fix: reject aliased cell-list layouts

[galjos]

Summary

• Add a runtime guard that rejects cell-list configurations where periodic neighbor offsets can alias the same cell.
• Add a regression test for the invalid 2-cell layout.
• Update the cell-list setup fixture to use a valid box/cutoff/cell-count combination.

Root cause

The half-neighbor cell-list stencil assumes enough cells per dimension to represent each neighbor offset without wrapping onto already represented cells. If nCells < 2 * nNeighbourCells + 1, periodic wrapping can alias offsets and make the neighbor list mathematically invalid.

Validation

• cmake --build build-stochastic-rescale-ci --target testCelllist testCelllistSetup --parallel 4
• ctest --test-dir build-stochastic-rescale-ci -R '^(testCelllist|testCelllistSetup)$' --output-on-failure

Reference

• LAMMPS developer docs on half neighbor-list bin stencils: https://docs.lammps.org/Developer_par_neigh.html

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.24%. Comparing base (d0069f7) to head (eddea27).

Additional details and impacted files

@@           Coverage Diff           @@
##              dev     #274   +/-   ##
=======================================
  Coverage   89.23%   89.24%           
=======================================
  Files         286      286           
  Lines       11025    11033    +8     
  Branches     3432     3437    +5     
=======================================
+ Hits         9838     9846    +8     
  Misses       1147     1147           
  Partials       40       40           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[pq-perf-bot]

⚡ Performance (instruction count) — ✅ no regressions

per-benchmark breakdown

benchmark	base Ir	PR Ir	Δ
bondedForces	41.46M	41.46M	-0.00%	✅
boxTransforms	10.67M	10.67M	-0.00%	✅
constraints	11.38M	11.38M	-0.00%	✅
coulombKernel	5.80M	5.80M	+0.00%	✅
forceKernel	15.33M	15.33M	-0.00%	✅
integrator	32.17M	32.17M	-0.00%	✅
kinetics	9.53M	9.53M	-0.00%	✅
linearAlgebra	2.08M	2.08M	+0.00%	✅
nonCoulombPairs	5.48M	5.48M	-0.00%	✅
shiftVector	5.81M	5.81M	-0.00%	✅
virial	11.63M	11.63M	-0.00%	✅

_{Deterministic callgrind instruction counts vs the base branch; gated at ±2%. Not wall-clock.}

Requested change addressed in eddea27; review re-requested.