libtool followup (was: Re: [Swarm-Support] Install problem on fc3

[Paul Johnson]

Well, I couldn't leave the FC3 alone. If you update gcc and attendant 
libs and cpp from here to version 3.4.3-4

http://download.fedora.redhat.com/pub/fedora/linux/core/development/i386/Fedora/RPMS/

The build does go past that internal compiler error.

When trying to build an RPM, I'm caught on some libtool magic.  The 
libswarm.so file is not created properly, there's no "so" in the file 
name. In the temporary install directory, look at this weirdness:


./usr/lib
./usr/lib/debug/usr/lib
./usr/lib/debug/usr/lib/swarm/libswarm.0.0.0.debug
./usr/lib/debug/usr/lib/swarm/libjavaswarm.1.0.0.debug
./usr/lib/swarm/libswarm.a
./usr/lib/swarm/libswarm.0.0.0
./usr/lib/swarm/libjavaswarm.a
./usr/lib/swarm/libswarm.la
./usr/lib/swarm/libswarm
./usr/lib/swarm/libjavaswarm.1.0.0
./usr/lib/swarm/libjavaswarm.la
./usr/lib/swarm/libjavaswarm
./usr/lib/swarm/libswarm.0
./usr/lib/swarm/libjavaswarm.1


In the output from make install, I see this:

 /bin/sh ../libtool --mode=install /usr/bin/install -c  libswarm.la 
/tmp/swarm-root/usr/lib/swarm/libswarm.la
/usr/bin/install -c .libs/libswarm.0.0.0 
/tmp/swarm-root/usr/lib/swarm/libswarm.0.0.0
(cd /tmp/swarm-root/usr/lib/swarm && rm -f libswarm.0 && ln -s 
libswarm.0.0.0 libswarm.0)
(cd /tmp/swarm-root/usr/lib/swarm && rm -f libswarm && ln -s 
libswarm.0.0.0 libswarm)
/usr/bin/install -c .libs/libswarm.lai 
/tmp/swarm-root/usr/lib/swarm/libswarm.la
/usr/bin/install -c .libs/libswarm.a 
/tmp/swarm-root/usr/lib/swarm/libswarm.a
ranlib /tmp/swarm-root/usr/lib/swarm/libswarm.a
chmod 644 /tmp/swarm-root/usr/lib/swarm/libswarm.a
libtool: install: warning: remember to run `libtool --finish /usr/lib/swarm'

And in the build dir, I have

$ find . -name "lib*"
./avcall/libtool
./avcall/.libs/libavcall.a
./avcall/.libs/libavcall.la
./avcall/libavcall.la
./libtool
./libobjc
./libobjc/libtool
./libobjc/.libs/libswarmobjc.a
./libobjc/.libs/libswarmobjc.la
./libobjc/libswarmobjc.la
./src/activity/.libs/libactivity.a
./src/activity/.libs/libactivity.la
./src/activity/libactivity.la
./src/analysis/.libs/libanalysis.a
./src/analysis/.libs/libanalysis.la
./src/analysis/libanalysis.la
./src/collections/.libs/libcollections.a
./src/collections/.libs/libcollections.la
./src/collections/libcollections.la
./src/defobj/.libs/libdefobj.la
./src/defobj/.libs/libdefobj.a
./src/defobj/libdefobj.la
./src/misc/.libs/libmisc.a
./src/misc/.libs/libmisc.la
./src/misc/libmisc.la
./src/objectbase/.libs/libobjectbase.a
./src/objectbase/.libs/libobjectbase.la
./src/objectbase/libobjectbase.la
./src/random/.libs/librandom.a
./src/random/.libs/librandom.la
./src/random/librandom.la
./src/simtools/.libs/libsimtools.a
./src/simtools/.libs/libsimtools.la
./src/simtools/libsimtools.la
./src/space/.libs/libspace.a
./src/space/.libs/libspace.la
./src/space/libspace.la
./src/simtoolsgui/.libs/libsimtoolsgui.a
./src/simtoolsgui/.libs/libsimtoolsgui.la
./src/simtoolsgui/libsimtoolsgui.la
./src/tclobjc/libtclobjc.la
./src/tclobjc/.libs/libtclobjc.a
./src/tclobjc/.libs/libtclobjc.la
./src/tkobjc/.libs/libtkobjc.a
./src/tkobjc/.libs/libtkobjc.la
./src/tkobjc/libtkobjc.la
./src/.libs/libswarm.0.0.0
./src/.libs/libswarm.0
./src/.libs/libswarm
./src/.libs/libswarm.a
./src/.libs/libswarm.lai
./src/.libs/libswarm.la
./src/libswarm.la
./java/c/.libs/libjavaswarmstubs.a
./java/c/.libs/libjavaswarmstubs.la
./java/c/.libs/libjavaswarm.1.0.0
./java/c/.libs/libjavaswarm.1
./java/c/.libs/libjavaswarm
./java/c/.libs/libjavaswarm.lai
./java/c/.libs/libjavaswarm.a
./java/c/.libs/libjavaswarm.la
./java/c/.libs/libjavaswarm.1.0.0T
./java/c/libjavaswarmstubs.la
./java/c/libjavaswarm.la


Maybe this happens because of my tools.  I used
automake-1.6
libtool-1.5.6
autoconf-2.59

I bet the autoconf is the wrong one, isn't it?  Should be 2.57..?

pj


Marcus G. Daniels wrote:
> Scott Christley wrote:
>
> > InputStream.m: At top level:
> > InputStream.m:428: internal compiler error: in encode_type, at
> > objc/objc-act.c:7393
>
>
> This one should be fixed in the current GCC 4.0 sources and leading edge 
> of 3.4 in CVS.   The latter is your best bet if you want to build Swarm 
> using a current GCC.  Also update your CVS checkout of Swarm.
> _______________________________________________
> Support mailing list
> address@hidden
> http://www.swarm.org/mailman/listinfo/support


--
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ku.edu/~pauljohn
1541 Lilac Lane, Rm 504
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66044-3177           FAX: (785) 864-5700