From owner-freebsd-stable Fri Sep 17 23:33:13 1999 Delivered-To: freebsd-stable@freebsd.org Received: from mail.rdc1.az.home.com (ha1.rdc1.az.home.com [24.1.240.66]) by hub.freebsd.org (Postfix) with ESMTP id 4A901152F0 for ; Fri, 17 Sep 1999 23:33:10 -0700 (PDT) (envelope-from elgreen@iname.com) Received: from ehome.local.net ([24.9.114.169]) by mail.rdc1.az.home.com (InterMail v4.01.01.00 201-229-111) with SMTP id <19990918063310.DTXH27294.mail.rdc1.az.home.com@ehome.local.net> for ; Fri, 17 Sep 1999 23:33:10 -0700 From: Eric Lee Green Organization: Myself @ Home To: freebsd-stable@freebsd.org Subject: Mysterious spontaneous reboot problem between 3_2_RELEASE and 3_3_RELEASE Date: Fri, 17 Sep 1999 23:12:41 -0700 X-Mailer: KMail [version 1.0.21] Content-Type: text/plain MIME-Version: 1.0 Message-Id: <99091723331000.21348@ehome.local.net> Content-Transfer-Encoding: 8bit Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG At some point in time between the two, my FreeBSD 3.x developed spontaneous reboot problems. I've verified this on two different machines. Symptom: I start a tape backup. Perhaps twenty minutes later, the machine spontaneously reboots -- screen freezes over, then goes black, then comes up to the BIOS boot screen. There are two different machines involved: Machine A: AMD K6-3/333, some cheapo motherboard (never bothered looking at it), 128mb of RAM, S3 video card, Advansys ISA SCSI card, Seagate NS-20 tape drive hooked to the SCSI, 6.4gb IDE hard drive. Via-Rhine (vr0) network card. ES1371 sound card. Machine B: Celeron 300, ASUS motherboard, 128mb of RAM, ATI video card, Diamond Fireport 40 SCSI card, Seagate TR-4 tape drive hooked to the SCSI, 8.0gb IDE hard drive, 8.0gb in assorted older SCSI hard drives (scraped the bottom of the barrel there). Digital "Tulip" network card (de0), RealTek-based Hawking network card (rl0) (I also use it to route between two networks). Soundblaster Awe-64 ISA sound card. Voodoo2 3d accelerator. I know that four weeks ago I could make a tape backup successfully on both machines. Something has changed between then and now, and I'm stumped as to what. There's no panic message, no error messages appear on the screen, I even did it from a console instead of my typical "X" windows and no kernel messages appeared... it just bombed. Went black, went to BIOS bootup screen, and voila. This happened first to me with BRU (I work for the BRU guys, duh), but thinking it was something to do with BRU (BRU *does* kind of exercise every bug an operating system can have in its filesystem code, there's a very obscure bug BRU found in the way directories are read off of Linux partitions under FreeBSD that I'm trying to track down at the moment), I tried 'tar'. The machine still spontaneously rebooted. I happened to be looking at the screen at the time, and the machine was backing up /usr/local/src/caldera/col23.iso (the ISO image for Caldera Linux 2.3, I was getting ready to install it on top of my defunct Linux partition and wanted a backup first). The only really interesting thing about my setup is that it's a dual boot with Linux on both of these machines, and I *am* backing up the Linux partitions too. But I haven't booted Linux for quite some time (probably been over a month, I only do it when new distributions come out so I can test against them), and I always mount the Linux filesystems "ro" (read-only), so that shouldn't be a problem... Linux filesystems can't degrade if they're not being used (grin). The other interesting thing is that these machines *NEVER* spontaneously reboot on me in everyday use. Never. Both machines have been up and going for quite some time, and until this problem, there wasn't one. And we're talking about machines that get USED. They're compiling all day, and if they're not compiling, they're doing something else strenuous like doing analysis of millions of pseudo-random numbers to verify the strength of a random number generator. So what should I do next? Should the latest 3_3 fix this? I'll try (says he, dubiously), 'make world' is going as we speak, but this is rather disturbing (reminds me of why I dumped Linux -- bloody Linux 2.2 kernel had more bugs than a roach motel, 2.2.10 leaked memory like a demon, and 2.2.12 still has one of the world's buggiest NFS implementations). -- Eric Lee Green http://members.tripod.com/e_l_green mail: e_l_green@hotmail.com There Is No Conspiracy To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message