From owner-freebsd-questions Thu Oct 28 14:14:16 1999 Delivered-To: freebsd-questions@freebsd.org Received: from mercury.gfit.net (ns.gfit.net [209.41.124.90]) by hub.freebsd.org (Postfix) with ESMTP id 7ECF814C18 for ; Thu, 28 Oct 1999 14:14:06 -0700 (PDT) (envelope-from tom@embt.com) Received: from PARANOR (timembt.iinc.com [206.67.169.229]) by mercury.gfit.net (8.8.8/8.8.8) with SMTP id QAA03642; Thu, 28 Oct 1999 16:17:46 -0500 (CDT) (envelope-from tom@embt.com) Message-Id: <3.0.3.32.19991028171255.0108f960@mail.embt.com> X-Sender: tembt@mail.embt.com X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.3 (32) Date: Thu, 28 Oct 1999 17:12:55 -0700 To: David Scheidt From: Tom Embt Subject: Re: upgrading/CVSUPping with bad ISP connection Cc: freebsd-questions@FreeBSD.ORG In-Reply-To: References: <3.0.3.32.19991028151057.01092eb8@mail.embt.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >I very strongly suggest building a new kernel, installing it, and rebooting >*before* building world. It is much easier to recover from a broken kernel >than a broken world. New kernels work with old worlds, but the reverse >isn't always true. At some point, this order will be required, since there >are a bunch of changes in -CURRENT that require it, so now is a good time to >start gettng used to it. When you try and config the kernel, you might get >an error saying that you haven't got a new enough config. If you do see >this, do the following: ># cp /usr/sbin/config /usr/sbin/config.old ># cd /usr/src/user.sbin/config ># make cleandir ># make && make install > >and try again. > >Doing things this way isn't any harder, and is much safer. It might take a >bit longer since you should do a reboot after make installworld, but I think >a machine that doesn't break is worth it. > >David Scheidt > I should've known this was coming ;) I had to do this recently on my -current box following the massive signal changes, however I've not yet adopted this as standard policy. I guess I've got the old "kernel and userland should be in sync" idea embedded too far into my head. Well, that and i don't recall ever having a bad kernel that wasn't My Own Stupid Fault. Is the "kernel before world" philosophy going to become SOP? Tom Embt tom@embt.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message