Return to BSD News archive
Newsgroups: comp.unix.bsd.freebsd.misc Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!pravda.aa.msen.com!spool.mu.edu!howland.reston.ans.net!news.sprintlink.net!in1.uu.net!news.iij.ad.jp!tyo2.gate.nec!newssv1-nec!dsp.cl.nec!necscc!csl.cl.nec!csls6!sakai From: sakai@csle59.csl.cl.nec.co.jp (Junji SAKAI) Subject: Re: 950726 Page Fault in Kernel In-Reply-To: j@bonnie.heep.sax.de's message of 28 Aug 1995 09:42:09 +0200 Message-ID: <SAKAI.95Aug30094611@csle59.csl.cl.nec.co.jp> Sender: news@csl.cl.nec.co.jp Organization: C&C Labs., NEC Corp., JAPAN References: <SAKAI.95Aug22002915@csle59.csl.cl.nec.co.jp> <41rs0h$poa@bonnie.tcd-dresden.de> Date: Wed, 30 Aug 1995 09:46:10 GMT Lines: 39 In article <41rs0h$poa@bonnie.tcd-dresden.de> j@bonnie.heep.sax.de (J Wunsch) writes: S> Fatal trap 12: page fault while in kernel mode S> fault virtual address = 0xf53cb708 (*) S> fault code = supervisor read, page not present : S> panic: page fault : S> Items with asterisk(*) vary case by case, but others are the same S> in all case. This panic happens almost every time I compile something. J> ...this rather smells like a hardware problem. Try relaxing the J> conditions for your RAM/cache (or EDO) RAM etc., or try to make the J> cache write-through. There are ASUS boards known to have a broken J> chipset, but i've forgot which one. Thank you for your response. When I dropped CPU clock down to 120MHz, most of the traps disappeared. (my CPU is printed as Pentium133 :-( ) So I bought a Peltier module and got it attached to my Pentium. This time, it is almost always OK at 120MHz, and generally good even at 133MHz (though more likely to trap than at 120MHz). J> I assume every memory-expensive program would cause this, the compiler J> is only just the best known expample. Try troffing a large .ms file, J> i guess you'll experience the same. I see. But it seems it is likely to trap when there are some disk access. With the Peltier module, the whole x11perf test has passed at 133MHz. Once the system suddenly stopped and rebooted while x11perf, as soon as it happened to start fsck (by /etc/daily). Does the IDE or NCR SCSI driver have some critical-timing routines ? or 33MHz PCI bus may be too fast for IDE/SCSI cards ? --- Junji SAKAI