- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello,
I am trying solve a generalized eigenvalues problem Ax = λBx by means of dfeast_scsrgv, but, I get SIGSEV error, I test the individual matrices A and B with dfeast_scsrev and all works fine, so I think that is not a problem of data representation.
Attached goes a code example for the problem described.
pd: The hermitian version zfeast_hcsrgv has the same issue. Sorry for the english.
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
which version of mkl did you try?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
2019 update 4, in Linux, I try with both ICC and GCC compilers.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
the fix of the issue available in MKL 2020.1.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I would like to reopen this issue as I am having the same problem.
You can use the very same example as given in this thread's initial post.
My OneAPI version is:
ls -l /opt/intel/oneapi/mkl
total 4
drwxr-xr-x 9 root root 4096 Oct 27 22:15 2025.0
lrwxrwxrwx 1 root root 6 Nov 25 12:29 latest -> 2025.0
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
Could you please let me know which compiler you are using? This will help me reproduce the issue you're facing.
Thanks and regards,
Aleksandra
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks Aleksandra for your help and sorry for not replying sooner.
Here is my output:
/opt/intel/oneapi/compiler/latest/bin/icx --version
Intel(R) oneAPI DPC++/C++ Compiler 2025.0.0 (2025.0.0.20241008)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /opt/intel/oneapi/compiler/2025.0/bin/compiler
Configuration file: /opt/intel/oneapi/compiler/2025.0/bin/compiler/../icx.cfg
Does this help?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@Aleksandra_KMy apologies, it seems that dfeast fails only in my program's environment... it runs when i compile it manually.
Will investigate and reply... please do not waste time on this for now.
Thanks,
Daniel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Daniel,
Thank you for the update. I also checked your example manually and it compiled correctly with icpx and the new MKL 2025.0 (there was a small mistake in calling the last function - after m0, the next argument should have been csr_gv_eigenvalues, but it was csr_gv_res, but you probably know about it).
Please let me know if there is something else I can help you in this topic, or confirm that I can close the issue.
Best regards,
Aleksandra
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
OK, I think I got it pin pointed!
I have a task that needs PARDISO call in order to solve linear system of equations first, then I have FEAST call to solve generalized eigenvalue problem.
I need to have them in threads in order to monitor progress.
Segmentation fault occurs in FEAST call if PARDISO call was in thread. If I call PARDISO directly, there is no problem.
Please check the example I provided in a hope that you can reproduce it also.
Problem happens with both gcc and intel compilers (Linux).
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Daniel,
Could you share more details on how you ran the code and the error you encountered? I really want to reproduce it but so far I can't get that error.
Thanks,
Aleksandra
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello Aleksandra,
here are the steps:
- download and extract main.zip
- open terminal in folder where you extracted main.cpp file
- start intel MKL development environment:
source /opt/intel/oneapi/setvars.sh
- compile main.cpp file into executable:
icpx -Wall -g -I/opt/intel/oneapi/mkl/latest/include -c main.cpp -o main.o
icpx -o main main.o -Wl,--start-group /opt/intel/oneapi/compiler/latest/lib/libiomp5.a /opt/intel/oneapi/mkl/latest/lib/libmkl_core.a /opt/intel/oneapi/mkl/latest/lib/libmkl_gnu_thread.a /opt/intel/oneapi/mkl/latest/lib/libmkl_intel_lp64.a -Wl,--no-as-needed -Wl,--end-group -ldl
- run executable:
./main
- Segmentation fault
I did find that using direct compilation:
icpx -I/opt/intel/oneapi/mkl/latest/include -L/opt/intel/oneapi/mkl/latest/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread -lm -ldl main.cpp -o main
will yield runnable executable... now I am trying to figure out the difference as I need separate compile and link times as I am using codeblocks IDE... perhaps you see right away what is wrong with my link line?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Please check with this short linking:
icpx main.cpp -o main -I${MKLROOT}/include -L${MKLROOT}/lib/intel64 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
Does it work or do you get errors?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you Aleksandra!
It works!
I guess it was libmkl_gnu_thread.a conflict that caused problems in my case... anyway I abandoned linking with static libraries.
Thanks,
Daniel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Daniel,
Great to hear that! Is there something else I can help you with this topic or can I close the issue?
Regards,
Aleksandra
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
No, that's it. Thank you, you can close the issue.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page