From owner-freebsd-current Mon Oct 19 02:37:56 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA12782 for freebsd-current-outgoing; Mon, 19 Oct 1998 02:37:56 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from cimlogic.com.au (cimlog.lnk.telstra.net [139.130.51.31]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id CAA12774 for ; Mon, 19 Oct 1998 02:37:52 -0700 (PDT) (envelope-from jb@cimlogic.com.au) Received: (from jb@localhost) by cimlogic.com.au (8.9.1/8.9.1) id TAA19300; Mon, 19 Oct 1998 19:40:24 +1000 (EST) (envelope-from jb) From: John Birrell Message-Id: <199810190940.TAA19300@cimlogic.com.au> Subject: Re: 3.0 Current to 3.0 Release Woes In-Reply-To: <000001bdfb3d$76113d20$0400000a@roaming> from "Jon E. Mitchiner" at "Oct 19, 98 04:50:02 am" To: jon@minotaur.com (Jon E. Mitchiner) Date: Mon, 19 Oct 1998 19:40:24 +1000 (EST) Cc: freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL40 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Jon E. Mitchiner wrote: > I did "make > aout-to-elf" and things appeared to work well, then throughout the process, > it could not find a directory -- I didn't write it down, but it was > something to the effect of freebsd2/aout which I had to create (missing > directory for "aout"), then re-ran make aout-to-elf which continued fine. Was this during the move of the aout libraries? If so, then the obj trees would have contained complete and up-to-date aout and elf objects, libraries and executables. A library path with "freebsd2" in it sounds like modula-3. I wonder if this is somehow related to a symlink in the library search path. > Then at one point the GENERICupgrade kernel had an error, so I had removed > the DPT_MEASURE_PERFORMANCE so the kernel would compile correctly. Then > after this, the system rebooted. Things seemed to look good. Did you check if GENERICupgrade was _*suitable*_ for your system? > Unfortunately, all the problems occurred here. Upon startup, it could not > run /bin/sh or any other shell I had in /bin -- it would do a signal 10. So > I was stumped and I was trying to find a way in to correct the problem. > Apparently it seems that some libs and the elf programs were not compiled > properly and everything was breaking. I couldn't get the system to work > correctly. Is it possible that you built this with the libc/exec problem that DES warned about on Oct 15: | If you cvsupped -current around 7 or 8 pm GMT, watch out for possible | breakage in src/lib/libc/gen/exec.c. It'll compile but may break some | programs, though it's not very serious (you'll still be able to | recompile a functioning libc even with a broken libc installed) Check | the revision number in RCS Id of that file; it should be 1.9: | | # ident /usr/src/lib/libc/gen/exec.c | /usr/src/lib/libc/gen/exec.c: | $Id: exec.c,v 1.9 1998/10/14 20:23:40 des Exp $ I've deleted the rest of your message because we don't have a hope of working out why things ended up the way they did due to consequential errors. It's unwise to hack a stage of the upgrade to get it to complete. If there is a problem that is preventing the upgrade from completing, that needs to be fixed first. To fix it, we need it reported to -current before you hack and change "evidence". This is the first report I've seen of anyone getting thoroughly hosed by the upgrade process. -- John Birrell - jb@cimlogic.com.au; jb@freebsd.org http://www.cimlogic.com.au/ CIMlogic Pty Ltd, GPO Box 117A, Melbourne Vic 3001, Australia +61 418 353 137 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message