Armadillo code:
#define ARMA_DONT_USE_WRAPPER
#define ARMA_USE_BLAS
#define ARMA_USE_LAPACK
arma::mat A = { { 0.0013 , 0.1741 , 0.9885 , 0.1662 , 0.8760 } ,
{ 0.1933 , 0.7105 , 0.1191 , 0.4508 , 0.9559 } ,
{ 0.5850 , 0.3040 , 0.0089 , 0.0571 , 0.5393 } ,
{ 0.3503 , 0.0914 , 0.5317 , 0.7833 , 0.4621 } ,
{ 0.8228 , 0.1473 , 0.6018 , 0.5199 , 0.8622 } };
A.print("A: ");
arma::mat B = arma::inv(A);
B.print("inv(A): ");
arma::mat I = A*B;
I.print("I: ");
MKL targets from CMAKE:
[cmake] -- MKL_VERSION: 2024.2.0
[cmake] -- MKL_ROOT: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2
[cmake] -- MKL_ARCH: None, set to ` intel64` by default
[cmake] -- MKL_LINK: None, set to ` dynamic` by default
[cmake] -- MKL_INTERFACE_FULL: None, set to ` intel_ilp64` by default
[cmake] -- MKL_THREADING: None, set to ` intel_thread` by default
[cmake] -- MKL_MPI: None, set to ` intelmpi` by default
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_scalapack_ilp64_dll.lib
[cmake] -- Found DLL: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/bin/mkl_scalapack_ilp64.2.dll
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_cdft_core_dll.lib
[cmake] -- Found DLL: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/bin/mkl_cdft_core.2.dll
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_intel_ilp64_dll.lib
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_intel_thread_dll.lib
[cmake] -- Found DLL: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/bin/mkl_intel_thread.2.dll
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_core_dll.lib
[cmake] -- Found DLL: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/bin/mkl_core.2.dll
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/lib/mkl_blacs_ilp64_dll.lib
[cmake] -- Found DLL: C:/Program Files (x86)/Intel/oneAPI/mkl/2024.2/bin/mkl_blacs_ilp64.2.dll
[cmake] -- Found C:/Program Files (x86)/Intel/oneAPI/compiler/latest/lib/libiomp5md.lib
[cmake] -- Imported oneMKL targets: MKL::mkl_scalapack_ilp64;MKL::mkl_cdft_core;MKL::mkl_intel_ilp64;MKL::mkl_intel_thread;MKL::mkl_core;MKL::mkl_blacs_ilp64;MKL::MKL
After starting the application crashes, and the following message is shown: Intel oneMKL ERROR: Parameter 1 was incorrect on entry to DGETRF.
Armadillo can be configured via editing the file include/armadillo_bits/config.hpp:
ARMA_DONT_USE_WRAPPER -> Disable going through the run-time Armadillo wrapper library (libarmadillo.so) when calling LAPACK, BLAS, ARPACK, and SuperLU functions. You will need to directly link with BLAS, LAPACK, etc (eg. -lblas -llapack)
ARMA_USE_LAPACK -> Enable use of LAPACK, or a high-speed replacement for LAPACK (eg. OpenBLAS, Intel MKL, or the Accelerate framework). Armadillo requires LAPACK for functions such as svd(), inv(), eig_sym(), solve(), etc.
ARMA_DONT_USE_LAPACK -> Disable use of LAPACK; overrides ARMA_USE_LAPACK
they were defined in the code, but the error persists.
I am working in Qt Creator IDE 14, using MSVC 2022 (64bit) (Qt 6.8), Armadillo 14.0.3, and Intel OneAPI MKL 2024.2
OneAPI MKL may be configured to use 64 bit ints. Standard BLAS and LAPACK uses 32 bit ints.
Enable the ARMA_BLAS_LONG_LONG option before including the Armadillo header to force the use of 64 bit ints when calling BLAS and LAPACK functions. This can be done like so:
#define ARMA_DONT_USE_WRAPPER
#define ARMA_BLAS_LONG_LONG
#include <armadillo>