Disable MiniSat simplifier after first incremental solve

[tautschnig]

MiniSat's SimpSolver runs variable elimination (eliminate()) before every solveLimited/solve call by default. When used incrementally in the all-properties verification loop, this repeated elimination can degrade the solver's ability to prove UNSAT, causing the CDCL search to hang indefinitely.

Fix by passing do_simp=false on all calls after the first. The first call still benefits from the full simplification pass, but subsequent incremental calls skip it, preventing the problematic variable elimination between iterations.

This fixes a hang when running:
cbmc --arrays-uf-always --unwind 1 --32 Array_UF23/main.c

• Each commit message has a non-empty body, explaining why the change was made.
• n/a Methods or procedures I have added are documented, following the guidelines provided in CODING_STANDARD.md.
• n/a The feature or user visible behaviour I have added or modified has been documented in the User Guide in doc/cprover-manual/
• Regression or unit tests are included, or existing tests cover the modified code (in this case I have detailed which ones those are in the commit message).
• n/a My commit message includes data points confirming performance improvements (if claimed).
• My PR is restricted to a single feature or bugfix.
• n/a White-space or formatting changes outside the feature-related changed lines are in commits of their own.

[Copilot]

Pull request overview

This PR fixes a hang in CBMC's incremental SAT solving when using MiniSat's SimpSolver. The issue occurs because MiniSat's simplifier runs variable elimination (eliminate()) before every solveLimited/solve call, which can degrade the solver's ability to prove UNSAT during the all-properties verification loop. The fix passes do_simp=false on all calls after the first, allowing the initial full simplification pass but skipping it on subsequent incremental calls.

Changes:

• Added a solver_was_called flag to satcheck_minisat2_baset to track whether the solver has been called before, and use if constexpr to pass do_simp=false to SimpSolver::solveLimited/solve on subsequent calls
• Reordered includes to follow the .clang-format priority ordering and added <type_traits> for std::is_same_v
• Added a regression test (Array_UF23) that reproduces the hang scenario with --arrays-uf-always --unwind 1 --32

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
src/solvers/sat/satcheck_minisat2.h	Added solver_was_called boolean member with in-class default initializer
src/solvers/sat/satcheck_minisat2.cpp	Used if constexpr to conditionally disable SimpSolver's simplifier after first solve, reordered includes
regression/cbmc/Array_UF23/main.c	New regression test C source with VLA array stores and an assertion
regression/cbmc/Array_UF23/test.desc	Test descriptor for the regression test

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.01%. Comparing base (20a9052) to head (c8ca115).
⚠️ Report is 1 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #8851      +/-   ##
===========================================
- Coverage    80.01%   80.01%   -0.01%     
===========================================
  Files         1700     1700              
  Lines       188338   188347       +9     
  Branches        73       73              
===========================================
+ Hits        150702   150707       +5     
- Misses       37636    37640       +4     

☔ View full report in Codecov by Sentry.
📢 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.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[kroening]

Is this a genuine issue in MiniSat, or are we using it wrong?

[tautschnig]

It's a known limitation of MiniSat's SimpSolver when used incrementally. The SimpSolver was designed primarily for single-shot solving — its solve/solveLimited methods run eliminate() before every call by default. While frozen variables are protected from elimination, the variable elimination and subsumption passes restructure the clause database between incremental calls, which can make subsequent UNSAT proofs much harder for the CDCL search.

MiniSat's API does provide the do_simp parameter precisely for this use case (the parameter defaults to true). We just weren't using it. Other incremental SAT solver interfaces (CaDiCaL, etc.) don't have this issue because they don't run preprocessing between incremental calls unless explicitly asked to.