From owner-freebsd-stable Wed Feb 3 07:46:30 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA26935 for freebsd-stable-outgoing; Wed, 3 Feb 1999 07:46:30 -0800 (PST) (envelope-from owner-freebsd-stable@FreeBSD.ORG) Received: from destiny.erols.com (destiny.erols.com [207.96.73.65]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA26908 for ; Wed, 3 Feb 1999 07:46:17 -0800 (PST) (envelope-from jdowdal@destiny.erols.com) Received: from destiny.erols.com (someone@destiny.erols.com [207.96.73.65]) by destiny.erols.com (8.9.2/8.6.12) with ESMTP id KAA13620; Wed, 3 Feb 1999 10:45:45 -0500 (EST) Date: Wed, 3 Feb 1999 10:45:44 -0500 (EST) From: John Dowdal To: Mike Zanker cc: freebsd-stable@FreeBSD.ORG Subject: RE: Kernel panic with recent RELENG_3 Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I have a PII-400 on an ABIT BX6 running 3.0 stable with the kernel cvsupped on 02/01 and the rest of the system on 01/29. The IBM Deskstar 16GP has its UDMA turned on with an a0ff flag. On 02/01 I compiled a new kernel which included a commit to if_wb.c which corrected an unitialized variable; the machine then crashed within 15 minutes of boot about 6 times using either the new kernel or the old kernel (indicating the commit to if_wb.c isn't the problem). The system paniced with a page fault in kernel mode error. First I tried squirting silicone lube on the cpu fan's bearing (does anyone know of a pentium CPU fan which does not get stiction within 6 months?). The fan then ran properly, but the machine still crashed. I made the following adjustments in a hail-mary to make it work at 12:30am: 1) Tear down and rebuild the computer. Blast every socket/slot/disk connector/card/dimm/cpu with canned air. 2) Disable bridging 3) ifconfig ed0 down (ed0 is a 10mbit SMC ELITE16 which has worked without problems since 1991; it was connected to an otherwise empty hub intended to connect windoze pc's for game playing. On the last working boot, there were numerous "unmuting ed0" "errors" logged to the console. Perhaps the random rebooting came from here. 4) Physically disconnect ed0 from its hub. Please remember that no change was made to the hardware, other than power cycling after a shutdown -h now (I have a very old scsi-1 disk which will not re-probe after BSD negotiates a transfer rmode other than the default). The machine has been up for about 30 hours now without crashing. I've recently seen several new complaints about UDMA, and am thinking something might be unitialized in the driver After all these complaints of similar panics/random lockups with UDMA, I'm suspect I might be *randomly* experiencing this problem, meaning that on some boots a config register has the correct random value, on other boots its wrong. If you get lucky, its stable :) I am completely aware that 3.0-STABLE will not be as stable as 2.2, and would like to help smoke out the bugs. On Wed, 3 Feb 1999, Mike Zanker wrote: > At 05:37 03/02/99 , John Saunders wrote: > > >Has anybody else noticed the IDE disk flackyness lately? > > Yes! I mentioned this over a week ago and no-one responded. I had noticed > that after prolonged heavy disk access my machine would completely lock up > for 4 or 5 seconds then reboot. I have now upgraded to a BX motherboard > with a Celeron 333, added 0xa0ff flags and so far things have been fine. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message