From owner-freebsd-hardware Wed May 17 18:40:55 1995 Return-Path: hardware-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id SAA20449 for hardware-outgoing; Wed, 17 May 1995 18:40:55 -0700 Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id SAA20442 for ; Wed, 17 May 1995 18:40:52 -0700 Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id SAA13066; Wed, 17 May 1995 18:40:22 -0700 From: "Rodney W. Grimes" Message-Id: <199505180140.SAA13066@gndrsh.aac.dev.com> Subject: Re: bad memory.. To: karl@bagpuss.demon.co.uk (Karl Strickland) Date: Wed, 17 May 1995 18:40:22 -0700 (PDT) Cc: freebsd-hardware@FreeBSD.org In-Reply-To: <199505180040.BAA03026@bagpuss.demon.co.uk> from "Karl Strickland" at May 18, 95 01:40:07 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1365 Sender: hardware-owner@FreeBSD.org Precedence: bulk > > system: 486/33 isa, aha1542, 16mb > > added another 16mb and it went crazy with programs dumping core, and > kernel traps all over the place. so i suspect the memory is bad (this > machine has run with 20Mb previously, and has bounce buffers configured > etc).. > Try adding 1 wait state to the memory setup in the BIOS if you can. I found that my ECS EISA/VLB board would run on ``fastest'' with 4 4MB-60nS simms, but add 4 more and I had to slow it down to ``faster''. This is the same as adding 1 wait state to main memory cycle times. > On FreeBSD 1.x, the memory test doesnt complain. I remember reading > the BIOS's memory test is crap. What should I use to test memory? FreeBSD is one heck of a memory tester, but it does not point to the location of the failure. You *should* get an NMI for a parity error, but gross memory failure (ie 2 bit or more) is often missed by this type of logic. > (currently the machine has 4x4Mb 30 pin 9-chip simms, giving 16Mb. > the new 16Mb is 4x4Mb 30 pin 3-chip simms. that shouldnt cause > any problems right?) No, that should not cause a problem. Are the memory simms the same speed, and at least 70nS or faster? What motherboard is this? -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD