Return to BSD News archive
Newsgroups: comp.os.386bsd.development Path: sserve!newshost.anu.edu.au!munnari.oz.au!bunyip.cc.uq.oz.au!harbinger.cc.monash.edu.au!yeshua.marcam.com!usc!howland.reston.ans.net!cs.utexas.edu!convex!seas.smu.edu!utacfd.uta.edu!rwsys!hammy!gordon From: gordon@sneaky.lonestar.org (Gordon Burditt) Subject: Re: [FreeBSD 1.0R] DMA Problems? References: <jmonroyCIHJA2.oy@netcom.com> <jmonroyCIo4yD.1G5@netcom.com> <2fnapb$6eb@u.cc.utah.edu> <jmonroyCIvwpv.8F3@netcom.com> Organization: /usr/lib/news/organi[sz]ation Date: Tue, 4 Jan 1994 12:17:26 GMT Message-ID: <CJ3u5H.9JG@sneaky.lonestar.org> Lines: 28 > To clarify the issue: DMA DRAM refresh skipping would be a > bad term to use. What happens is the timer for channel 0, > the DRAM refresh timer, is reprogrammed so that there are Timer 0 is the DRAM refresh timer? Not on my system. (Timer 0 = timer interrupt, Timer 1 = refresh, Timer 2 = speaker) > It is possible, via software, to reprogram the refresh cycle. > My example program does this. The results are had when the > system crashes. That is, to make this a software *probe* > the program would (if it existed for *BSD) slowly turn down > the timer till the system crashed. At this point, some The proper way to do this is to test this while operating the environmental controls to extremes of temperature and power supply voltages, and using lots of patterns in RAM. Oh, your system doesn't have controls like that? This probe routine would likely take an hour to run an adequate test even at ONE set of temperatures and power supply voltages. That's a long time to boot. Sure, you can probably do it a lot faster if you have expensive analog measuring equipment attached to the system, but you don't. Also, remember that refresh failure doesn't guarantee a crash. Even if you test it properly, when it becomes known that your OS fiddles with refresh this way, you can be sure that the reaction when you come to a DRAM manufacturer and say "bad RAM", they'll start laughing. Gordon L. Burditt sneaky.lonestar.org!gordon