Intel® oneAPI Math Kernel Library
Ask questions and share information with other developers who use Intel® Math Kernel Library.
7266 ディスカッション

Segmentation default caused by DFTI function

TravelerPhys
ビギナー
1,869件の閲覧回数

I wrote a c language code in which DFTI is used, and it is attached. The filename is "test.c"

The compilation command line is " icx test.c -g -qmkl=parallel -lgsl -lgslcblas -lmkl_core -lmkl_intel_lp64 -lmkl_intel_thread -liomp5 -lpthread -qopenmp -lm -o test ". 

I then used gdb: " gdb ./test ". 

It can be seen that after DftiCommitDescriptor(fft_handle) was called, the address that the pointer "uv" pointed to cannot be accessed.

I would be very grateful if this problem could be solved.

ラベル(1)
0 件の賞賛
3 返答(返信)
IntelSupport
管理者
1,815件の閲覧回数

Hi,

 

Thank you for posting in Intel Communities..

 

Please use below command line to compile the attached test code.

" icx test.c -g -qmkl=parallel -lmkl_core -lmkl_intel_lp64 -lmkl_intel_thread -liomp5 -lpthread -qopenmp -lm -o test ".

 

We have removed "-lgsl -lgslcblas " flags from command line as we could see that your code is not using any GSL specific implementations. We could see the code is working fine without any issues after this change.

 

Best Regards,

Jilani Shaik

 

IntelSupport
管理者
1,696件の閲覧回数

Hi,


A gentle reminder:

Has the information provided helped? Could you please let us know if you need any other details?


Regards,

Jilani


IntelSupport
管理者
1,625件の閲覧回数

Hi,


A gentle reminder:

Has the information provided helped? Could you please let us know if you need any other details?


Regards,

Jilani


返信