[platelett]

Fixes #1908.

A case that sets chem_params%diffusion = 'T' without chemistry = 'T' passes ./mfc.sh validate, then segfaults on the first time step. This PR rejects that case in the validator. It also adds the chemistry guard that #1908 asks for to the matching deallocation. No working case changes and no goldens move.

Why it crashes

s_initialize_rhs_module allocates the arrays the diffusion path writes only under if (chemistry):

• the species slots of flux_src_n;
• flux_src_n(E), when nothing else allocates it.

Several sites test chem_params%diffusion on its own, so they still run in a non-reacting case:

• m_riemann_state.fpp: if (chem_params%diffusion) zeroes flux_src_vf(E:species%end).
• m_rhs.fpp: s_compute_additional_physics_rhs runs on viscous .or. surface_tension .or. chem_params%diffusion .or. heat_conduction. Inside it, each direction's if (chem_params%diffusion) block reads flux_src_n(E).

For example, examples/1D_sodshocktube with that one line added:

 Simulating a regular 399x0x0 case on 1 rank(s) on CPUs.
 [  0%] step 1 of 21 (t_step 0) ...
Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Change

• The rule. check_chemistry gets one rule, next to the other chemistry checks:

  self.prohibit(diffusion and not chemistry, "chem_params%diffusion requires chemistry = T")

  It follows the existing "parameter set without its feature" checks, such as sigma is set but surface_tension is not enabled. Guarding the six-plus Fortran sites instead would spread one condition over several files.
• The deallocation. s_finalize_rhs_module gets the chemistry .and. guard its allocation already has, as flux_src_n energy-slot deallocation lacks the chemistry guard its allocation has #1908 asks. With the new rule it is unreachable, but allocate and deallocate are now symmetric.
• Tests. Two unit tests in toolchain/mfc/test_case_validator.py: one for the rejection, and one showing that diffusion with chemistry is still accepted.

Side effects

• A case with chem_params%diffusion = 'T' and no chemistry is now rejected at validation. It used to crash on step 1.
• All 20 examples that set chem_params%diffusion also set chemistry, so they still validate.
• No solver results change.

Verification

Check	Result
Sod shock tube + chem_params%diffusion = 'T', master (GNU 13.4 + Open MPI, CPU)	passes validation, then SIGSEGV on step 1
Same case, this PR	Case parameter constraint violations: 1. chem_params%diffusion requires chemistry = T
pytest toolchain/mfc/test_case_validator.py	96 passed, including the 2 new tests
All 15 chemistry regression tests, including the diffusion cases 8E56ACC1, 4587725A and 41AD0140 (GNU 13.4 + Open MPI)	15 passed against the existing goldens
./mfc.sh precheck, which validates every example case	passes
Full regression suite, GNU 13.3 + Open MPI 4.1.6, CPU (on a branch that also includes our other fixes)	725 passed, 0 failed (39 convergence cases skipped by default)

Contribution Policy

All contributions are expected to demonstrate:

• A clear understanding of the codebase
• Alignment with product direction
• Thoughtful reasoning behind changes
• Evidence of real-world usage or hands-on experience with the problem

If these expectations are not met, we would prefer to implement the changes ourselves rather than spend time reviewing low-effort submissions.

Acknowledgement

• I confirm this PR meets the above expectations and reflects my own understanding and real-world context.

---

This PR is from Diablo (Tsinghua University) participating in the SC26 Student Cluster Competition.

[github-actions]

Claude Code Review

Head SHA: 26a3d39

Files changed:

• 3
• src/simulation/m_rhs.fpp
• toolchain/mfc/case_validator.py
• toolchain/mfc/test_case_validator.py

Findings:
No issues found. Checked in step 4: src/simulation/m_rhs.fpp (the new chemistry .and. condition in the deallocation matches the allocation at line 261, which sits inside if (chemistry)), toolchain/mfc/case_validator.py, and toolchain/mfc/test_case_validator.py.

[codecov]

Codecov Report

❌ Patch coverage is 0% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 61.95%. Comparing base (ee99151) to head (26a3d39).

Files with missing lines	Patch %	Lines
src/simulation/m_rhs.fpp	0.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #2108   +/-   ##
=======================================
  Coverage   61.95%   61.95%           
=======================================
  Files          86       86           
  Lines       22906    22906           
  Branches     3378     3378           
=======================================
  Hits        14192    14192           
  Misses       6206     6206           
  Partials     2508     2508           

☔ 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.