From owner-freebsd-hackers Mon Apr 15 22:15:34 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id WAA10600 for hackers-outgoing; Mon, 15 Apr 1996 22:15:34 -0700 (PDT) Received: from Root.COM (implode.Root.COM [198.145.90.17]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id WAA10590 for ; Mon, 15 Apr 1996 22:15:29 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by Root.COM (8.7.5/8.6.5) with SMTP id WAA03240; Mon, 15 Apr 1996 22:15:18 -0700 (PDT) Message-Id: <199604160515.WAA03240@Root.COM> X-Authentication-Warning: implode.Root.COM: Host localhost [127.0.0.1] didn't use HELO protocol To: Brian Tao cc: Jaye Mathisen , hackers@FreeBSD.org Subject: Re: Hmmm, OK, SMC driver is the de driver. Is it broke? In-reply-to: Your message of "Tue, 16 Apr 1996 00:47:44 EDT." From: David Greenman Reply-To: davidg@Root.COM Date: Mon, 15 Apr 1996 22:15:18 -0700 Sender: owner-hackers@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk >On Sun, 14 Apr 1996, David Greenman wrote: >> >> >Well, I punted and went to -current, and it seems to be just fine. I >> >think something is broke in -stable. Dropping back to a kernel from >> >around Mar 15th works fine as well, but 3/23 and 4/11 both fail miserably. >> >> Do you use NFS? Do you have it and all other filesystems that you use >> specified in your kernel config file? The only significant change I can think >> of that was made to -stable in that time period was a change to vnode.h to >> change the size of some fields. If you have *any* LKMs that haven't been >> rebuilt, this will cause the system to fail. > > I tried upgrading our 2.1.0R Web/FTP server machine to 2.2-960323-SNAP >(see my message "Subject: pmap_zero_page and kmem_malloc panics"), but >I'm getting an "unimplemented trap" kernel panic as soon as it tries >to ifconfig the de0 interface. It doesn't matter whether >/etc/netstart calls it, or if I type it in from a single-user shell. >That may not be the exact panic message, but I can double-check that >when I'm at work tomorrow. The panic occurred in this region of the >kernel: > >f01ab180 F trap.o >f01ab310 T _trap >f01ab794 t _trap_pfault >f01abb0c t _trap_fatal <--- inside this routine >f01abdd8 T _dblfault_handler >f01abe24 T _trapwrite >f01abf04 T _syscall > > At first I had only upgraded the kernel and the lkm's to the >snapshot, then I completely reinstalled the OS. No go either way. >The machine is now back to 2.1.0R (after much grumbling of users and >admin alike). > > I checked the if_de.c code in both 960323 and -current as of April >16, and both have the same revision number (1.44). Has something else >changed? Should I try -stable or even -current? > > The kernel config file used to build the 2.2 kernel is the exact >same one used to build the 2.1 kernel. NFS is compiled in. Any >ideas? I can see nothing wrong with your config file. I can only think that somehow your sources aren't in sync or some other build related problem. All of my machines here use the 'de' driver and -current works just fine with them. Perhaps it could be related to ARP? What is the ifconfig line that you're running at startup? Do you do IP aliasing? -DG David Greenman Core-team/Principal Architect, The FreeBSD Project