Dear,
I am using the interface dfeast_scsrev for computing eigenvalues and eigenvectors of a sparse matrix sorted using a CSR format (3-vector).
It works fine with small sparse matrices with a size of about ~10,000. However, I got a segmentation fault with a sparse matrix of size ~130,000 or bigger.
Below is the error message I got:
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
snpblup 000000000438CD21 tbk_trace_stack_i Unknown Unknown
snpblup 000000000438AE5B tbk_string_stack_ Unknown Unknown
snpblup 00000000043418B4 Unknown Unknown Unknown
snpblup 00000000043416C6 Unknown Unknown Unknown
snpblup 00000000042EACD7 Unknown Unknown Unknown
snpblup 00000000042EF0B0 Unknown Unknown Unknown
libpthread-2.17.s 00002AAAAACDE100 Unknown Unknown Unknown
snpblup 000000000047D059 Unknown Unknown Unknown
snpblup 0000000000451F4A Unknown Unknown Unknown
snpblup 000000000041DBFA Unknown Unknown Unknown
snpblup 0000000000407A10 Unknown Unknown Unknown
snpblup 000000000040741E Unknown Unknown Unknown
libc-2.17.so 00002AAAABD50B15 __libc_start_main Unknown Unknown
snpblup 0000000000407329 Unknown Unknown Unknown
To run it I used 10 (OMP) threads. I set ulimit -s unlimited and exported KMP_STACKSIZE equal to 20G.
I am not sure where is the issue. Any help would be appreciated.
Thank you.
Jeremie