Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!pacbell.com!ucsd!ucbvax!SNOW-WHITE.MERIT-TECH.COM!goss From: goss@SNOW-WHITE.MERIT-TECH.COM (Mike Goss) Newsgroups: comp.sys.sgi Subject: Re: Multi-thread on iris Message-ID: <9102010006.AA18053@snow-white.merit-tech.com> Date: 1 Feb 91 00:06:30 GMT Sender: daemon@ucbvax.BERKELEY.EDU Organization: The Internet Lines: 23 > Date: Thu, 31 Jan 91 10:35:02 EST > From: "Dan Karron (karron@nyu.edu" > . > . > . > In fact, with one cpu is it easier to multi-thread because you > have 0 probability of collision enforced by the scheduler ! Not true! Although things can't really happen simultaneously, the wonders of preemptive scheduling mean that a process can be interrupted at any time (unless it has a non-degrading priority or some such thing), another process can be run which accesses the same data, and then the original process restored without knowing it was ever interrupted. The effect is that collisions are possible, although a bit less likely than on a multi-processor system. ------------------------------ Mike Goss Merit Technology Inc. (214)733-7018 goss@snow-white.merit-tech.com Disclaimer: This offer void except where prohibited by law.