meep output step-functions (e.g., mp.output_efield) are NOT MPI-safe when called frequently?

[aa16oaslak]

Hello!

I am currently trying to use the output step functions in MEEP to write the field files in the h5py format using the pre-defined output step functions in MEEP:

mp.after_time(t0, mp.at_every(dt, mp.output_hfield))

When I run this in my local device for a small system, it's running flawlessly! But when I try to use it in a HPC facility using mpirun, its showing me the following segmentation-fault error:

Caught signal 11 (Segmentation fault: address not mapped to object at address 0x100000001)

If I comment out wherever the output step functions are used, I don't run into any issues!

Is it happening because of the output step functions are called multiple times in a short timespan and that's causing some I/O h5py dispute when rank not equals to zero?? (Note: Here rank is the MPI process identifier (an integer) that uniquely identifies each parallel process when running with MPI. Rank = 0 is the master work)

Could you please help me out with this issue?

P.S: I can write my own functions to extract the fields which are MPI-safe- But just wanted to confirm if I can still use the output step functions!