From owner-freebsd-hackers Sat Nov 9 15:20:45 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id PAA27868 for hackers-outgoing; Sat, 9 Nov 1996 15:20:45 -0800 (PST) Received: from dg-rtp.dg.com (dg-rtp.rtp.dg.com [128.222.1.2]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id PAA27854 for ; Sat, 9 Nov 1996 15:20:41 -0800 (PST) Received: by dg-rtp.dg.com (5.4R3.10/dg-rtp-v02) id AA08985; Sat, 9 Nov 1996 18:20:02 -0500 Received: from ponds by dg-rtp.dg.com.rtp.dg.com; Sat, 9 Nov 1996 18:20 EST Received: from lakes.water.net (lakes [10.0.0.3]) by ponds.water.net (8.7.5/8.7.3) with ESMTP id OAA28790 for ; Sat, 9 Nov 1996 14:14:56 -0500 (EST) Received: (from rivers@localhost) by lakes.water.net (8.7.5/8.6.9) id OAA10664 for freebsd-hackers@freefall.cdrom.com; Sat, 9 Nov 1996 14:16:09 -0500 (EST) Date: Sat, 9 Nov 1996 14:16:09 -0500 (EST) From: Thomas David Rivers Message-Id: <199611091916.OAA10664@lakes.water.net> To: ponds!freefall.cdrom.com!freebsd-hackers Subject: Even more info on daily panics... Content-Type: text Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Ok - Just so everyone doesn't think I've vanished from the face of the earth :-) I thought I'd provide a little status. I'm now running a 2.1.5-S kernel I grabbed on Oct. 17th. This kernel has two changes to is: 1) The "patch" Terry suggested to vfs_subr.c [which, in itself didn't stop my problem from reoccuring.] 2) The change in vrele() to catch when the usecount goes negative. I installed that kernel on Nov. 7th... It is now Sat Nov 9 14:11:33 EST 1996, and uptime shows: 2:11PM up 2 days, 1:14, 1 user, load averages: 0.81, 0.90, 0.83 That is, the machine has not panic'd since I installed this kernel. There is one possibility - perhaps, when I was trying Terry's suggested patch, I accidently installed the wrong kernel... maybe Terry's patch does work-around the problem. There is another possibility - in the past, the machine would sometimes go for two or three days without a panic. It would rarely go longer than that. I could be in one of these "lulls." There is still another possibility, gcc 2.6.3 has a bug which is tickled by the vrele() code. Changing the source caused the bad code to not be generated [I'll be the first one to admit that I'm reaching here...] Anyway, as soon as something happens, either the typical panic's I've been getting, or a new panic from usecount<0 in vrele() - I'll let everyone know... - Dave Rivers -