From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/3905 Path: news.gmane.org!not-for-mail From: Newsgroups: gmane.linux.lib.musl.general Subject: RE: problems with dynamic linking since 0.9.1 Date: Wed, 14 Aug 2013 15:26:24 -0700 Message-ID: <20130814152624.dc30d64f61e5ec441c34ffd4f788e58e.2015f1a843.wbe@email22.secureserver.net> Reply-To: musl@lists.openwall.com NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable X-Trace: ger.gmane.org 1376519198 32616 80.91.229.3 (14 Aug 2013 22:26:38 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Wed, 14 Aug 2013 22:26:38 +0000 (UTC) To: musl@lists.openwall.com Original-X-From: musl-return-3909-gllmg-musl=m.gmane.org@lists.openwall.com Thu Aug 15 00:26:41 2013 Return-path: Envelope-to: gllmg-musl@plane.gmane.org Original-Received: from mother.openwall.net ([195.42.179.200]) by plane.gmane.org with smtp (Exim 4.69) (envelope-from ) id 1V9jWb-0007b0-HW for gllmg-musl@plane.gmane.org; Thu, 15 Aug 2013 00:26:41 +0200 Original-Received: (qmail 9780 invoked by uid 550); 14 Aug 2013 22:26:41 -0000 Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: Original-Received: (qmail 9772 invoked from network); 14 Aug 2013 22:26:40 -0000 X-SID: CmST1m0012YkKj001 X-Originating-IP: 46.121.219.189 User-Agent: Workspace Webmail 5.6.41 Xref: news.gmane.org gmane.linux.lib.musl.general:3905 Archived-At:

=0A

=0A
=0A----= ---- Original Message --------
=0ASubject: Re: [musl] problems with dyna= mic linking since 0.9.1
=0AFrom: Rich Felker <dalias@aerifal.cx>
=0ADate: Wed, August 14, 2013 1:= 58 pm
=0ATo: musl@lists.openw= all.com
=0A
=0AOn Wed, Aug 14, 2013 at 10:47:08PM +0200, Jens wro= te:
=0A> >My recollection was that it failed to support -Bsymbolic= -functions at
=0A> >all and would produce an error when encounteri= ng it, so this makes me
=0A> >wonder how generation of libc.so succeeded at all...
=0A>
=0A> The = musl libc in this case is built with binutils-2.20.1, since the
=0A> = older binutils (2.17) didnt work. You helped me with this exact
=0A> = problem some months ago.
=0A>
=0A> I have a build-environment = where I specify all the dependencies for
=0A> each build. binutils-2.= 20.1 is then a dependency for musl (where
=0A> binutils 2.17 is the d= efault).
=0A>
=0A> So for my use-case I can always specify a l= ater binutils as a
=0A> dependency for all musl builds. Though dynami= c linking is a low
=0A> priority for me, since all resulting binaries= must be statically
=0A> linked.
=0A
=0AOK. It may be that binu= tils-2.20.1, at least with the options we're
=0Ausing, is somehow taggin= g the libc.so file as incompatible with olde= r
=0Abinutils. This would be unfortunate, and it's an issue I'd like to<= br>=0Aresolve if that's what it is. I'll see if I can reproduce it or if=0Asomeone else (preferably someone on x86_64) can do so.
=0A
=0ARic= h
=0A=0A
=0A

I ran into the same is= sue two or three months ago while building gcc against musl on Ubuntu 12.04= (x86_64) with binutils 2.22. For all I can tell (from memory), the p= roblem had nothing to do with musl, but rather with ld being built against = one libc (musl), and gcc being built against another libc (gnu). I so= lved it by setting PATH, LIBRARY_PATH and LD_LIBRARY_PATH appropriately.

zg