Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!usc!pollux.usc.edu!addison From: addison@pollux.usc.edu (Richard Addison) Newsgroups: comp.sys.amiga.tech Subject: Re: MutualExclude for Gadgets Keywords: Amiga MutualExclude Message-ID: <20076@usc.edu> Date: 23 Sep 89 21:29:55 GMT References: <1849@cbnewsd.ATT.COM> <125013@sun.Eng.Sun.COM> <7964@cbmvax.UUCP> <125083@sun.Eng.Sun.COM> <7980@cbmvax.UUCP> Sender: news@usc.edu Reply-To: addison@pollux.usc.edu (Richard Addison) Distribution: usa Organization: University of Southern California, Los Angeles, CA Lines: 28 In article <7980@cbmvax.UUCP> jesup@cbmvax.UUCP (Randell Jesup) writes: > Looking at *0? Must we add 0-detect circuits in order to not >"support" it? Could do this with the MMU (wish I had one). Trap all access to that page of memory, and continue if it is an allowed read (like, oh I don't know, um, maybe location 4?) but complain if it is a read from location 0. Yeah, I know it slows things down, but it's a debugging tool. > Sorry if I sound testy, I just found some major stuff that was playing >evil games and calling internal BCPL routines that shouldn't have been. Grrrr. Gee, I was wondering when you guys would notice. (-; I have a question that is unrelated to the subject line, but this remark reminded me: I've noticed that some of the ROM libraries (at least those routines I've traced to find out what is really going on) push their parameters from the registers onto the stack. Considering that at least one C compiler for the Amiga eliminates the need to have stub routines, wouldn't it be nice if the ROM routines themselves could use the registerized parameters directly? Like maybe in 1.4? Just a thought. Richard Addison