What happened?
clang and GNU C and C++ compilers with gfortran 16.1, MacOS 26.6 on M-series: Builds and runs fine without HWM14, but when built with HWM14 it will crash after a minute or so running a simulation with a dipole grid. Seems to work fine with gcc 15.2 (M-series) and works fine on MacOS intel (any compiler combo). Seems to work on our HPC with gcc 13.2, as well. Test data here: https://www.dropbox.com/scl/fo/ssmyqo0l0d6530wybvxej/AGn7tc-NMm4Nfw9JM7UFrJI?rlkey=wx99l0dwj6bzrtm1titnb4m7h&dl=0
Relevant log output
4 MPI processes detected
./gemini.bin Process: 1 / 3 at 20260813T102936.789
./gemini.bin Process: 3 / 3 at 20260813T102936.789
./gemini.bin Process: 2 / 3 at 20260813T102936.789
./gemini.bin Process: 0 / 3 at 20260813T102936.789
******************** input config ****************
simulation directory: /Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing
start year-month-day: 2016-03-03
start time: 4500.000
duration: 86400.000
output every: 3600.000
gemini.f90: using input data files:
/Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/simsize.h5
/Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/simgrid.h5
/Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/initial_conditions.h5
no neutral disturbance specified.
no precipitation specified
no Efield specified
no solar flux specified
GLOW disabled
MSISE00 enabled for neutral atmosphere calculations.
EIA enables with peok equatorial drift: 10.000000000000000
Variable background neutral atmosphere disabled.
Background precipitation has total energy flux and energy: 1.0000000000000000E-003 3000.0000000000000
Parallel current calculation disabled.
Inertial capacitance calculation type: 0
Diffusion solve type: 2
Milestone output disabled.
Gravitaional drift terms disabled.
Two-way coupling disabled.
Lagrangian grid disabled
Using default magnetic pole.
Anomalous electrojet heating disabled.
Nonlinear current disabled.
Using legacy model for electron inelastic collisions.
**************** end input config ***************
process grid (Number MPI processes) x2, x3: 2 2
Process:0 at process grid location: 0 0
get_subgrid_size: 3D run
process grid (Number MPI processes) x2, x3: 2 2
Process:2 at process grid location: 0 1
get_subgrid_size: 3D run
process grid (Number MPI processes) x2, x3: 2 2
Process:1 at process grid location: 1 0
get_subgrid_size: 3D run
process grid (Number MPI processes) x2, x3: 2 2
Process:3 at process grid location: 1 1
get_subgrid_size: 3D run
init_storage_root done
gather_grid_root done
calc_coord_diffs_root done
enforce_gridmpi_periodic done
grid_check done
Done with read_grid_in...
enforce_gridmpi_periodic done
grid_check done
Done with read_grid_in...
enforce_gridmpi_periodic done
enforce_gridmpi_periodic done
grid_check done
Done with read_grid_in...
grid_check done
Done with read_grid_in...
Starting from beginning of simulation...
Starting from beginning of simulation...
Starting from beginning of simulation...
Starting from beginning of simulation...
/Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/simsize.h5 input dimensions: 128 128 24
Target (output) grid structure dimensions: 128 128 24
File handle opened for reading initial conditions: /Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/initial_conditions.h5
Initial conditions read from file: /Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/inputs/initial_conditions.h5
Sent ICs to workers in 6.184E-03 seconds.
Initial conditions (root):
------------------------
Min/max input density: 3.00E+00 6.98E+11
Min/max input velocity: 0.00E+00 0.00E+00
Min/max input temperature: 1.97E+02 1.07E+03
Min/max input electric potential: 0.00E+00 0.00E+00
Min/max input electric potential (full grid): 0.00E+00 0.00E+00
NOTE: Neutral background to be taken from empirical models...
NOTE: Neutral background to be taken from empirical models...
NOTE: Neutral background to be taken from empirical models...
Recomputed initial dist. fields via pot2perpfield:
gemini 0.0000000000000000 0.0000000000000000
gemini 0.0000000000000000 0.0000000000000000
gemini 0.0000000000000000 0.0000000000000000
gemini 0.0000000000000000 0.0000000000000000
Priming inputdata
NOTE: Neutral background to be taken from empirical models...
Setting Lagrangian-ness
Recomputed initial fields including background (BGfieldLagrangian):
0.0000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000
Recomputed initial drifts:
-36.101351563943545 28.842169633203074
-39.717429173066911 38.766254508921456
Starting main loop
Updating empirical background atmosphere...
Updating empirical background atmosphere...
Updating empirical background atmosphere...
Updating empirical background atmosphere...
System sizes according to Phiall: 128 128 24
HDF5 Output file name: /Users/zettergm/simulations/sdcard/simulations_ESF/Eq/Eq_ground_HWM_testing/20160303_04500.000000.h5
INFO: Input file has selected full output or milestones, large files may result!
Writing topside boundary conditions for non-inverted-type grid...
Throttling dt to: 2.000E-06
Throttling dt to: 4.000E-06
Throttling dt to: 8.000E-06
Throttling dt to: 16.000E-06
Throttling dt to: 32.000E-06
Throttling dt to: 64.000E-06
Throttling dt to: 128.000E-06
Throttling dt to: 256.000E-06
Current time 2016-03-03 4500.000511; dt=0.000256
Throttling dt to: 512.000E-06
Throttling dt to: 1.024E-03
Throttling dt to: 2.048E-03
Throttling dt to: 4.096E-03
Throttling dt to: 8.192E-03
Throttling dt to: 16.384E-03
Throttling dt to: 32.768E-03
Throttling dt to: 65.536E-03
Current time 2016-03-03 4500.222085; dt=0.039932
Throttling dt to: 404.462E-06
Current time 2016-03-03 4500.388267; dt=0.000225
Throttling dt to: 449.292E-06
Throttling dt to: 898.585E-06
Throttling dt to: 1.797E-03
Throttling dt to: 3.594E-03
Throttling dt to: 7.189E-03
Throttling dt to: 14.377E-03
Throttling dt to: 28.755E-03
Throttling dt to: 6.404E-03
Current time 2016-03-03 4500.472229; dt=0.006404
Throttling dt to: 12.808E-03
Throttling dt to: 25.616E-03
Throttling dt to: 9.864E-06
What happened?
clang and GNU C and C++ compilers with gfortran 16.1, MacOS 26.6 on M-series: Builds and runs fine without HWM14, but when built with HWM14 it will crash after a minute or so running a simulation with a dipole grid. Seems to work fine with gcc 15.2 (M-series) and works fine on MacOS intel (any compiler combo). Seems to work on our HPC with gcc 13.2, as well. Test data here: https://www.dropbox.com/scl/fo/ssmyqo0l0d6530wybvxej/AGn7tc-NMm4Nfw9JM7UFrJI?rlkey=wx99l0dwj6bzrtm1titnb4m7h&dl=0
Relevant log output