topic I will try to provide it. in Intel® oneAPI Math Kernel Library

mkl scalapack + dapl fails

Marius_B_ — Sun, 17 Sep 2017 16:17:30 GMT

Hi,

I try to run mkl scalapack in fortran code on an infinity band network using I_MPI_FABRICS=shm:dapl, however mkl scalapack does not work correctly when running on several nodes, e.g. pzheev exits with error code 16. When switching to shm:tcp network fabrics it works. It also works with netlib scalapack reference implementation + mkl lapack/blas. I tried intel-2016 update 4 and intel-2017 update 4, both give the same errors.

Any idea on this error ?

best,

marius

this is an unknown problem

Gennady_F_Intel — Mon, 18 Sep 2017 03:21:16 GMT

this is an unknown problem with MKL 2017 u3. If pzheev works with Netlib's implementation and fails with MKL, then probably, this is the bug. Could you give us reproducer?

thanks for the swift reply.

Marius_B_ — Mon, 18 Sep 2017 03:35:49 GMT

thanks for the swift reply.

I did some more testing.I used to use -mkl=cluster for which I got the errors mentioned above, however when using the linker options supplied by intel mkl link advisor, i.e. " ${MKLROOT}/lib/intel64/libmkl_scalapack_lp64.a -Wl,--start-group ${MKLROOT}/lib/intel64/libmkl_intel_lp64.a ${MKLROOT}/lib/intel64/libmkl_sequential.a ${MKLROOT}/lib/intel64/libmkl_core.a ${MKLROOT}/lib/intel64/libmkl_blacs_intelmpi_lp64.a -Wl,--end-group -lpthread -lm -ldl", it works. Using -mkl=cluster together with TCP fabrics works, tough.

So is it in general not a good idea to use "-mkl=cluster" over giving the linker options explicitly?

This is strange because of

Gennady_F_Intel — Mon, 18 Sep 2017 05:19:48 GMT

This is strange because of when we use -mkl=cluster compiler option, that's mean we link with sequential MKL cluster components and these components are used Intel MPI. Here is should be identical what you explicitly used.

The difference is that -mkl

Marius_B_ — Tue, 19 Sep 2017 00:35:57 GMT

The difference is that -mkl=cluster links the libraries are linked dynamically while following link advisor links them statically. If I use -mkl=cluster -static-intel -static_mpi to statically link the mkl libraries scalapack works.

then, could you give us the

Gennady_F_Intel — Tue, 19 Sep 2017 02:12:57 GMT

then, could you give us the reproducer?

I will try to provide it.

Marius_B_ — Thu, 05 Oct 2017 23:53:15 GMT

I will try to provide it.