From owner-freebsd-doc Mon Jan 26 10:46:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA04124 for freebsd-doc-outgoing; Mon, 26 Jan 1998 10:46:32 -0800 (PST) (envelope-from owner-freebsd-doc@FreeBSD.ORG) Received: from gateman.zeus.leitch.com (gateman.zeus.leitch.com [204.187.61.193]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA04118 for ; Mon, 26 Jan 1998 10:46:22 -0800 (PST) (envelope-from woods@zeus.leitch.com) Received: from zeus.leitch.com (0@tap.zeus.leitch.com [204.187.61.10]) by gateman.zeus.leitch.com (8.8.5/8.7.3/1.0) with ESMTP id NAA11882; Mon, 26 Jan 1998 13:46:39 -0500 (EST) Received: from bitter.zeus.leitch.com (bitter.zeus.leitch.com [204.187.61.66]) by zeus.leitch.com (8.7.5/8.7.3/1.0) with ESMTP id NAA06386; Mon, 26 Jan 1998 13:46:27 -0500 (EST) Received: (woods@localhost) by bitter.zeus.leitch.com (8.8.7/8.6.9) id NAA17711; Mon, 26 Jan 1998 13:46:27 -0500 (EST) Date: Mon, 26 Jan 1998 13:46:27 -0500 (EST) Message-Id: <199801261846.NAA17711@bitter.zeus.leitch.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit From: woods@zeus.leitch.com (Greg Woods) To: freebsd-doc@FreeBSD.ORG Cc: R.E.Wolff@BitWizard.nl Subject: questions about validity of FAQ entry 4.9 regarding SIGSEGV X-Mailer: VM 6.38 under Emacs 19.34.1 Reply-To: woods@planix.com (Greg A. Woods) X-Mailer: ViewMail (vm) Version 6.38 with GNU Emacs 19.34.1 (i386--freebsd, X toolkit) of Wed Mar 12 1997 on han.cs.berkeley.edu Organization: Planix, Inc.; Toronto, Ontario; Canada Sender: owner-freebsd-doc@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I have a few comments regarding the FAQ entry: 4.9. My programs occasionally die with ``Signal 11'' errors. and the subsequent document refered to from this entry: http://www.bitwizard.nl/sig11/ This entry (and the document it refers to) suggests that SIGSEGV core dumps are most likely caused by memory (DRAM or cache) or other data errors (motherboard, cpu, etc.) Either this is a sad comment on the state of PC hardware or it is misleading at best. I admit that I'm somewhat biased against PC hardware because of it's shoddy design and bad reputation, but still.... 99.99% of the SIGSEGV failures I've *ever* encountered have been due to programmer error (i.e. they were true access violations due to wild pointers and such). Of course the percentage of time I've used PCs, esp. ones without true parity or even ECC memory vs. other kinds of hardware that does have ECC memory is rather low. I'd be very dismayed to hear that this experience does not hold true for PC users too. I'm very dismayed not to find any reference to parity or ECC in the FAQ entry and indeed no mention of ECC in the second document. My ancient Sun-3 computers were using ECC memory back in the mid 1980's even before memory was fast and dense enough to even worry about anything but alpha-particle radiation being much of a problem. I don't know what to suggest, other than to mention programmer error as a likely cause and to strongly recommend using true-parity & ECC memory (and a trusted high-integrity motherboard) on anything but a games playing machine. BTW the comp.sys.pc.hardware FAQ (part 1, 2.20-2.22) has a decent discussion of the merits of true-parity and ECC memory. -- Greg A. Woods +1 416 443-1734 VE3TCP Planix, Inc. ; Secrets of the Weird