VASP aborting ... with vasp.6.4.2+vtstcode6.4

[MCV]

Dear VTST developpers,

I have compiled vasp.6.4.2 with vtstcode6.4 according to instructions in your group [webpage](https://theory.cm.utexas.edu/vtsttools/ ... ation.html). I have done that in two HPC clusters with the help of my institue's computer engineer. Although plain vasp calculations work perfectly, we don't manage run VTST and this is what we get in the standard ouput:

running 36 mpi-ranks, on 1 nodes
distrk: each k-point on 9 cores, 4 groups
distr: one band on 9 cores, 1 groups
vasp.6.4.2 20Jul23 (build Jul 1 2024 14:43:42) complex
POSCAR found : 3 types and 43 ions
scaLAPACK will be used
LDA part: xc-table for Pade appr. of Perdew
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...
VASP aborting ...

Do you have a hint on why this happens? Could you please help us to solve this issue?

Thank you very much.

Kind regards,
MCV

[MCV]

P.S.: I forgot to mention one additional comment. Below there are the last lines writtent in the OUTCAR file. There is a message from the VTST script that makes no sense.

"[...]
VTST: version 4.1, (05/19/21)

CHAIN: initializing optimizer

OPT: Using Conjugate-Gradient optimizer
OPT: CG, Init
OPT: CG, FDSTEP 0.005000
OPT: CG, MAXMOVE 0.200000

Must set EDIFFG < 0 when using IOPT > 0"

EDIFFG is defenitely <0 in my INCAR file.

Regards,
MCV

[graeme]

If you post a .tar.gz of your calculation, I can take a look.

[MCV]

Hello,

There it is.

Thanks.
MCV

[graeme]

change:

EDDIFG = -0.01

to

EDIFFG = -0.01

[MCV]

Dear Graeme,

Oops, my bad! Sorry for such an improductive thread in the forum.

Despite many years working with VASP, I'm afraid I can still produce this kind of typos...

Thanks a lot,
MCV

[graeme]

I'm right there with you!