From owner-freebsd-bugs  Tue Feb  4 06:56:17 1997
Return-Path: <owner-bugs>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id GAA19317
          for bugs-outgoing; Tue, 4 Feb 1997 06:56:17 -0800 (PST)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA19312
          for <bugs@freebsd.org>; Tue, 4 Feb 1997 06:56:14 -0800 (PST)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id BAA24540; Wed, 5 Feb 1997 01:51:14 +1100
Date: Wed, 5 Feb 1997 01:51:14 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199702041451.BAA24540@godzilla.zeta.org.au>
To: bugs@freebsd.org, jlemon@americantv.com
Subject: Re: gdb/libiberty failure
Sender: owner-bugs@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>-current falls over on building dependencies on gdb/libiberty.

It works fine in -current here.

>.bin/gdb/libiberty/Makefile.usr.bin.gdb.libiberty.Makefile /usr/tuna/src/gnu/usr 
>*** gnu.usr.bin.gdb.libiberty.Makefile  Mon Feb  3 15:37:51 1997
>--- /usr/tuna/src/gnu/usr.bin/gdb/libiberty/Makefile    Mon Feb  3 16:16:38 1997
>***************
>*** 5,12 ****
>  
>  SRCS= argv.c basename.c concat.c cplus-dem.c fdmatch.c getopt.c \
>        getopt1.c getruntime.c hex.c floatformat.c obstack.c spaces.c \
>!       strerror.c strsignal.c vasprintf.c xatexit.c xexit.c \
>!       xmalloc.c xstrdup.c xstrerror.c insque.c
>  
>  CFLAGS+= -I$(.CURDIR) -I$(.CURDIR)/../gdb/.
>  NOPROFILE=no
>--- 5,11 ----
>  
>  SRCS= argv.c basename.c concat.c cplus-dem.c fdmatch.c getopt.c \
>        getopt1.c getruntime.c hex.c floatformat.c obstack.c spaces.c \
>!       insque.c
>  
>  CFLAGS+= -I$(.CURDIR) -I$(.CURDIR)/../gdb/.
>  NOPROFILE=no

However, it shouldn't be using all that cruft, especially vasprintf(),
which was nuked in the old gdb when a better version was put in the
library.

There seems to be a linkage bug.  vasprintf.o from liberty.a is used,
but strerror.o from liberty.a is not used - the standard (shared) version
is used instead.  I would have expected the (broken) behaviour for
strerror.o in both cases.

Bruce