Hessian calculation runs into segmentation fault when atoms are frozen

[TamasFoldes]

Describe the bug
I attempted to calculate the Hessian of a system with several atoms frozen. The calculation terminated with an error showing segmentation fault:

writing file \<hessian>.

 projected vibrational frequencies (cm⁻¹)

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0  0x153a79b5fd01 in ???
#1  0x153a79b5eed5 in ???
#2  0x153a797e720f in ???
#3  0x556310f7e468 in ???
#4  0x556310e711a8 in ???
#5  0x556310e6bf98 in ???
#6  0x153a797c80b2 in ???
#7  0x556310e6bffd in ???
#8  0xffffffffffffffff in ???
Segmentation fault (core dumped)

To Reproduce
Input and output files are attached.

Expected behaviour
Normal termination with the results printed out.

Additional context
output.txt
structure.txt
xcontrol.txt

[awvwgk]

Thanks for reporting, this looks indeed faulty. I'll have to investigate what is throwing up here.

[jiachenchem]

A little bit more detail for this issue. xtb 6.4.1 seems to work if number of freezing atoms is smaller than 7. Control and output files are attached. When number of freezing atoms are larger, then I have the segmantation fault.
xcontrol_2.txt
output_2.txt

[TamasFoldes]

Thank you for the notification! I tested v 6.4.1 when it came out and noticed the limitation. Is there a plan to further extend the number of frozen atoms possible in a single calculation?

[lxsep]

Hello, I am trying to do the same in version 6.6.1 and having the same error as in here. I already tried to extend the memory for the calculation and it is not working.