Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site utah-gr.UUCP Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!think!harvard!seismo!utah-cs!utah-gr!donn From: donn@utah-gr.UUCP (Donn Seeley) Newsgroups: net.lang.f77 Subject: Re: Re: G_floating in f77 under 4.2BSD? Message-ID: <1489@utah-gr.UUCP> Date: Thu, 27-Jun-85 03:47:54 EDT Article-I.D.: utah-gr.1489 Posted: Thu Jun 27 03:47:54 1985 Date-Received: Sat, 29-Jun-85 00:15:01 EDT References: <331@sbcs.UUCP> <1608@hao.UUCP> Organization: University of Utah CS Dept Lines: 18 I have heard rumors to the effect that the Ultrix compilers support at least the G floating format, as a result of trying to get software to run on the MicroVAX I when the latter had F and G floating point instead of F and D... Ask a DECcie. I know of no current plans to provide G and H support in 4.3 BSD compilers. (Maybe 4.4...) If you need G floating point but don't need D floating point, you could try interrupting the compiling process between the stages of compiling and assembly, and running 'sed' on the input to the assembler, converting instructions like 'cvtfd' to 'cvtfg' and pseudo-ops like '.double 0d1.0' to '.gfloat 0g1.0'. When this sort of postprocessing is embedded in a default rule in a makefile, it isn't as painful as it sounds. No guarantees that this works, of course... Donn Seeley University of Utah CS Dept donn@utah-cs.arpa 40 46' 6"N 111 50' 34"W (801) 581-5668 decvax!utah-cs!donn