Multiwfn forum / segmentation fault during NTO analysis for ORCA SF-TDDFT

Re: segmentation fault during NTO analysis for ORCA SF-TDDFT

dummy@example.com (sobereva) — Thu, 02 May 2024 15:03:07 +0000

Please compress your input files and send me via E-mail, I will test

Re: segmentation fault during NTO analysis for ORCA SF-TDDFT

dummy@example.com (zander_z) — Thu, 02 May 2024 13:01:54 +0000

sobereva wrote:

Hello,
Please note that Multiwfn manual indicated: "Note: If ulimit -s unlimited does not work properly on your system, try to use ulimit -Sn unlimited instead"
So, please try "ulimit -Sn unlimited"

Thanks for your reply! I forgot to mention that ulimit -Sn unlimited didn't work as well.

Re: segmentation fault during NTO analysis for ORCA SF-TDDFT

dummy@example.com (sobereva) — Thu, 02 May 2024 11:11:59 +0000

Hello,

Please note that Multiwfn manual indicated: "Note: If ulimit -s unlimited does not work properly on your system, try to use ulimit -Sn unlimited instead"
So, please try "ulimit -Sn unlimited"

segmentation fault during NTO analysis for ORCA SF-TDDFT

dummy@example.com (zander_z) — Thu, 02 May 2024 04:03:40 +0000

Hi,

I'm using Multiwfn for NTO analysis from ORCA SF-TDDFT output. I encountered segmentation fault during the analysis: (output attached)

Calculating local term of density matrix...
Calculating cross term of density matrix...
forrtl: severe (174): SIGSEGV, segmentation fault occurred----] 50.0 % |
Image PC Routine Line Source
multiwfn 00000000020257F3 Unknown Unknown Unknown
libpthread-2.31.s 000014934C25C910 Unknown Unknown Unknown
multiwfn 00000000011AF5FC mkl_blas_def_xdge Unknown Unknown
multiwfn 00000000010EFA5E Unknown Unknown Unknown
multiwfn 00000000010E6574 Unknown Unknown Unknown
multiwfn 000000000044312F Unknown Unknown Unknown
multiwfn 0000000000986AF1 Unknown Unknown Unknown
multiwfn 0000000000968189 Unknown Unknown Unknown
multiwfn 0000000000854327 Unknown Unknown Unknown
multiwfn 000000000042B922 Unknown Unknown Unknown
libc-2.31.so 000014934C08424D __libc_start_main Unknown Unknown
multiwfn 000000000042B829 Unknown Unknown Unknown

I have set ulimit -s unlimited. Running cat /proc/sys/kernel/shmmax returns 18446744073709551615

For one molecule (44 atoms), the NTO analysis can be done successfully for the excited state 1 but had segmentation fault for the rest of states. The other molecule I tested (34 atoms), the NTO analysis encountered segmentation fault for any states during the step of calculating cross term of density matrix.

I suspect it's due to some sort of insufficient memory, but I don't know what's the limiting part and how to adjust.

Thank you for your help!