Re: Linking Problem with Solaris ld

[Bob Friesenhahn]

On Sun, 17 Apr 2005, Moritz Schulte wrote:

> On Sat, Apr 16, 2005 at 01:52:39PM -0500, Bob Friesenhahn wrote:
>
> > I believe that this problem is due to a missing LDADD in the
> > libgcrypt Automake makefiles and I formally reported this as a bug
> > to the libgcrypt developers some six or eight months ago.  I think I
> > even supplied a patch.
>
> I am confused now.  The Makefile.am in question looks like this:
>
> ----------------------------------------
> [...]
> lib_LTLIBRARIES = libgcrypt.la
>
> [...]
>
> libgcrypt_la_CFLAGS = @GPG_ERROR_CFLAGS@
> libgcrypt_la_SOURCES = g10lib.h types.h cipher.h \
>        misc.c global.c sexp.c \
>        stdmem.c stdmem.h secmem.c secmem.h \
>        mpi.h missing-string.c module.c \
>        ath.h ath.c
>
> libgcrypt_la_LDFLAGS = $(libgcrypt_version_script_cmd) -version-info \
>        @LIBGCRYPT_LT_CURRENT@:@LIBGCRYPT_LT_REVISION@:@LIBGCRYPT_LT_AGE@
> libgcrypt_la_DEPENDENCIES = ../cipher/libcipher.la ../mpi/libmpi.la \
>        $(srcdir)/libgcrypt.vers
> libgcrypt_la_LIBADD = ../cipher/libcipher.la ../mpi/libmpi.la \
>        @LTLIBOBJS@ @GPG_ERROR_LIBS@
> ----------------------------------------
>
> There is no LDADD, but LIBADD (which I though to be appropriate for
> this purpose).  What exactly should be changed here?

Looking back at what I did to get the libgcrypt package building the 
last time around, I see this difference:

Index: tests/Makefile.am
===================================================================
RCS file: /cvs/gnupg/libgcrypt/tests/Makefile.am,v
retrieving revision 1.16
diff -u -r1.16 Makefile.am
--- tests/Makefile.am   3 Mar 2004 08:08:05 -0000       1.16
+++ tests/Makefile.am   17 Apr 2005 14:52:09 -0000
@@ -21,7 +21,7 @@
 TESTS = prime register ac basic tsexp keygen pubkey benchmark

 INCLUDES = -I$(top_srcdir)/src
-LDADD = ../src/libgcrypt.la
+LDADD = ../src/libgcrypt.la ../cipher/libcipher.la

 EXTRA_PROGRAMS = testapi
 noinst_PROGRAMS = $(TESTS)

This fix may have been for a symptom rather than for the underlying 
problem.

Bob
======================================
Bob Friesenhahn
address@hidden, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/