From owner-freebsd-stable Mon Jun 17 23:51:12 1996 Return-Path: owner-stable Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA04883 for stable-outgoing; Mon, 17 Jun 1996 23:51:12 -0700 (PDT) Received: from ref.tfs.com (ref.tfs.com [140.145.254.251]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id XAA04873 for ; Mon, 17 Jun 1996 23:51:09 -0700 (PDT) Received: from Root.COM (implode.Root.COM [198.145.90.17]) by ref.tfs.com (8.7.5/8.7.3) with ESMTP id XAA06878 for ; Mon, 17 Jun 1996 23:50:57 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by Root.COM (8.7.5/8.6.5) with SMTP id XAA01247; Mon, 17 Jun 1996 23:45:57 -0700 (PDT) Message-Id: <199606180645.XAA01247@Root.COM> X-Authentication-Warning: implode.Root.COM: Host localhost [127.0.0.1] didn't use HELO protocol To: Juergen Lock cc: stable@freebsd.org Subject: Re: Trap 12/supervisor read, page not present In-reply-to: Your message of "Tue, 18 Jun 1996 08:16:00 +0200." <199606180616.IAA10166@saturn.hb.north.de> From: David Greenman Reply-To: davidg@Root.COM Date: Mon, 17 Jun 1996 23:45:57 -0700 Sender: owner-stable@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> Like I said, this is a "known" problem. What I don't understand is why you >>see it happen so often. It takes literally *months* of heavy system activity >>for me to reproduce this sometimes, which is why it has taken so long to fix >>it. We'll be looking at this problem again this week, but I can give no > >Hmm. im not deep enough inside the vm code to take a guess whether >it may be related or not but, i just made a new -stable kernel again >(now ctm-cvs-cur 2130) and the hold count problem (kern/1195) is >still there. system just booted, otherwise completely idle except for >a few xterms and a xperfmon: after having written stables /usr/obj this >time i have 21 of those (DIAGNOSTIC-panic-patched-into-printf, see PR) >messages on the xconsole... > >find obj | afio -G 2 -o -v -f -b 5120 -c 1000 -s 100000x -Z /dev/nrst0 >(this was the test. CPU 486dx100, tape Wangtek 525ES, PCI NCR SCSI.) This appears to be caused by an unrelated bug in the physio code. > ATB and sorry if this is old news and of no use I appreciate all the information you've provided on the problem. I'll be looking forward to having you test a fix if/when we come up with one for the "invalid kernel page directory" problem (and perhaps also the bug you've reported above). >PS: is the Root.COM a real address? i cant nslookup right now >(still no flat rates in germany), therefore cc'd to the list... Of course. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project