights into hardware and software optimization opportunities as well as perform pattern-centric

System Development Corporation (SDC) has developed an informal approach for establishing correspondence (or lack thereof) between the formal specifications of a system and lower levels of specification, including implementation code. Establishing correspondence means showing that one level meets the intent of a higher lever. This paper presents the motivation, technique, and current SDC experience with the method, contains an example of the hypothetical correlation analysis, and evaluates the limits and the usefulness of this approach. This paper results from recent applications of the method.

Since June 1963, a Time-Sharing System has been operational at the System Development Corporation in Santa Monica. This system was produced under the sponsorship of ARPA and has utilized ideas developed at both Massachusetts Institute of Technology and Bolt, Beranek, and Newman, as well as some original techniques. Time-sharing, in this case, means the simultaneous access to a computer by a large number of independent (and/or related) users and programs. The system is also general purpose, since there is essentially no restriction on the kind of program that it can accommodate. The system has been used for compiling and debugging programs, conducting research, performing calculations, conducting games, and executing on-line programs using both algebraic and list-processing languages.

Parameter mining of traces identifies formal properties that describe a program's dynamic behaviour. These properties are useful for developers to understand programs, to identify defects, and, in general, to reason about them. The dynamic behavior of programs typically follows a distinct pattern of calls and returns. Prior work uses general logic to identify properties from a given set of templates. Consequently, either the properties are inadequate since the logic is not expressive enough, or the approach fails to scale due to the generality of the logic. This paper uses nested words and nested word automata that are especially well suited for describing the dynamic behaviour of a program. Specifically, these nested words can describe pre/post conditions and inter-procedural data-flow and have constant memory requirements. We propose a framework for mining properties that are in the form of nested words from system traces. As a part of the framework, we propose a novel scalable algorithm optimized for mining nested words. The framework is evaluated on traces from real world applications.

Supporting source-level performance analysis of programs written in data-parallel languages requires a unique degree of integration between compilers and performance analysis tools. Compilers for languages such as High Performance Fortran infer parallelism and communication from data distribution directives, thus, performance tools cannot meaningfully relate measurements about these key aspects of execution performance to source-level constructs without substantial compiler support. This paper describes an integrated system for performance analysis of data-parallel programs based on the Rice Fortran 77D compiler and the Illinois Pablo performance analysis toolkit. During code generation, the Fortran D compiler records mapping information and semantic analysis results describing the relationship between performance instrumentation and the original source program. An integrated performance analysis system based on the Pablo toolkit uses this information to correlate the program's dynamic behavior with the data parallel source code. The integrated system provides detailed source-level performance feedback to programmers via a pair of graphical interfaces. Our strategy serves as a model for integration of data-parallel compilers and performance tools.

There are important classes of programming errors that are hard to diagnose, both manually and automatically, because they involve a program's dynamic behavior. This article describes a compile-time analyzer that detects these dynamic errors in large, real-world programs. The analyzer traces execution paths through the source code, modeling memory and reporting inconsistencies. In addition to avoiding false paths through the program, this approach provides valuable contextual information to the programmer who needs to understand and repair the defects. Automatically-created models, abstracting the behavior of individual functions, allow inter-procedural defects to be detected efficiently. A product built on these techniques has been used effectively on several large commercial programs. Copyright © 2000 John Wiley & Sons, Ltd.

Annual Review of Information Science and Technology

Everyone talks about the computer user, but virtually no one has studied him in a systematic, scientific manner. There is a growing experimental lag between verified knowledge about users and rapidly expanding applications for them. This experimental lag has always existed in computer technology. Technological innovation and aggressive marketing of computer wares have consistently outpaced established knowledge of user performance—a bias in computer technology largely attributable to current management outlook and practice. With the advent of time-sharing systems, and with the imminence of the much-heralded information utility, the magnitude of this scientific lag may have reached a critical point. If unchecked, particularly in the crucial area of software management, it may become a crippling humanistic lag—a situation in which both the private and the public use of computers would be characterized by overriding concern for immediate machine efficiency and economy, and by an entrenched neglect of human needs, individual capabilities, and long-range social responsibilities.

Although online searching has been available to libraries since the late 1960s, it is only in very recent years that a significant number of academic institutions have begun offering information retrieval through commercial vendors such as Dialog, BRS (Bibliographic Retrieval Service), and SDC's (System Development Corporation's) Orbit. Because online searching is so new compared with traditional library services, it is still developing as a standard offering in most academic libraries.

