[CIG-SEISMO] SPECFEM3D: Segmentation fault - xgenerate_databases

gmatharu at eos.ubc.ca gmatharu at eos.ubc.ca
Thu Feb 14 16:56:34 PST 2013


So I seemed to have traced the issue down to how the tomography file is
being read in. The error report directs me to line 360 in the
model_tomography.f90 source code. I compared an older version and to the
current one and there has been the introduction of an imaterial variable
(seems more for CUBIT users). I checked the latest manual which suggests
two ways of specifying a tomography model. I chose 1) set MODEL=tomo in
Par_file as I had previously done.

if( imaterial_id < 0 .and. trim(undef_mat_prop(2,abs(imaterial_id))) ==
'tomography' ) then

Is the line causing the error. Is this an error in the way I've specified
the input? Or does the code no longer allow input in the way it used to?

Regards,

Gian



> Hi
>
>  I have been running SPECFEM3D on a couple of different clusters on the
> Westgrid computer consortium. As I'm not sure what details are relevant
> I've included links which explain the builds of either cluster.
>
> http://www.westgrid.ca/support/quickstart/lattice
> http://www.westgrid.ca/support/quickstart/orcinus
>
> I'm currently generating a mesh on 256 processors which is 256*128*60
> (X/Y/Z). A few months back on an older build of the software I had no
> issues generating a mesh on the orcinus cluster and running the solver.
> However with an updated package I am encountering segmentation fault
> errors when running xgenerate_databases on both Lattice and Orcinus (using
> ifort,icc and mpif90 on both). My submission scripts seem correct and the
> initial meshing seems to work ok, however I'm at a loss as to what's
> causing this error. Do you have any recommendations on how to resolve this
> issue? If you require further information let me know.
>
> Best Regards,
>
> Gian
>
> Below is a snippet of the error message I receive.
>
>
> starting MPI internal mesher on 256 processors
>
> generating databases...
> cp: cannot stat `/OUTPUT_FILES/output_mesher.txt': No such file or
> directory
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
> Image              PC                Routine            Line        Source
> libifcore.so.5     00002B8064AF3683  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000473530  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000045248E  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000417D08  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000416351  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040D10B  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BD09  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BC9C  Unknown               Unknown
> Unknown
> libc.so.6          0000003F7FA1D994  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BBA9  Unknown               Unknown
> Unknown
> forrtl: severe (174): SIGSEGV, segmentation fault occurred
> Image              PC                Routine            Line        Source
> libifcore.so.5     00002B2EB0DF5683  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000473530  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000045248E  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000417D08  Unknown               Unknown
> Unknown
> xgenerate_databas  0000000000416351  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040D10B  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BD09  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BC9C  Unknown               Unknown
> Unknown
> libc.so.6          000000336A21D994  Unknown               Unknown
> Unknown
> xgenerate_databas  000000000040BBA9  Unknown               Unknown
> Unknown
>
>




More information about the CIG-SEISMO mailing list