Fix random MPI crashes caused by array out-of-bounds access

[ChristopherMayes]

Fix remaining MPI array out-of-bounds crashes

Summary

Hardens the MPI build against intermittent segmentation faults (e.g. mpirun -n 8 ImpactZexe-mpi) caused by array out-of-bounds access. The crashes are non-deterministic because they depend on the runtime particle distribution across ranks.

Master already fixes the main offender via V2.7.6 (distparam = distparam0 in init_CompDom), which eliminates the uninitialized initial computational domain. This PR fixes three further, independent out-of-bounds bugs that V2.7.6 does not touch, plus a defensive guard in charge deposition. All are exposed by compiling with -fcheck=all. A reproducer is attached as error-test.zip.

Changes

1. src/Contrl/Input.f90 — obtype(0) when parsing comments

The lattice input parser reads lines in a loop, incrementing index i only for data lines (non-comment). However, the -99 end-of-lattice check on obtype(i) was outside the if(comst.ne."!") block, so when the first line is a comment, i is still 0 and obtype(0) is accessed.

Fix: Move the obtype(i).eq.-99 check inside the data-reading branch.

2. src/Contrl/Output.f90 — glbin(0) in 12 percentile search loops

The 90th/95th/99th percentile emittance calculations use cumulative histograms. Twelve do i = 1, nbin loops access glbin(i-1), which gives glbin(0) when i=1. The cumulative sum is built starting from i=2, so glbin(0) is never initialized and is out of bounds.

Fix: Start all 12 search loops at i = 2 instead of i = 1. This is safe because the interpolation formula uses glbin(i-1) and glbin(i), and glbin(1) already holds the raw count for bin 1.

3. src/Func/Ptclmger.f90 — zero-sized MPI buffer allocations

When a rank has no particles to exchange in a direction, jleft/jright/jdown/jup (or nsmall, derived from Nptlocal/numbuf) can be 0. This produces zero-sized allocations for temp1 and the left/right/up/down send buffers, after which MPI_RECV(temp1(1,1), ...) / MPI_SEND(left(1,1), ...) indexes element (1,1) out of bounds.

Fix: Allocate with max(..., 1) for the six affected buffers.

4. src/Appl/BeamBunch.f90 — defensive bounds check in deposit_BeamBunch

Defense-in-depth for the CIC charge deposition: skip any particle whose stencil (ix..ix+1, jx..jx+1, kx..kx+1) falls outside the local rho grid. Skipped particles are counted, reduced with MPI_ALLREDUCE, and reported once from rank 0 with a WARNING, so any charge loss is visible rather than causing a silent crash.

The out-of-bounds deposition that originally crashed the reproducer is fixed at the source by V2.7.6 (uninitialized initial domain). With master, this guard is not triggered by the reproducer; it remains as cheap insurance against any future decomposition producing out-of-bounds indices.

Testing

• Reproducer (error-test.zip): 10,000 particles through a 4-dipole chicane, extended diagnostics, mpirun -n 8.
• On top of master (V2.7.6): 5/5 runs complete cleanly with zero particles dropped (no WARNING).
• Debug build with -fcheck=all: no out-of-bounds reports.
• Regression: Example1, Example2, and Example3 still run correctly.

Files Changed

File	Change
src/Contrl/Input.f90	Move -99 check inside the data branch
src/Contrl/Output.f90	12 percentile loops: do i = 1 → do i = 2
src/Func/Ptclmger.f90	6 MPI buffer allocations guarded with max(..., 1)
src/Appl/BeamBunch.f90	Defensive deposition bounds check + skipped-particle reporting

Acknowledgement

Bug diagnosis and fixes developed with AI assistance (GitHub Copilot, Claude Opus 4.8). The original problem shows up in https://christophermayes.github.io/lume-impact/examples/z/elements/csr-zeuthen/.