[CIG-SEISMO] SPECFEM2d: "Integer overflow when calculating the amount of memory to allocate"

[Dimitri Komatitsch]

Dear Sumedh,

It is probably a 4-byte integer counter that reaches the 2GB limit in 
the case of an extremely large mesh. Easy to fix. Please email me your 
input files (no need to cc the whole mailing list) so that I can test 
and I will fix it and commit the changes to SVN.

Thank you,
Best regards,

Dimitri.

On 10/11/2013 10:21 PM, Sumedh Joshi wrote:
> Hi Folks,
>
> I've been getting the following error when running a particularly large
> job while running meshfem2d:
>
> fortran runtime error: Integer overflow when calculating the amount of
> memory to allocate
>
> The job then fails.  This only happens for larger runs (6000 x 4000
> elements with 9 grid points per element), and I'm wondering if I'm just
> hitting the memory limit of my system (a 64-core 128 GB machine).  Has
> anyone seen this error before?  Is my 6000 x 4000 x 9 unreasonably large
> for my machine?  I wouldn't think so, as 6000 x 4000 x 9 x 16 bytes ~= 3
> GB, but I'm not sure.
>
> Thanks!
>
> Sumedh
>
>
> _______________________________________________
> CIG-SEISMO mailing list
> CIG-SEISMO at geodynamics.org
> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-seismo
>

-- 
Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France    http://komatitsch.free.fr