From owner-freebsd-current Tue Mar 7 07:21:57 1995 Return-Path: current-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id HAA16711 for current-outgoing; Tue, 7 Mar 1995 07:21:57 -0800 Received: from Root.COM (implode.Root.COM [198.145.90.1]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id HAA16699 for ; Tue, 7 Mar 1995 07:21:53 -0800 Received: from corbin.Root.COM (corbin.Root.COM [198.145.90.18]) by Root.COM (8.6.8/8.6.5) with ESMTP id HAA00174; Tue, 7 Mar 1995 07:23:33 -0800 Received: from localhost (localhost [127.0.0.1]) by corbin.Root.COM (8.6.9/8.6.5) with SMTP id HAA00179; Tue, 7 Mar 1995 07:21:17 -0800 Message-Id: <199503071521.HAA00179@corbin.Root.COM> X-Authentication-Warning: corbin.Root.COM: Host localhost didn't use HELO protocol To: Bruce Evans cc: current@FreeBSD.org, phk@ref.tfs.com Subject: Re: the chatterbug categorized In-reply-to: Your message of "Tue, 07 Mar 95 21:10:44 +1000." <199503071110.VAA13544@godzilla.zeta.org.au> From: David Greenman Reply-To: davidg@Root.COM Date: Tue, 07 Mar 1995 07:21:12 -0800 Sender: current-owner@FreeBSD.org Precedence: bulk >>I belive that David and I have understood the chatterbug, now we just >>need to find it and fix it. > >>What happens is that some vnodes are not properly freed when the vm_object >>is freed, this has two effects: one is that the list of free vnodes are >>too short to make the name-cache efficient (the chatter) and the other >>is that the system will continue to allocate more vnodes, (the gradual >>loss of available memory) > >>Now to isolate it, we need to know when it happens, I can reproduce it here >>with: >> cd /usr/src/release >> make release CHROOTDIR=<500Mb space> > >There seems to be a problem with the increment of object->ref_count in >vm_mmap.c. That would be an understatement. John and I managed to introduce multiple object reference count bugs when we rewrote vm_mmap(). Object references are being gained by subtle side effects that were overlooked; if I hadn't put diagnostic code in there to prove this, I wouldn't believe it. I'm not altogether certain that there aren't other external (object reference count related) bugs...it's going to take me a bit to follow all the code paths. I hope to have a fix by late tonight. Thanks to Poul and Bruce for helping track this one down. Geeze, and I tried so hard to get the damn reference counts correct in there! -DG