Xref: utzoo comp.os.minix:5742 comp.sys.atari.st:16143
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!unmvax!pprg.unm.edu!hc!lll-winken!uunet!portal!cup.portal.com!DeadHead
From: DeadHead@cup.portal.com (Bruce M Ong)
Newsgroups: comp.os.minix,comp.sys.atari.st
Subject: Re: Real time clock from MINIX?
Message-ID: <18018@cup.portal.com>
Date: 6 May 89 04:51:30 GMT
References: <2398@brahma.cs.hw.ac.uk>
Distribution: usa
Organization: The Portal System (TM)
Lines: 48

I got a question...

	is minix's sleep() broken, or it's because I dont know
	exactly what sleep does?

	I have two programs

	--- test1.c ---
	main()
	{
		sleep(50);
	}

	--- test2.c ---
	main()
	{
		int i = 0;
		while (i < 10)
		{
			++i;
			print("\nGoing to sleep for a while");
			sleep(3);
			printf("\nI am now awake");
		}
	}

	if you execute the programs like this
	$test1 &
	130
	$test2

	Then test2 will terminate after looping few times, but
	test1 will never wake up...

	I havent tried it on other unix systems, yet... am I doing
	something wrong with sleep()?  I remember reading somewhere
	in the text saying the wake up call does not know which one
	to wake up, it just sends a wake up signal.. is that why?

	I am using sleep() to implement a printer spooler (the daemon
	scans the directory every several seconds); and I am also
	using creat(2) to create queue locks - if it cant lock a queue file
	it will sleep 2 seconds or so and try again. But after
	running the queue program, the printer daemon just slips
	into a coma...  Does anybody else have the problem?  I am using
	MINIX 1.2 for the AT.

	bruce	 	deadhead@cup.portal.com