From owner-freebsd-questions Wed Jun 26 07:07:47 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id HAA10085 for questions-outgoing; Wed, 26 Jun 1996 07:07:47 -0700 (PDT) Received: from ns3.mke.ab.com (ns3.mke.ab.com [130.151.86.191]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id HAA10076; Wed, 26 Jun 1996 07:07:30 -0700 (PDT) Received: from by ns3.mke.ab.com (4.1/SMI-4.1) id AB26296; Wed, 26 Jun 96 09:07:10 CDT Received: from pegasus.ven.ra.rockwell.com (pegasus.ven.ra.rockwell.com [130.151.17.156]) by zeus.ven.ra.rockwell.com (8.6.12/8.6.9) with SMTP id KAA14047; Wed, 26 Jun 1996 10:02:40 -0400 Date: Wed, 26 Jun 1996 10:02:40 -0400 Message-Id: <2.2.16.19960626100546.362f66d6@zeus.ven.ra.rockwell.com> Organization: Rockwell Automation de Venezuela X-Sender: eparis@zeus.ven.ra.rockwell.com X-Mailer: Windows Eudora Pro Version 2.2 (16) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: questions@freebsd.org, hardware@freebsd.org From: "Eloy A. Paris" Subject: (SOLVED!) My Cyrix microprocessor/motherboard problems Cc: Gary Chrysler , "Lenzi, Sergio" , "M.R.Murphy" , "Michael L. VanLoon -- HeadCandy.com" , sos@freebsd.org, Terry Lambert , Didier Derny , Alex Nash , "Andrew V. Stesin" , "Brant M. Katkansky" , "David Alderman" , "Jacob M. Parnas" Sender: owner-questions@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hello everyone. Well, finally the problems with my Cyrix 486DX4/100 and my motherboard seem to be solverd. As you know, I was under severe attack of signals 10 (bus error) and 11 (segmentation fault) - almost all of them are, and this made my system too unstable to work as a dedicated server. In my last message I made a plan to find out the causes of my problem. These are the results: > 1) Replace the Cyrix 486DX4 with a genuine Intel DX4. OK, I got a genuine Intel 486DX4-100. I set up the jumpers of my motherboard accordingly and guess what? I got the segmentation fault when executing "man -a /usr/local/bin/*| col -b > /tmp/junk_test". This took me to the next step of my plan: >2) ... I am going to take my IDE harddrive with FreeBSD to another machine with an >Intel DX2/66 with a motherboard I know that works (it's a Digital computer, like one >that has been running Linux for about a year now with no problems - very good quality). I took the IDE hard drive with FreeBSD to a Digital DECpc LPx 466d2, with 16 Megs. of RAM and a 486DX2-66. Again, when trying "man -a ..." I got the segmentation fault. This really confussed me and made me suspect the problem was not hardware since I really trust the Digital mother board. The third step was not needed since step #2 failed: > 3) Get a different motherboard, one of good quality. I'll try to test the Cyrix DX4 on > it just to know if the problem is the microprocessor or the motherboard. I decided to made a final test to be sure the problem was not hardware: I repeated the same "man -a /usr/local/bin/* ..." test on my Dell Latitude XPi notebook with a genuine Intel Pentium @ 90 MHz. I had been running FreeBSD 2.1.0-RELEASE on this machine for more than 6 months with no problems. Initially I did not get the segmentation fault but when I looked at the /usr/local/bin directory I saw the Elm files were not there, as compared to the FreeBSD in the installation I was problems with. I installed Elm in the notebook and I GOT THE SEGMENTATION FAULT. I upgraded to -stable and I still was getting the fault. Finally, I determined that the command "man -a elm" was causing the faults. Summarizing, I found two causes for the segmentation faults: 1) I was using to test my configuration a command ("man -a elm") that has some kind of bug (I guess). 2) I was configuring the jumpers of the mother board for a Cyrix DX4-100, as is my case. If I do this I get more segmentation faults than ever, beginning with the modload command that loads the screen saver module. To overcome this situation I configured the jumpers for a genuine Intel DX4-100. After I did this my system is stable!!! :-) I know this does not make sense (to have a Cyrix working in a mother board set up for an Intel) but it works. Well, I hope this helps people having the same problems I had. Thank you all people that helped me with suggestions and fresh ideas. See ya! E-loy.- -- Eloy A. Paris Global Technical Services Rockwell Automation de Venezuela Telephone: 58-2-9432311 Fax: 58-2-9430323