Intel® oneAPI Base Toolkit
Support for core tools and libraries to build and deploy high-performance data-centric applications
Announcements
The Intel sign-in experience is changing in February to support enhanced security controls. If you sign in, click here for more information.
338 Discussions

Segmentation fault for large array using MKL while compiled with OpenMP flag

SamW1
Beginner
769 Views

My target is to implement the Intel MKL solver into my SMP program to replace the current SOR method. I hope it would be faster.

I altered the sample code of Preconditioned CG to solve a linear system of i,j,k = 100 (uploaded code)

If the code compiled with

ifort -qmkl cg_jacobi_precon.f90

it runs normally with implicit parallization as expected.

However,  if I compile it with qopenmp flag (not even adding openMP directives inside code)

ifort -qmkl -qopenmp -mcmodel=large -shared-intel cg_jacobi_precon.f90

and run with ./a.out , sigmentation fault occurred.

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source
libifcoremt.so.5   0000149902209359  for__signal_handl     Unknown  Unknown
libpthread-2.27.s  00001498F61B4980  Unknown               Unknown  Unknown
libmkl_avx512.so.  00001498F501C64C  mkl_spblas_lp64_a     Unknown  Unknown
libmkl_intel_thre  00001498FE44E595  mkl_spblas_lp64_d     Unknown  Unknown
libiomp5.so        00001498F944A893  __kmp_invoke_micr     Unknown  Unknown
libiomp5.so        00001498F93BDCB3  Unknown               Unknown  Unknown
libiomp5.so        00001498F93BEF7D  __kmp_fork_call       Unknown  Unknown
libiomp5.so        00001498F937A425  __kmpc_fork_call      Unknown  Unknown
libmkl_intel_thre  00001498FE44E11B  mkl_spblas_lp64_d     Unknown  Unknown
libmkl_intel_thre  00001498FE1CD051  mkl_spblas_lp64_m     Unknown  Unknown
a.out              0000000000401CC0  Unknown               Unknown  Unknown
a.out              0000000000400F62  Unknown               Unknown  Unknown
libc-2.27.so       00001498F5DD2C87  __libc_start_main     Unknown  Unknown
a.out              0000000000400E6A  Unknown               Unknown  Unknown

If I lower the size of the system to i, j = 100 , k = 10 , everything seems fine.

Back to i,j,k=100, I tried

ulimit -s unlimited

export KMP_STACKSIZE=1G

but nothing solved.

I am not familiar to Intel-MKL, please tell me if I've done anything wrong with it.

 

※By the way, 100^3 isn't really a large array, I do computational simulation with a 10x larger system on the same environment with openMP and no problem occur. 

 

Same question posted on Intel-Fortran-Compiler Section and no solution yet.

https://community.intel.com/t5/Intel-Fortran-Compiler/Segmentation-fault-for-large-array-using-MKL-w...

0 Kudos
6 Replies
VidyalathaB_Intel
Moderator
572 Views

Hi,


Thanks for reaching out to us.

The issue raised by you is reproducible. We are working on it, we will get back to you soon.

Meanwhile could you please let us know the MKL version being used in this case?


Regards,

Vidya.


SamW1
Beginner
554 Views

I have "2022.1.0" and "latest" under /opt/intel/oneapi/mkl/

If 2022.1.0 is the version of MKL. Thank you.

VidyalathaB_Intel
Moderator
492 Views

Hi,

 

Thanks for the details.

Here is an update regarding the issue which you have raised.

 

We tested the sample code which you have provided and based on our observations, the issue is not because of the mkl calls. Here is the modified reproducer test.f90 to avoid all oneMKL calls and then removed "-qmkl" from the compilation line. The updated reproducer still segfaults when compiled with "-qopenmp".

 

We are working on this issue and will be back as soon as we get an update.

 

Regards,

Vidya.

 

VidyalathaB_Intel
Moderator
350 Views

Hi @SamW1 ,

 

Thanks for your patience.

Here is the reason why we are getting segmentation fault when including -qopenmp flag

Without openmp flag, the large arrays in the code are globals

With openmp flag, they are local to the MAIN function and blow the stack hence seg-fault error

If we add SAVE attribute to the large arrays in the code then the issue is gone (attaching the modified sample code)

Please find the output of the screenshot below which were tested with the sample reproducer that you have provided (cg_jacobi_precon.f90) as you can see now it works fine both with and without including qopenmp flag.

VidyalathaB_Intel_1-1672929814382.png

 

Please let us know if it resolves the issue and you can get back to us if you still have any queries regarding the same.

 

Regards,

Vidya.

 

VidyalathaB_Intel
Moderator
196 Views

Hi @SamW1 ,

 

As we haven't heard back from you, could you please provide us with an update regarding the issue?

 

Regards,

Vidya.

 

VidyalathaB_Intel
Moderator
76 Views

Hi @SamW1 ,

 

As we haven't heard back from you, we are closing this thread. Please post a new question if you need any additional assistance from Intel as this thread will no longer be monitored.

 

Regards,

Vidya.

 

Reply