From owner-cvs-all Wed May 24 11:38:22 2000 Delivered-To: cvs-all@freebsd.org Received: from relay.nuxi.com (nuxi.cs.ucdavis.edu [169.237.7.38]) by hub.freebsd.org (Postfix) with ESMTP id 8DA3437BD2A; Wed, 24 May 2000 11:38:10 -0700 (PDT) (envelope-from obrien@NUXI.com) Received: from dragon.nuxi.com (obrien@relay.nuxi.com [169.237.7.38]) by relay.nuxi.com (8.9.3/8.9.3) with ESMTP id LAA98715; Wed, 24 May 2000 11:38:06 -0700 (PDT) (envelope-from obrien@dragon.nuxi.com) Received: (from obrien@localhost) by dragon.nuxi.com (8.9.3/8.9.1) id LAA11900; Wed, 24 May 2000 11:38:25 -0700 (PDT) (envelope-from obrien) Date: Wed, 24 May 2000 11:38:25 -0700 From: "David O'Brien" To: Peter Wemm Cc: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/gnu/usr.bin/cc/cc1 Makefile src/gnu/usr.bin/cc/cc1obj Makefile src/gnu/usr.bin/cc/cc_int Makefile Message-ID: <20000524113825.D11538@dragon.nuxi.com> Reply-To: obrien@FreeBSD.org References: <20000524125226.7CF101CE1@overcee.netplex.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2i In-Reply-To: <20000524125226.7CF101CE1@overcee.netplex.com.au>; from peter@netplex.com.au on Wed, May 24, 2000 at 05:52:26AM -0700 X-Operating-System: FreeBSD 5.0-CURRENT Organization: The NUXI BSD group X-PGP-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Keyid: 34F9F9D5 Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Wed, May 24, 2000 at 05:52:26AM -0700, Peter Wemm wrote: > > > Did you return to previous version ? > > Yes. I got told off for touching this so it got backed out. > > David O'Brien is the maintainer. But I think it was bad to give our users whiplash once you did change the files. > I've done a bit of research and I've found out exactly what is going on and > why the new binutils has changed the behavior of the gcc build. The old > binutils ld had a bug in it's symbol handling and the new ld behavior is > correct and exactly as expected. So we are almost to the point we need to be at. It would be nice to know with `ld' 2.9.1 which symbols were used and does that explain any of the other toolchain bugs every one brings up. If you had made a full explanation of the problem I would not have gone off on you. But things with the tool chain cannot move forward with out understanding the issues. Since we are using Binutils version 2.10 which has *tons* of changes from 2.9.1, yet most Binutils developers attention is on 2.11 and I have seen several bugs fixed in 2.11 that have not be MFC to the 2.10 (release should be any day now) I am a little leary of 2.10. IMHO, the FreeBSD release pardygme is supier and GNU Binutils and GCC .0 releases are known to have bugs and is usally replaced by a .1 release farily quickley. We need to know if issues are in the Binutils code, our configuration, or other FreeBSD code. When I hit the IA-64 cross-compiling, having a lot of unknowns could be fatal. I now have an email from a main Binutils maintainer that explains the change happened in late December in order to be compatible with the Solaris linker. (note this email came w/in 30 minutes of me asking a question of him) -- -- David (obrien@FreeBSD.org) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message