segv in weight generation from mpas to ne60 conservative

[oehmke]

Ryan found this when he was working on an issue in ESMPy...

Hi Bob, it's been a while, hope you are doing well!

While looking into an issue with incorrect test reporting counts in ESMPy I ran into a segv with the offline weight generation. It is test case 7 in the esmpy regridweightgen test, I was able to reproduce it with the offline app:

(vesmpy) [Linux.gfortran.64.openmpi.default 1088]$ ./ESMF_RegridWeightGen -s /home/ryan/Dropbox/sandbox/esmf-test-data/grids/mpas_uniform_10242_dual.nc -d /home/ryan/Dropbox/sandbox/esmf-test-data/grids/ne60np4_pentagons_100408.nc -m conserve --src_type ESMFMESH -w mpas2ne60conservesegv.nc
Starting weight generation with these inputs:
Source File: /home/ryan/Dropbox/sandbox/esmf-test-data/grids/mpas_uniform_10242_dual.nc
Destination File: /home/ryan/Dropbox/sandbox/esmf-test-data/grids/ne60np4_pentagons_100408.nc
Weight File: mpas2ne60conservesegv.nc
Source File is in ESMF format
Source Grid is a global grid
Source Grid is an unstructured grid
Use the center coordinates of the source grid to do the regrid
Destination File is in SCRIP format
Destination Grid is a global grid
Destination Grid is an unstructured grid
Use the center coordinates of the destination grid to do the regrid
Regrid Method: conserve
Pole option: NONE
Line Type: greatcircle
Norm Type: dstarea
Extrap. Method: none

Segmentation fault (core dumped)
(vesmpy) [Linux.gfortran.64.openmpi.default 1089]$ ls
ESMF_PrintInfo* ESMF_Regrid* ESMF_Scrip2Unstruct* ESMX_Builder* ''$'\001\340''Z@@'
ESMF_PrintInfoC* ESMF_RegridWeightGen* ESMF_WebServController* PET0.RegridWeightGen.Log

There is a strange file created in place of the weight file output, the PET0 file didn't have anything interesting, just the normal output.

Let me know if you'd like me to dig further, otherwise I'll keep sifting through the ESMPy test output.

Ryan

[anntsay]

Action: Ann to follow up with @rokuingh to see if he has time to dig deeper to resolve this.