Path: utzoo!attcan!uunet!samsung!usc!jarthur!bridge2!mips!hal!mark From: mark@mips.COM (Mark G. Johnson) Newsgroups: comp.arch Subject: Re: Program to discern machine architecture information Message-ID: <39380@mips.mips.COM> Date: 14 Jun 90 14:58:53 GMT References: Sender: news@mips.COM Reply-To: mark@mips.COM (Mark G. Johnson) Organization: MIPS Computer Systems, Inc. Lines: 22 For floating-point, the program MACHAR from netlib might be just the ticket: MACHAR machar.f Fortran 77 source. Both single- and double-precision versions can be extracted from the supplied source code with simple editing changes. NOTE: at least one version MUST be extracted before the source will compile. macharc C source for machar and a driver. Float and double versions are selected with compiler directives. MACHAR is an evolving subroutine for dynamically determining thirteen fundamental parameters associated with floating-point arithmetic. The original version was published in Cody and Waite, Software Manual for the Elementary Functions, Prentice-Hall, 1980. The present version has been modified to operate correctly with IEEE floating-point arithmetic. It will malfunction on many CRAY and most CYBER systems, however. See W. J. Cody, "MACHAR: A subroutine to dynamicall determine machine parameters," TOMS 14, December, 1988 for details. -- -- Mark Johnson MIPS Computer Systems, 930 E. Arques M/S 2-02, Sunnyvale, CA 94086 (408) 524-8308 mark@mips.com {or ...!decwrl!mips!mark}