Fix integer overflow

[wilfonba]

Description

Fixes integer overflow for simulations with more than 2^31 cells.

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Something else

Scope

• This PR comprises a set of related changes with a common goal

If you cannot check the above box, please split your PR into multiple PRs that each have a common goal.

[sbryngelson]

Something about some of these changes doesn't quite seem right to me (I could be wrong). For example, if you write int(p), this shouldn't do anything if p is already int. Likewise for other quantities.

[wilfonba]

int(p) casts an 8-byte integer p to a 4-byte integer. The MPI routines only accept 4-byte integers.

[sbryngelson]

int(p) casts an 8-byte integer p to a 4-byte integer. The MPI routines only accept 4-byte integers.

Got it - though I see a couple other ones, like for s_compute_finite_difference_coefficients.

[wilfonba]

I didn't change the integer types in simulation because the global cell count is never computed in simulation. s_compute_finite_difference_coefficients is used in both simulation and post_process and expects a 4-byte integer.

[sbryngelson]

I didn't change the integer types in simulation because the global cell count is never computed in simulation. s_compute_finite_difference_coefficients is used in both simulation and post_process and expects a 4-byte integer.

Does it not automatically up (or down) cast the input precision? That would make this much easier to maintain, otherwise every time one of the "special" variables (e.g., p) is used in both simulation and another code (pre/post process), the developer needs to know that they need to do int(p).

[wilfonba]

It does not automatically cast the inputs. I'll just make simulation use 8-byte integers for m, n, and p, then everything will be the same. The only reason I didn't was because adding all the casting in the MPI routines would be tedious.

[sbryngelson]

It does not automatically cast the inputs. I'll just make simulation use 8-byte integers for m, n, and p, then everything will be the same. The only reason I didn't was because adding all the casting in the MPI routines would be tedious.

How is this different than what you already did?

[wilfonba]

I'll just explain it at our meeting this afternoon. It'll be easier that way.