From owner-freebsd-smp Mon Apr 3 2:51:51 2000 Delivered-To: freebsd-smp@freebsd.org Received: from relativity.student.utwente.nl (wit389306.student.utwente.nl [130.89.234.166]) by hub.freebsd.org (Postfix) with ESMTP id 57E3237BDC9 for ; Mon, 3 Apr 2000 02:51:42 -0700 (PDT) (envelope-from djb@wit389306.student.utwente.nl) Received: by relativity.student.utwente.nl (Postfix, from userid 1000) id B20F35DBF; Mon, 3 Apr 2000 11:51:24 +0200 (CEST) Date: Mon, 3 Apr 2000 11:51:24 +0200 From: Dave Boers To: Matthew Dillon Cc: Tommy Hallgren , Jeremiah Gowdy , bart@ixori.demon.nl, freebsd-smp@FreeBSD.ORG Subject: Re: SMP and vn Message-ID: <20000403115124.A374@relativity.student.utwente.nl> Reply-To: djb@relativity.student.utwente.nl References: <20000327183911.18682.qmail@web124.yahoomail.com> <20000329132919.A10781@relativity.student.utwente.nl> <200003291604.IAA63016@apollo.backplane.com> <20000329203525.A379@relativity.student.utwente.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0.1i In-Reply-To: <20000329203525.A379@relativity.student.utwente.nl>; from djb@relativity.student.utwente.nl on Wed, Mar 29, 2000 at 08:35:25PM +0200 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org It is rumoured that Dave Boers had the courage to say: > I reconfigured my kernel. It's now booting on the serial terminal as a > console. Also I removed DDB_UNATTENDED from my kernel config and added > DIAGNOSTIC to it. > > _IF_ there's an unnoticed kernel panic going along with the system hang > then I will be dropped in DDB on the serial terminal if all is well. This > way, I won't miss out on any kernel messages while using X. > Talk to you all in 5-7 days or so :-) Well, it's been a little less than 5-7 days. The system just crashed on me. This time, I am _absolutely sure_ that the system is _not_ doing a panic. I configured a serial terminal as console and I repeatedly tested DDB on it. It works fine. However during the latest crash, there was no debugger or panic message on the serial console. The lockup was again complete; the keyboard NUMLOCK key didn't even respond anymore. The lockup occurred while XFree86 was running, along with some editor sessions (vim), some ssh sessions, Licq and (important, I guess) I was burning a cd-recordable using cdrecord. So the system was using the posix realtime scheduling a lot. The crash left my scsi devices in such a state that I had to do a full system power down before I got my cdrom devices back online. Furthermore, this shows that the problem was not the specific kernel configuration or the ATA driver (I was running a kernel configuration sent to me by Lyndon Nerenberg who does not have the lockups on the same hardware; I removed the ATA driver completely from the kernelconfig). This leaves XFree86, cdrecord or licq to trigger the problem. I do have some harddisk space to spare; would it be a good idea to add some logging to some parts of the kernel to a logfile somewhere? Maybe that way we can trace what happened just before the lockups? I don't have any kernel programming experience, so I don't know were to look. Can someone here help out? I don't have a problem with a logfile of, say, 7 Gb or so. Regards, Dave. -- djb@ifa.au.dk To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message