From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID autolearn=no autolearn_force=no version=3.4.4 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 103376,27e79426c7fe2b67 X-Google-Attributes: gid103376,public From: ka@socrates.hr.att.com (Kenneth Almquist) Subject: Re: GNAT 2.04 Linux binaries? Date: 1995/04/21 Message-ID: #1/1 X-Deja-AN: 101370095 sender: news@nntpa.cb.att.com (Netnews Administration) references: <3mm4vn$f52@mica.inel.gov> organization: AT&T newsgroups: comp.lang.ada Date: 1995-04-21T00:00:00+00:00 List-Id: vladimir@speedy.intrepid.com (Vladimir Vukicevic) asks: > Or did someone just go off on a whim and say "Hey! Let's make > setjmp a macro, and break interfacing to every other language! Yeah!"? There was a machine on which implementing setjmp as a function was difficult if not impossible. (I don't remember the machine or the details; it was something to do with not being able to access the frame pointer for anything besides the current procedure.) Calling setjmp from a language other than C is not a particularly good idea anyway; for example if the interface is implemented using stub routines setjmp won't work at all. The difficulty GNAT faces is that the method of performing a setjmp is machine-specific but is not specified in the machine descriptions used by GCC. Kenneth Almquist