Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msunews!uwm.edu!chi-news.cic.net!newsfeed.internetmci.com!howland.reston.ans.net!nntp.crl.com!crl.crl.com!not-for-mail From: ggrant@crl.com (Gary E. Grant) Newsgroups: comp.unix.bsd.freebsd.misc Subject: Re: Am I having Mother Board problems? (long) Date: 12 Nov 1995 17:08:10 -0800 Organization: CRL Network Services (415) 705-6060 [Login: guest] Lines: 64 Message-ID: <4865pq$psp@crl.crl.com> References: <480g74$s8o@central.ro.com> NNTP-Posting-Host: crl.com In article <480g74$s8o@central.ro.com>, Mike R. Prevost <mprevost@ro.com> wrote: > >Hi. I've been running Linux for a long time and (for several reasons) I am >trying to bring up a FreeBSD server (to run a pretty good sized news >server on). I think I may be having weird motherboard problems and would >like to get a second opinion before hassling the vendor. > >486dx4/100 (AMD I think) ISA/PCI >3Com 3c509 >Adaptec 2940 (PCI, not wide) >1x1GB Conner SCSI disk (for / swap and news history and temp space) >4x2GB Seagate Baracuuda SCSI disks >32 MB RAM > >I have installed several versions of FreeBSD (2.0.5-RELEASE, >2.1-11??95-SNAP, 2.1.0-STABLE, etc) and recompiled the kernel to support >just the devices I have but it ALWAYS hangs during boot. It hangs right >after is says 'probing for devices on ISA bus' (or whatever). However, >it ALWAYS will do fine if I either 1) type '-v' and the boot: prompt, or >2) type '-c' at the boot: prompt and immediatley type 'quit' at the >config prompt. > >But wait.... There's more.... > >When I first install FreeBSD (2.0.5) I had problems with processes dying >from signal 11 errors. I also had files that would somehow get typos in >them without my intervention. Like compiling the kernel -- they had typos >all in them. Somtimes I would download a tar.gz file and it would test >out fime (with -tzvf), but then after doing something else I go back and >the file is corrupted. I read the FAQ (like a good boy) and saw that bad >memory causes the signal 11 problems and figured that it would also cause >my other problems. I had the vendor bring a hardware SIMM tester and sure >enought the 32 MB SIMM tested bad. He put 2x16MB simms (no-parity -- only >ones he had on him) in there (tested ones). > >Still more... > >So I got brave and installed a 2.1.0 snapshot (11??95). It was working >ok, but processes started dying from (of all things) SIG_BUS errors. I >have very frequent kernel "panics" becauseof kernel page faults. I even >saw an "privileged instruction" fault. (these were from trap() evidently). > >I'm at a loss of what to do next. If you have any ideas at all, please >mail me or post here. > >Thanks. > > >--- Mike R. Prevost > mprevost@ro.com What it looks? like to me is a hardware problem, specifically in the Cache Tag Ram. I have seen this problem before, and when I replace the Cache Rams (9 Chips) the problem goes away. To verify that this indeed is your problem: Go into the CMOS write on paper the original settings ( so you can put them back if this doesnt fix the problem) turn OFF all caching, and ROM Shadowing. Then if the problem goes awaym the problem is in the CACHE subsystem. Else you could try re-seating the Memory SIMS. also clean the edge connectors , with an eraser and brush, for the ISA/EISA Bus. Gary :-(