The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear Oscar,

Did you try smaller systems, before your 252 atom one ?

Xavier

On 15 Nov 2007, at 08:33, ramirez_o273@utpb.edu wrote:

> Dear Abinitioners,
>      I'm running a structural relaxation of a gold slab with a  
> supercell size of 252 atoms. I'm constraining the positions using  
> the iatfix and natfix variables to allow only the top two layers to  
> relax. Now, oddly, the run terminates less than 10 seconds with the  
> error: Fortran Runtime error: Attempting to allocate negative  
> memory: possible integer overflow, which occurs right in the middle  
> of the assignment of the occupation numbers for the Au atoms in the  
> output file. Could it be that this is a typical error when one does  
> not configure the fortran compiler correclty? By the way, my  
> internal tests and some tutorial input files completed and worked  
> just fine. It's just that with this large supercell, this fortran  
> "negative memory" attempt occured.
>
> I hope this is enough information to shed light on the cause of  
> just a minor problem (I hope). A lengthy input/output files are  
> available upon request.
>
> Best regards,
> Oscar M. Ramirez