On Apr 10, 2006, at 9:43 AM, David Gunter wrote:
> After much fiddling around, I managed to create a version of open-mpi
> that would actually build. Unfortunately, I can't run the simplest
> of applications with it. Here's the setup I used:
> export CC=gcc
> export CXX=g++
> export FC=gfortran
> export F77=gfortran
> export CFLAGS="-m32"
> export CXXFLAGS="-m32"
> export FFLAGS="-m32"
> export FCFLAGS="-m32"
> export LDFLAGS="-L/usr/lib"
That looks about right. I'm surprised you need the LDFLAGS, but I
can see where it might be needed, and I doubt it's causing the
problems you are seeing. Actually, mmm... depending on which LANL
system you are on, that might be needed for gm, so that makes sense.
> ./configure --prefix=/net/scratch1/dog/flash64/openmpi/
> openmpi-1.0.2-32b --build=i686-pc-linux-gnu --with-bproc --with-g
> m --enable-io-romio --with-romio --with-io-romio-flags='--build=i686-
I'm surprised you need the --build= flag, but again, not what is
causing issues here.
> Configure completes, as does 'make' and then 'make install'. Next I
> tried to compile a simple MPI_Send test application, which fails to
> (flashc 104%) gcc -m32 -I/net/scratch1/dog/flash64/openmpi/
> openmpi-1.0.2-32b/include -o send4 send4.c -L/net/scratch1/dog/
> flash64/openmpi/openmpi-1.0.2-32b/lib -lmpi
> warning: epoll_wait is not implemented and will always fail
> warning: epoll_ctl is not implemented and will always fail
> warning: epoll_create is not implemented and will always fail
That's extremely weird. For some reason, we're trying to use the
epoll code, but it clearly isn't supported (as the linker so nicely
informs us). I'm a little curious how this can happen. Would you
mind sending the output of configure as well as the config.log file
it generates? Also, do you know what OS is running on that machine?
Open MPI developer