List Info

Thread: LAM: Test suite 7.1.2 on Mac OS X: Make errors




LAM: Test suite 7.1.2 on Mac OS X: Make errors
user name
2006-05-10 13:50:45

Hi Didier, 

* Didier Lucor wrote on Wed, May 10, 2006 at 02:15:35PM CEST:
> >
> > I'm using the latest LAM-MPI 7.1.2 (Mac OS X Package:
> > lam-7.1.2.dmg.gz) on my G5 (OS 10.4.6; Dual 2.3 GHz PowerPC) and
> > I'm trying to run the LAM 7.1.2 test suite. 

> > While compiling the "comm" module, it seems that there is a
> > multiple definitions of symbol __cxa_atexit that causes conflict
> > during the linking. At this point it won't generate the executables
> > and just stops compiliing. 

> > mpic++ -o comm_get_name_cxx comm_get_name_cxx.o -L../reporting
> > -llamtest
> > /usr/bin/ld: multiple definitions of symbol ___cxa_atexit
> > /usr/local/lib/gcc/powerpc-apple-darwin8.6.0/4.2.0/crt3.o private
> > external definition of ___cxa_atexit in section (__TEXT,__text)
> > /usr/lib/libdl.dylib(atexit.So) definition of ___cxa_atexit
> > collect2: ld returned 1 exit status
> > mpic++: No such file or directory
> > make[1]: *** [comm_get_name_cxx] Error 1
> > make: *** [all-recursive] Error 1 

Please go into the 'comm' subdirectory of your build tree and issue
  mpic++ -v -o comm_get_name_cxx comm_get_name_cxx.o -L../reporting \
    -llamtest 

post all output. I have no idea what's going on here. 

Cheers,
Ralf 


%%%%%%%%%%%

Thank you Ralph.
For some reason, I've not received your email (maybe I'm doing something wrong here with the mailing list?)
This is what I've got:


> mpic++ -v -o comm_get_name_cxx comm_get_name_cxx.o -L../reporting -llamtest

Using built-in specs.
Target: powerpc-apple-darwin8.6.0
Configured with: ../gcc-4.2-20060318/configure --enable-languages=c,c++,fortran,java
Thread model: posix
gcc version 4.2.0 20060318 (experimental)
/usr/local/libexec/gcc/powerpc-apple-darwin8.6.0/4.2.0/collect2 -dynamic -arch ppc -macosx_version_min 10.3 -multiply_defined suppress -weak_reference_mismatches non-weak -o comm_get_name_cxx -lcrt1.o /usr/local/lib/gcc/powerpc-apple-darwin8.6.0/4.2.0/crt2.o -lcrt3.o -L../reporting -L/usr/local/lib -L/usr/local/lib/gcc/powerpc-apple-darwin8.6.0/4.2.0 -L/usr/local/lib/gcc/powerpc-apple-darwin8.6.0/4.2.0/../../.. comm_get_name_cxx.o -llamtest -u _lam_darwin_malloc_linker_hack -multiply_defined suppress -force_flat_namespace -flat_namespace -multiply_defined suppress -llammpio -llammpi++ -lmpi -llam -ldl -lstdc++ -lgcc_s.10.4 -lgcc -lSystemStubs -lSystem
/usr/bin/ld: multiple definitions of symbol ___cxa_atexit
/usr/local/lib/gcc/powerpc-apple-darwin8.6.0/4.2.0/crt3.o private external definition of ___cxa_atexit in section (__TEXT,__text)
/usr/lib/libdl.dylib(atexit.So) definition of ___cxa_atexit
collect2: ld returned 1 exit status
mpic++: No such file or directory

best regards,

Didier
LAM: Test suite 7.1.2 on Mac OS X: Make errors
user name
2006-05-10 14:09:42
* Didier Lucor wrote on Wed, May 10, 2006 at 03:50:45PM
CEST:
>
> For some reason, I've not received your email (maybe
I'm doing  
> something wrong here with the mailing list?)

I have no idea.  I've Cc:ed you now.

> This is what I've got:
> 
> > mpic++ -v -o comm_get_name_cxx comm_get_name_cxx.o
-L../reporting - 
> llamtest

Ouch. I meant
  mpic++ -showme -o ... 

Sorry about that.

I have a vague feeling that it's due to an old Autoconf
bug[1] lingering
somewhere, but I can't see where Fortran would come into
play here.
Another reason to get Autoconf-2.60 out the door soon,
luckily that
seems pretty near now.

Cheers,
Ralf

[1] http://lists.gnu.org/archive/html/autoco
nf-patches/2004-02/msg00010.html
_______________________________________________
This list is archived at http://www.l
am-mpi.org/MailArchives/lam/
[1-2]

about | contact  Other archives ( Real Estate discussion Medical topics )