Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!news.uoknor.edu!ns1.nodak.edu!netnews.nwnet.net!ogicse!uwm.edu!news.moneng.mei.com!howland.reston.ans.net!pipex!uunet!EU.net!sun4nl!philapd!apdnews!spoetnix.idca.tds.philips.nl!wilko From: wilko@idca.tds.philips.nl (Wilko Bulte) Newsgroups: comp.os.386bsd.questions Subject: Re: ISA strayintr 7 - please help!! Message-ID: <wilko.753607415@spoetnix.idca.tds.philips.nl> Date: 18 Nov 93 07:23:35 GMT Article-I.D.: spoetnix.wilko.753607415 References: <2c6nga$2e8@darkstar.UCSC.EDU> <wilko.753391395@spoetnix.idca.tds.philips.nl> <2cafgi$rqs@olivaw.apanix.apana.org.au> Sender: news@idca.tds.philips.nl Lines: 47 hart@apanix.apana.org.au (Leigh Hart) writes: >wilko@idca.tds.philips.nl (Wilko Bulte) writes: >>krostrin@cats.ucsc.edu (Ken Ostrin) writes: >>>I am running NetBSD 0.9 on a 386DX-40 with 8MB RAM and a mathco. When NetBSD >>>boots, I get the above error message several times and then the message "Too >>>many errors. will stop logging" >>>I have removed all cards except the hard disk controller. Has anyone else had >>>this problem? What causes it? How do I fix it? >>Essentially it unclean code in the area of handling the interrupt >>controller chips. You can't fix it, other than by uncommenting >>the message in the kernel sources. There is nothing wrong with >>your machine in general. >*ROFL* >It's got nothing to do with clean/unclean code in the interrupt handler >routines - the fact is there ISN'T a routine handling the interrupt, THAT >is what the kernel is complaining about! It has! If no device is configured there, and the hardware is OK then there should be no IRQs coming in. >If you look at any PC book, or even run diagnostic software on a DOS PC >you will quickly see that IRQ7 is the first parallel port (printer port). Sure. IRQ5 is also in use for LPT as an alternative. >Printer ports usually have no need for interrupt routines because they >very rarely are input devices. As a consequence, when the IRQ 7 line is >happily floating high and low whenever it feels like it, there's no routine >activated to service the interrupt. You have apparantly never seen correctly designed hardware. On interrupt lines there should always be a pullup resistor (assuming active low IRQs). Floating interrupt lines will _always_ cause trouble not only for int7. Wilko -- | / o / / _ Wilko Bulte mail: wilko@idca.tds.philips.nl |/|/ / / /( (_) |d|i|g|i|t|a|l| Equipment Corporation voice: +3155-432062 fax: +3155-432103 DTN: 829 2062 PO Box 245 - 7300 AE Apeldoorn - The Netherlands