From owner-freebsd-questions Mon Apr 24 22:37: 1 2000 Delivered-To: freebsd-questions@freebsd.org Received: from yoda.dccnet.com (mail.deltacable.com [207.230.239.70]) by hub.freebsd.org (Postfix) with ESMTP id C2C5937BC93; Mon, 24 Apr 2000 22:36:54 -0700 (PDT) (envelope-from kg@versys.net) Received: from versys.net (unverified [209.5.131.193]) by yoda.dccnet.com (Rockliffe SMTPRA 3.4.6) with ESMTP id ; Mon, 24 Apr 2000 22:30:15 -0700 Message-ID: <39052DCF.DC16C2E7@versys.net> Date: Mon, 24 Apr 2000 22:31:59 -0700 From: "Kevin G. Eliuk" X-Mailer: Mozilla 4.72 [en] (X11; I; Linux 2.2.12 i386) X-Accept-Language: en MIME-Version: 1.0 To: Ryan Davis Cc: freebsd-questions@freebsd.org, freebsd-stable@freebsd.org Subject: Re: 3.4S -> 4.0S botched, can't build anything now References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Ryan Davis wrote: > > buildworld worked fine. > > installworld failed. > > I followed the instructions in /usr/src/UPDATING as best I could, but > I still think it is conflicting on whether to build the kernel first > or not. Doesn't matter, trying to build the kernel first failed, > presumably because config has changed. > > I have gone through several phases: > > 1) The first and most innocent ended with: > > ===> lib/libcom_err/doc > install-info --quiet --defsection="Programming & development tools." > --defentry="* libcom_err: (com_err). A Common Error > Description Library for UNIX." com_err.info /usr/share/info/dir > install-info: unrecognized option `--defsection=Programming & > development tools.' > Try `install-info --help' for a complete list of options. > > I fixed this with something from the mail archives. > > 2) Next was the immediate coredump on any make: > > "/usr/src/Makefile", line 101: warning: "LC_TIME=C date" returned > non-zero status > *** Signal 12 > > I figured out that /bin/sh would core. I copied /bin/bash over it and > it worked fine. > > 3) Finally, after trying to clean up and rebuild: > > /usr/obj/usr/src/i386/usr/bin/byacc -> /usr/obj/usr/src/i386/usr/bin/yacc > cd /usr/src/usr.bin/colldef; make obj; make depend; make all; make install > /usr/obj/usr/src/i386/usr/src/usr.bin/colldef created for > /usr/src/usr.bin/colldef > yacc -d /usr/src/usr.bin/colldef/parse.y > *** Signal 12 > > I am stuck. The only thing I've found that is relevant to #3 is to cd > to yacc's directory and make install, but I'm apparently not far > enough in the general build anymore for that to work. The wierd thing > is, unlike #2 where any use of /bin/sh would core, yacc works fine on > the commandline. > > I am trying to restore from tape now to see if I can roll out the > partial install until I get this under control. I am really > disappointed in how bad this transition has been. I usually brag to > linux bigots about how painless keeping up to date BSD is w/ cvsup > and /usr/src. Just dId the same today (cautiously) quite cleanly. (Starting from 3.4R) # cd /usr/src # make buildworld # cd /usr/src/sbin/mknod # make install # cd /usr/src/sys/modules # make install # cd /usr/src (Having edited my kernel config file "QUARTZ" while waiting for buildworld to finish) # make buildkernel KERNEL=QUARTZ # make installkernel KERNEL=QUARTZ # cp /usr/src/etc/MAKEDEV /dev # cd /dev # ./MAKEDEV ad0 (in my case) # ./MAKEDEV ad0s2a (again in my case) # vi /etc/fstab (change wd* to ad* entries) # shutdown -r now Rebooting and interupting the reboot ok set boot_single ok unload ok load /QUARTZ ok boot (I must admit I held my breath) After booting and manually mounting the slices ... # cd /usr/src # make -DNOINFO installworld # make installworld # mergemaster -cv -t /usr/tmp/root-20000424 # shutdown -r now After fixing a couple of other things like snd devices and ethernet devices it was a lot less tramatic than I was made to believe. -- Regards, Kevin G. Eliuk Technician | VerSys Communications Ltd. | http://www.versys.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message