Error running dimrun example VTST 2.04b & 2.03b VASP 4.6.36

[mtgarvey]

I have two versions of VASP 4.6.36 compiled on different clusters with different versions of VTST (2.03b and 2.04b). Both give me the same error when I try to run the example contained in dimrun.tar.gz downloaded from the UT site. The stdout is given below:

VTST 2.04b
------------------------------------------------------------------------
running on 16 nodes
distr: one band on 1 nodes, 16 groups
vasp.4.6.36 17Feb09 complex
POSCAR found : 2 types and 5 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR found : 2 types and 5 ions
The IMAGES tag should not be set for this implementation of the dimer method.
------------------------------------------------------------------------

VTST 2.03b
------------------------------------------------------------------------
running on 12 nodes
distr: one band on 1 nodes, 12 groups
vasp.4.6.36 17Feb09 complex
POSCAR found : 2 types and 5 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
The IMAGES tag should not be set for this implementation of the dimer method.
------------------------------------------------------------------------

I have seen in a previous post that using MPI_CHAIN could cause this. There is no -DMPI_CHAIN in my makefile.

I do not have any other example files to run, but the compiled code does run the VASP bench_Hg example properly.

Is there anyone who has an idea as to what is causing this? Thanks in advance for any advice.

Mike Garvey
University of Wisconsin-Milwaukee

[graeme]

Do you have the IMAGES tag set in the INCAR?

[mtgarvey]

Professor Henkelman,

Thanks for the swift response. The IMAGES tag is not set in the INCAR file. Sorry, I should have stated that right away. I have included the contents of the INCAR file below, it should be the same as it is in the example.

INCAR
--------------------------------------------------------------
SYSTEM = Typical Dimer Run
IBRION=3
POTIM=0.0
ISYM=0
NSW=600
EDIFF=1E-7
EDIFFG=-0.001
PREC=Norm
LREAL=.FALSE.
LWAVE=.FALSE.
LCHARG=.FALSE.

# DIMER PARAMETERS
ICHAIN=2
DdR=0.005
DRotMax=4
DFNMin=0.01
DFNMax=1.0

# OPTIMIZER PARAMETERS
IOPT=2
--------------------------------------------------------------

[graeme]

This is very strange. In fact, I don't think that MPI_CHAIN is ever set anymore. So I'm not sure how your compiler is including the statement which is printing the message.

If you have time to run a check, I would be interested to know if it is fixed when you remove the white space before the

#if defined(MPI_CHAIN)
...
#endif

in dimer.F so that it reads:

#if defined(MPI_CHAIN)
...
#endif

If this doesn't fix it, you can delete everything between and including those lines. This should really be cleaned up in the code anyway.

[mtgarvey]

graeme wrote:
> If you have time to run a check, I would be interested to know if it is fixed when
> you remove the white space before the
>
> #if defined(MPI_CHAIN)
> ...
> #endif
>
> in dimer.F so that it reads:
>
> #if defined(MPI_CHAIN)
> ...
> #endif

This worked for 2.03b. I was able to run the dimrun example and it seems to be in reasonable agreement with the output from the website. Thanks for the help, I will pass this along to our hpc facilitator to see if it solves the problem with 2.04b on our larger cluster.

I have tried it on 2.04b on our departmental cluster and I am getting a new error when compiling. However, I cannot be certain it is related to this because I am not sure if I ever successfully compiled 2.04b on our cluster. I will include the error below. Maybe it will be obvious what the problem is.

Error when compiling VASP 4.6 w/ VTST 2.04b
-----------------------------------------------------------------------------------------------------------------------------------------------------------
dimer.F(494): error #6460: This is not a field name that is defined in the encompassing structure. [TYPE]
IF (T_INFO%TYPE(1)/=' ') THEN
-----------------^
dimer.F(494): error #6158: The structure-name is invalid or is missing. [T_INFO]
IF (T_INFO%TYPE(1)/=' ') THEN
----------^
compilation aborted for dimer.F (code 1)
make: *** [dimer.o] Error 1
-----------------------------------------------------------------------------------------------------------------------------------------------------------

[graeme]

Ok, that we have seen; here is the fix:
viewtopic.php?f=2&t=1185

[mtgarvey]

Thanks, that seems to have done it. I only had one more small issue after that. I had to remove the white space in front of all of the #if and #endif lines in chain.F Now both versions can run the dimrun example and give the same results which seem to be close to the results from the website.

[ckande]

I also faced the same problem with VASP 5.2.12 and Intel 12.1.1.256. The issue got fixed after removing the lines with #if defined (MPI_CHAIN)... #endif in dimer.F.

[mwchang]

[quote="ckande"]I also faced the same problem with VASP 5.2.12 and Intel 12.1.1.256. The issue got fixed after removing the lines with #if defined (MPI_CHAIN)... #endif in dimer.F.[/quote]




I also encountered the same problem.

As your suggestion, It can be solved by removing the lines of #if defined (MPI_CHAIN)... #endif in dimer.F


Thanks.