Fix Periodic Errors Near IB Surface

[danieljvickers]

Contribution Policy

We get an error when we try to look for ghost points that are on the other side of processor boundaries. We had a check to prevent looking inside walls, but it also applied to periodic conditions that should be open. This fixes the issue. While I was in there, I grouped some ugly code into a single macro.

[github-actions]

Lines of Code

File	Lines	Diff
src/simulation/m_ibm.fpp	1397	-11

Directory	Lines	Diff
simulation	28054	-11
total	46992	-11

[github-actions]

Claude Code Review

Head SHA: 97fc643

Files changed:

• 1
• src/simulation/m_ibm.fpp

Findings:

• src/simulation/m_ibm.fpp: In the new #:for loop, the domain-boundary flags are written to ghost_points_in(local_idx)%DB(${ID}$), but every other field for this same local_idx in this block (e.g. ghost_points(local_idx)%z_periodicity, ghost_points(local_idx)%slip in the unchanged lines immediately above) is written to ghost_points, not ghost_points_in. If ghost_points_in is a distinct array (rather than a typo for ghost_points), the %DB component of ghost_points(local_idx) is left unset/uninitialized by this code path, silently breaking the domain-boundary logic used elsewhere for ghost-point boundary treatment.