Path: utzoo!mnetor!uunet!husc6!bloom-beacon!mit-eddie!uw-beaver!cornell!batcomputer!itsgw!imagine!pawl6.pawl.rpi.edu!jesup
From: jesup@pawl6.pawl.rpi.edu (Randell E. Jesup)
Newsgroups: comp.sys.amiga
Subject: Re: Unix on the Amiga
Message-ID: <857@imagine.PAWL.RPI.EDU>
Date: 6 May 88 08:40:12 GMT
References: <517@viscous> <51577@sun.uucp> <1923@sugar.UUCP>
Sender: news@imagine.PAWL.RPI.EDU
Reply-To: jesup@pawl6.pawl.rpi.edu (Randell E. Jesup)
Organization: RPI Public Access Workstation Lab - Troy, NY
Lines: 37
Keywords: Unix SYSV 4.2 Amiga

In article <1923@sugar.UUCP> peter@sugar.UUCP (Peter da Silva) writes:
>Here's one thing that will never, ever, be implemented on the Amiga...
>
>	pid = fork();
>	switch(pid) {
>		case -1: perror(argv[0]); exit(1);
>		case 0: ...
>		default: ...
>	}
>
>Don't tell me that AmigaDOS tasks are more efficient, or more powerful,
>or less filling. I think they taste great myself. The point remains that
>fork() is not compatible with them.

	I think they taste filling.  :-)

>Consider this a challenge, if you like... I don't think it's possible to
>implement fork on the Amiga in the general case.

>-- Peter da Silva      `-_-'      ...!hoptoad!academ!uhnix1!sugar!peter

	Challenge accepted.

	I went over all this when people wer talking about a minix port
for the amiga.  The way to do it is to copy the data segments and stack
to save areas, then spawn another process.  You'll have to modify the
task variables to swap the data/stacks if needed when one of the two
tasks gets CPU, plus some shenanigans for ending tasks.  (By modify the
task variables, I mean the code called when getting/losing CPU).

	I never said it would be pretty.

     //	Randell Jesup			      Lunge Software Development
    //	Dedicated Amiga Programmer            13 Frear Ave, Troy, NY 12180
 \\//	beowulf!lunge!jesup@steinmetz.UUCP    (518) 272-2942
  \/    (uunet!steinmetz!beowulf!lunge!jesup) BIX: rjesup
(-: The Few, The Proud, The Architects of the RPM40 40MIPS CMOS Micro :-)