[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: --preserve-dup-deps seems not to work

From: Ralf Wildenhues
Subject: Re: --preserve-dup-deps seems not to work
Date: Wed, 13 Sep 2006 08:54:22 +0200
User-agent: Mutt/1.5.13 (2006-09-01)

Hello Stefan,

Thanks for the bug report.

* Stefan Traby wrote on Wed, Sep 13, 2006 at 04:01:19AM CEST:
> I tried libtool-1.4.3 (which introduced --preserve-dup-deps)
> in addition to 1.5.22 which I use normally:
> The long (real-world) output is here:
> here a short variant:
> /bin/sh ./libtool --preserve-dup-deps --mode=link gcc  -g -O2 -o prog prog.o 
> gcc -g -O2 -o prog prog.o  ./.libs/liba.a ./.libs/libb.a
> I don't know how to make libtool to honor duplicate dependencies.

In this special case, you have two convenience archives and which have a circular dependency (as opposed to: one or both
libraries are to-be-installed static or shared libraries).  I agree
that this case should work.  But this case is also the easiest to work
around: you could add as a dependency to the link line of
  $LIBTOOL --mode=link --tag=CC $CC $CFLAGS $LDFLAGS -o \
           b.lo [OBJECTS...]

This is a bit wasteful in that all objects from liba will also appear
in some form or other in libb, but it should be portable.  It will
require you to add a dependency of on in your Makefile
(in case you use Automake it should take care of that, I believe).

I think we should fix this.  Proposed test case below.


        * tests/ New file.  Test circular depending
        convenience archives (currently failing).
        * Update.
        Report by Stefan Traby <address@hidden>.

RCS file: /cvsroot/libtool/libtool/,v
retrieving revision 1.198
diff -u -r1.198
--- 12 Sep 2006 18:02:31 -0000      1.198
+++ 13 Sep 2006 05:48:01 -0000
@@ -399,6 +399,7 @@
                  tests/ \
                  tests/ \
                  tests/ \
+                 tests/ \
                  tests/ \
                  tests/ \
                  tests/ \
--- /dev/null   2006-09-05 22:40:33.520458500 +0200
+++ tests/     2006-09-13 07:48:16.000000000 +0200
@@ -0,0 +1,64 @@
+# Hand crafted tests for GNU Libtool.                         -*- Autotest -*-
+# Copyright 2006 Free Software Foundation, Inc.
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2, or (at your option)
+# any later version.
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# GNU General Public License for more details.
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+# 02110-1301, USA.
+AT_SETUP([preserve duplicate convenience deps])
+# --preserve-dup-deps should work for convenience archives.
+# Create a circular dependency of liba and libb:
+# a1 pulls in b1, that pulls in a2.
+cat >a1.c <<\EOF
+extern int b1 ();
+int a1 () { return b1 (); }
+cat >a2.c <<\EOF
+int a2 () { return 0; }
+cat >b1.c <<\EOF
+extern int a2 ();
+int b1 () { return a2 (); }
+cat >main.c <<\EOF
+extern int a1 ();
+int main () { return a1 (); }
+for file in a1.c a2.c b1.c; do
+  $LIBTOOL --mode=compile $CC $CPPFLAGS $CFLAGS -c $file
+$CC $CPPFLAGS $CFLAGS -c main.c
+$LIBTOOL --mode=link --tag=CC $CC $CFLAGS $LDFLAGS -o a1.lo a2.lo
+# This could be worked around by adding to
+# (in that case all objects from liba would be merged into
+# libb.a as well, possibly renamed.)
+$LIBTOOL --mode=link --tag=CC $CC $CFLAGS $LDFLAGS -o b1.lo
+AT_CHECK([$LIBTOOL --mode=link --tag=CC \
+         $CC $CFLAGS $LDFLAGS -o main main.$OBJEXT],
+         [0], [ignore], [ignore])
+# This currently fails:
+$LIBTOOL --mode=link --tag=CC $CC $CFLAGS $LDFLAGS -o b1.lo
+AT_CHECK([$LIBTOOL --mode=link --preserve-dup-deps --tag=CC \
+         $CC $CFLAGS $LDFLAGS -o main main.$OBJEXT],
+         [0], [ignore], [ignore])

reply via email to

[Prev in Thread] Current Thread [Next in Thread]