This paper reviews the introduction of factory concepts and practices, based on tools and methods from the evolving field of software engineering, at major software producers, in particular those that explicitly adopted the factory label to describe their software facilities or approach to software development: Hitachi, Toshiba, NEC, and Fujitsu in Japan, as well as System Development Corporation in the United States. The other United States firm discussed in detail is International Business Machines, which, without adopting the factory label, introduced numerous measures to organize and control software development, especially basic software. The paper emphasizes that the difficulty of the technology, shortages of skilled engineers, and large-scale projects have encouraged producers to become more systematic or factory-like in managing a series of projects, even though some characteristics of the technology and the industry have made software seem difficult to control and more suitable to a loosely structured project-centered or craft approach to development.

In April, 1979, the 20th General Assembly of the Japan Medical Congress was held in Tokyo. The Assembly had about 120 scientific lectures and symposiums in which medical doctors and other health professionals participated from all over the country.In addition to those meetings, the General Medical Exhibition was arranged and various programs were provided such as research progresses, medical instruments, pharmaceutical drugs, and information/communication systems.“Dynamics of Medical Information Services in Future” was shown at a greater part of the Exhibition Place.Presented there were the Japan Information Center of Science and Technology's on-line information system (JOTS), and the Lockheed Information Service (DIALOG) and the System Development Corporation's Search Service through international telephone-lines. As useful devices of inter-library loan service, telexes and tele-facsimiles are demonstrated.Our Committee of Medical Information/Communication Systems suggested a future net-work of delivery of medical information at the nation-wide level, including a plan of a national medical library and information center.In commemoration of the Assembly, a 130-page book “How to find out medical literature (Igaku Bunken no Sagashikata)” was published.

John McCarthy designed the LISP programming language for the symbolic computation typical of artificial intelligence at the time, and LISP was strongly associated with AI at least through the 1990s. By 1963, a number of ideas had been proposed for how LISP 1.5 could be improved and extended. This lead to the LISP 2 project, which was carried out at System Development Corporation by personnel from SDC and Information International, Inc., using the military AN/FSQ-32 computer. LISP 2 extended LISP 1.5 with ALGOL-like syntax, static type checking, efficient support for arithmetic and arrays, pattern-matching, and more, but the LISP 2 language failed to displace LISP 1.5. Despite this, LISP 2 foreshadowed later languages, such as ECL, Common Lisp, and Java. An online archive of LISP 2 documents and source code exists at the Computer History Museum, providing a snapshot of mid-1960s ARPA-funded research and development.

Karen Sparck Jones' Cambridge PhD thesis of 1964 has had an interesting and unusual history. Entitled Synonymy and Semantic Classification (henceforth SSC) it was reproduced only in the simple mimeo book form then used by the Cambridge Language Research Unit where she worked. It was finally published in 1986, in an Edinburgh University Press series. Even that late publication managed to be ahead of a great deal of later work that recapitulates aspects of it, usually from ignorance of its existence. There is no doubt that SSC was developing statistical and symbolic techniques for the use of what we now call language resources so far ahead of other work that it was almost impossible for contemporary researchers to understand the book or to relate it to their own activity. At the time SSC was being written, Olney and Revard (1968) were exploring the content of Webster's Third Dictionary quantitatively on punched cards at Systems Development Corporation (where Sparck Jones also was by chance at the time and joined in their work with her 1967) and their work met a similar lack of reception, it, too, being twenty to thirty years ahead of its time.

At System Development Corporation there is a conviction that one of the most plausible ways to make the cost of software decline is to build general-purpose software that is capable of solving a variety of problems. SDC's most successful effort in this field has been in the area of general-purpose data management. Our initial large-scale, time-shared data management system, TSS-LUCID, enabled the nonprogrammer to describe, load, query, and maintain a data base. In use for over two years, this system provided enough generality to solve such diverse problems as comparison of salary data in different segments of the aerospace industry, analysis of statistical data for a customer in the oil industry, and monitoring of public-supported cancer research projects. Currently being implemented on the IBM 360 family of computers is an improved, more powerful version called TDMS (Time-Shared Data Management System).

Interactive computer systems establish a dialog between the man and the machine. By means of displays, the user receives immediate feedback of the results of his actions, and he is able to modify his decisions in order to obtain a system response that is most relevant to his needs.BOLD (Bibliographic On‐Line Display) is an example of a highly automated interactive document storage and retrieval system that is in operation at System Development Corporation. It enables the user to browse through the data base by subject categories or search for specific documents. The response is rapid and the results are satisfying.