IEFBR14

Abstract: IBM's High Level Assembler (HLASM) is a low level programming language for z/Architecture mainframe computers. Many legacy codebases contain large subsets written in HLASM for various reasons, and such components usually had to be manually rewritten in COBOL or PL/I before a migration to a modern framework could take place. Now, the Raincode ASM370 compiler for .NET supports HLASM syntax and emulates the data types and behaviour of the original language, allowing one to port, maintain and interactively debug legacy mainframe assembler code under .NET.

Abstract: The purposes of this book are to give the reader a basic understanding of the IBM mainframe architecture from the perspective of the instruction set and to serve as an assembly language text for introductory college courses in computer organization. Many computer architectures have come and gone in the past 50 years and only a relatively few designs remain. While most computer users are familiar with the ubiquitous Intel x86 microprocessor line, few have had much experience with another widely used architecture, the venerable IBM mainframe. In the not too distant past it was assumed that the rise of microprocessors and server centric computing would bring an end to the age of big iron (as mainframes were referred to). In the early 90s the rapid rise of companies such as Sun, Apollo, Silicon Graphics, MIPS, Digital Equipment Corporation (DEC), Tandem, Prime, Data General, SDS (also knows as XDS) and many more seemed to point to a world without the mainframes. All these companies, however, are now gone. Likewise vanished are the other mainframe manufacturers: NCR, Burroughs, Control Data Corporation, General Electric, Honeywell, RCA and UNIVAC. But the IBM mainframe architecture has endured and, rather than declining, is growing in popularity and becoming the platform of choice for such rapidly emerging technologies as cloud computing and virtualization. As the world chokes in an energy and space intensive microprocessor based server sprawl, a single mainframe system can deliver thousands of virtual Linux platforms for a fraction of the cost in dollars, energy and space. Now, however, the baby-boom generation programmers who wrote the billions of lines of legacy mainframe assembly code since the introduction of the IBM 360 in 1964 are beginning to retire. Consequently, the demand for replacement programmers to maintain, support and extend these applications as well as develop new ones is growing rapidly. To those looking to meet this need, this text will serve as a basic road map and introduction. The other purpose of this book is to provide an alternative assembly language platform for introductory college level courses in computer organization. From an academic perspective, the question of which architecture to use poses several challenges. Once, the answer was simple: the Intel x86 assembly language. However, as the Intel x86 architecture has grown into an intricate and unwieldy collection of exceptions and special cases, its value as a viable pedagogical platform in an introductory course has diminished. So, in many cases, the choice defaults to an assembly language based on a simulator for a non-existent, abstract RISC based architecture. Although this gives students a taste of a machine level architecture, it provides little real experience that can be used to actually further their careers and is often regarded by students themselves as an exercise in futility. For most, the basic impediment to teaching mainframe assembly language is actual access to a mainframe. While many colleges and universities use mainframes, students seldom have access to these. This text, however, is designed to be used in combination with a free, portable, open sourced, GPL licensed, Java based z390 mainframe emulator and macro assembler which permits a student to assemble and execute IBM mainframe programs on a Windows based PC2. This text concentrates on the basic architecture in order to keep to a manageable size suitable for a 5 week section in a computer organization course. While there are a number of instructions and features not discussed, these are, for the most part, logically intuitive extensions of the basic architecture discussed here.

Abstract: The assembly language course at Georgia State University is taught using a PDP-11/40 Assembler and Simulator, written in Pascal, running on a Sperry UNIVAC mainframe. In the past two years most of the computer science program has been switched to IBM PC compatible microcomputers. This, coupled with the fact that the UNIVAC system is about to be retired, motivated us to translate the PDP-11/40 system to the microcomputers. The translation is from the mainframe Pascal version to Ada for the PC's. The use of the assembler and simulator, the rationale for the use of Ada, and the problems encountered in the translation are all discussed.

Abstract: To an Ada-knowledgeable Software Engineer, the additional information that a production quality financial application can be delivered in Ada is probably pleasant and reassuring, but probably not surprising. I hope you find this paper informative and not too tedious. It is the Information Systems professional faced with the trauma of delivering that financial application , in Ada for the first time in that environment, that this paper seeks to assuage. Modern Information Systems software development requires the ability to support a large number of concurrent, interactive users. To be sure, there is still a requirement for scheduled batch processing , but usually, most of the development effort is of the interactive variety. Interactive computing on the IBM mainframe is old news. IBM provided a very powerful (if inelegant by modern standards) teleprocessing monitor targeted to its own line of mainframe computers called CICS, an acronym for Customer Information and Control System. CICS solved the problems of concurrent file access, record locking, tasking and terminal control for any number of users. CICS runs under the control of the operating system, for the purposes of this paper, IBM's MVS/XA. At the time CICS was introduced , it was the best game in town. CICS easily interfaced with the popular languages on that platform, including COBOL, Assem-bler and PL/1, and it was, and still is, well supported by the vendor. CICSts popularity continues for several reasons including: (1) a large installed base; (2) CICS'S relative ease of use; (3) the availability of programming talent; (4) vendor support and (5) third party tool support. Johnson [JOHN91] states IBMIs CICS group generates $325 million annually from 35,000 worldwide licenses, including 99% of the Fortune 2000. Johnson continues by giving the incredible figure of 300 plus billion lines of code that have been written for CICS and that there 250,000 programmers who know CICS. Clearly, any interactive system development targeted to an IBM mainframe must consider CICS. Until 1987, there was no requirement for a CICS-Ada interface. After all, CICS projects center on Information Systems and Ada was not intended to deliver Information Systems. However, the STANFINS–Redesign project changed all that. STANFINS-Redesign is an acronym Permission tocopytithout fee allorpartofthh material is granted providadthatthe copiesarenotnrade ordistributedfor diractconurercial 415 advantage, the ACM copyright notice and the title of the publition and its date appear, and notice is gwen that copying IS by perrmssion of the Aasocfation for Comp@ngMachine~. …