topic Problem running PARDISO example in Intel® oneAPI Math Kernel Library

Problem running PARDISO example

Polk__Jay — Sat, 07 Sep 2019 05:50:19 GMT

Hi--I am trying to use the PARDISO solver on a MacOS and get this error when I try to complie and run the pardiso_unsym_f.f file. I'm compiling like this:

gfortran pardiso_unsym_f.f -o pardiso_unsym_fexec ${MKLROOT}/lib/libmkl_intel_ilp64.a ${MKLROOT}/lib/libmkl_intel_thread.a ${MKLROOT}/lib/libmkl_core.a -liomp5 -lpthread -lm -ldl

and I get this error when I run it:

LMC-062490:Fortran jpolk$ ./pardiso_unsym_fexec
Reordering completed ...
The following ERROR was detected: -1
STOP 1

I would appreciate any help!

Hello,

Kirill_V_Intel — Mon, 09 Sep 2019 20:47:29 GMT

Hello,

As documentation states (https://software.intel.com/en-us/mkl-developer-reference-c-pardiso#E44B4021-701A-48DA-BA29-70CFA20766AA) this error code means that the input is inconsistent. So I assume something is wrong with the matrix or input parameters.

You can try to turn on the matrix checker, i.e. set iparm(25)=1 (https://software.intel.com/en-us/mkl-developer-reference-fortran-pardiso-iparm-parameter) and see whether you'll get more verbose information.

It is hard to say more without looking at your code and input data.

Best,
Kirill

Hi Kirill, I discovered that

Polk__Jay — Tue, 10 Sep 2019 03:47:58 GMT

Hi Kirill, I discovered that I need to compile with lp64 instead of Ilp64. The example file only works with the 32-bit integer option, apparently. Or maybe I needed to change the type of some of the integer inputs to get it to work with the ilp64 option--maybe that's why it was giving an input inconsistency error. Thanks for looking into it!

Jay

for ILP64 mode: you may

Gennady_F_Intel — Tue, 10 Sep 2019 09:47:47 GMT

for ILP64 mode: you may redefine all integers as MKL_INT data types and linking your case with -i8 compiler option in the case if you use intel Fortran compiler or -fdefault-integer-8 -m64 options in the case of gnu compiler.