Access violation when using Intel MKL Pardiso in 32 bit.

[omar....@demo3d.com]

I'm getting an access violation in pardiso (on the following call in patdiso_loader.c) when running with 32-bit:

func_pardiso(pt, maxfct, mnum, mtype, phase, n, a, ia, ja, perm, nrhs, iparm, msglvl, b, x, error);

It works fine when I build for 64-bit. This is with just the osqp_demo. Has anyone managed to get Intel MKL Pardiso working in 32-bit with OSQP?

Thanks

[Paul Goulart]

Could you provide a bit more information about how you produced this error?   What is the OS, what is the exact error message that you got and how did you produce it?

In particular, it's unclear to me whether you are trying to compile with the DFLOAT option on a 64-bit OS or whether you are compiling on a 32 bit OS.

[omar....@demo3d.com]

Hi Paul,

I am just running the included demo. The only change I made was to set the linear solver to MKL_PARDISO_SOLVER. I am using the Intel C++ compiler under Visual Studio.

The OS is Windows 10 (64-bit).

I'm not compiling with DFLOAT enabled, though DLONG is defined and I have just found that undefining DLONG does fix the issue. Thanks for pointing me towards this!