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!nntp.coast.net!zombie.ncsc.mil!news.mathworks.com!newsfeed.internetmci.com!news.io.com!news.fc.net!obiwan!bob From: bob@luke.pmr.com (Bob Willcox) Subject: Re: Strange new errors with 32mb upgrade Sender: news@obiwan.pmr.com (News Subsystem) Organization: Bob's Place, Austin TX Message-ID: <DurFCv.yu@obiwan.pmr.com> References: <4sbhat$q7c@skydiver.jaguNET.com> X-Nntp-Posting-Host: luke.pmr.com Date: Thu, 18 Jul 1996 22:19:43 GMT Lines: 54 In article <4sbhat$q7c@skydiver.jaguNET.com>, Sean Emery <emerys@confucius.omniscient.com> wrote: >Hello all, I have been getting several strange effects after adding some >memory to my system. (total of 48mb now.) > >First thing I noticed is that vi (and it's derivitives nvi, etc.) and talk >would core dump upon invocation. reboot fixed. > >Second thing, find command always produced a bus error. Reboot fixed. > >Third thing, a 'du' locked up my system and necessitated manual fsck'ing >upon reboot. > >Fourth thing, X sometimes exits prematurely with a buserror or segfault. >Not usually on its own, though, when I've done something (new window or >something) to upset it. After this happens, X exits almost immediately >with the same error when restarted (until reboot, of course.) > >Fifth thing, have rebooted my machine left the room to get a coke and >returned to find it rebooting again... weird, huh? > >I bought 2 4x32mb simms (rated at 60ns although I use them at 70) and >added them to the four 4MB sims that I already had. Is this the >possible effects of defective RAM? or should I recompile the kernel with >the option of more than 32 MB of memory or something like that. When one of my systems start behaving this way it has *always* been hardware related. Usually its been the memory, but I've had L2 cache failures and (what I believe to be) internal cache failures also cause it. Generally, when upgrading my memory, CPU and/or MB I will recompile the kernel in a loop for a few hours and watch for problems. These usually show up as segment violations or the like. Bus errors and segment violations when compiling programs could well be memory problems, especially if on recompiling them they compile ok. You could try changing some of your CMOS chipset settings, though I'm afraid that without knowing your specific motherboard and BIOS I can't be of much specific help there. Basically, try adding wait states to your memory accesses and/or try turning off your internal/external caches to see if that helps. > >Thanks, > >Sean Emery >emerys@confucius.omniscient.com > -- Bob Willcox bob@luke.pmr.com Austin, TX