mirrors/z3
3
0
Fork 0
mirror of https://github.com/Z3Prover/z3 synced 2026-07-16 03:55:42 +00:00
Code Activity
z3/src/smt
History
Lev Nachmanson a5454ec375
[snapshot-regression-fix] smt_parallel: report unknown on theory-incomplete cubes instead of hanging (#9999) 
## Summary

Fixes a hang (wall-clock timeout) in the native parallel SMT solver when
a cube is incomplete for a reason that cannot change. Originating
discussion: https://github.com/Z3Prover/bench/discussions/2746

Benchmark: `iss-3707/bug-1.smt2` (`QF_NRA`, runs with
`parallel.enable=true`).

## Divergence

The recorded oracle vs. current z3 (`z3 -T:20`):

```diff
-(incomplete (theory difference-logic))
-unknown
+timeout
```

z3 should terminate with `unknown` (incomplete theory) but instead spins
until the 20s timeout.

## Root cause

In `src/smt/smt_parallel.cpp` the per-cube worker handled an `l_undef`
cube by unconditionally calling `update_max_thread_conflicts()` and
re-splitting/re-checking. That only helps when the cube was abandoned at
the per-cube conflict limit (`max-conflicts-reached`). When the cube is
incomplete for a permanent reason (incomplete theory, quantifiers,
resource limits), the verdict never changes, so the worker re-checks the
same cube forever. The `batch_manager` had no `unknown` terminal state,
so `get_result()` could only end as sat/unsat/exception — there was no
way to settle on `unknown`, hence the hang. This is the `smt_parallel`
analogue of the `parallel_tactical.cpp` regression fixed earlier.

## Fix

Minimal, mirroring the tactic-side fix:
- add an `is_unknown` batch-manager state + `m_reason_unknown`;
- a worker reporting `l_undef` whose `last_failure` is not
`max-conflicts-reached` calls `set_unknown(reason)` and stops
re-splitting;
- `set_sat`/`set_unsat` may still override `is_unknown` so a definitive
answer wins;
- `get_result()` maps `is_unknown -> l_undef` and the reason propagates
to the parent context.

## Validation

Rebuilt z3 (`make -C build -j16`) and re-ran the benchmark 5× with
`-T:20`. Every run finished in well under the timeout with output
matching the oracle byte-for-byte:

```
(incomplete (theory difference-logic))
unknown
```

Created as a **draft** for human review.




> Generated by [Fix a Z3 snapshot-regression
divergence](https://github.com/Z3Prover/bench/actions/runs/28358375255)
· 553.9 AIC · ⌖ 27.2 AIC · ⊞ 9K ·
[◷](https://github.com/search?q=repo%3AZ3Prover%2Fz3+%22gh-aw-workflow-id%3A+snapshot-regression-fixer%22&type=pullrequests)

<!-- gh-aw-agentic-workflow: Fix a Z3 snapshot-regression divergence,
engine: copilot, version: 1.0.63, model: claude-opus-4.8, id:
28358375255, workflow_id: snapshot-regression-fixer, run:
https://github.com/Z3Prover/bench/actions/runs/28358375255 -->

<!-- gh-aw-workflow-id: snapshot-regression-fixer -->
<!-- gh-aw-workflow-call-id: Z3Prover/bench/snapshot-regression-fixer
-->

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
2026-06-29 06:55:24 -07:00
..
proto_model Remove copies (#8583) 2026-02-18 21:02:22 -08:00
tactic Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
arith_eq_adapter.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
arith_eq_adapter.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
arith_eq_solver.cpp Fix static analysis issues: null dereferences, unsafe casts, branch clones, uninitialized members (#9424) 2026-04-29 13:37:11 -07:00
arith_eq_solver.h Remove redundant non-virtual destructors with = default (#8462) 2026-02-18 20:58:01 -08:00
CMakeLists.txt remove lattice component 2026-02-19 15:49:45 -08:00
database.h Make sure all headers do #pragma once. (#6188) 2022-07-23 10:41:14 -07:00
database.smt
diff_logic.h Replace fall-through comments with Z3_fallthrough macro (#8219) 2026-02-18 20:57:31 -08:00
dyn_ack.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
dyn_ack.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
expr_context_simplifier.cpp Fix static analysis issues: null dereferences, unsafe casts, branch clones, uninitialized members (#9424) 2026-04-29 13:37:11 -07:00
expr_context_simplifier.h move smt params to params directory, update release.yml 2025-06-09 10:47:22 -07:00
fingerprints.cpp Remove unused defined_names artifacts and simplify fingerprint_set::contains (#9702) 2026-06-03 08:16:46 -07:00
fingerprints.h remove side definitions 2026-06-02 21:43:55 -07:00
mam.cpp Update used_enodes properly (#9695) 2026-06-03 13:36:37 -07:00
mam.h Use = default for virtual constructors. 2022-08-05 18:11:46 +03:00
old_interval.cpp Centralize and document TRACE tags using X-macros (#7657) 2025-05-28 14:31:25 +01:00
old_interval.h Use noexcept more. (#7058) 2023-12-16 12:14:53 +00:00
qi_queue.cpp remove side definitions 2026-06-02 21:43:55 -07:00
qi_queue.h add an option to register callback on quantifier instantiation 2025-08-06 21:11:55 -07:00
seq_axioms.cpp add options for logging learned lemmas and theory axioms 2022-08-08 11:18:56 +03:00
seq_axioms.h na 2022-01-01 17:59:31 -08:00
seq_eq_solver.cpp tweaks to string solver 2026-06-28 17:16:52 -07:00
seq_ne_solver.cpp Adopt C++17 structured bindings for map/pair iteration (#8159) 2026-02-18 20:57:09 -08:00
seq_offset_eq.cpp Refactor seq_offset_eq::find to use std::optional (#8300) 2026-02-18 20:57:53 -08:00
seq_offset_eq.h Refactor seq_offset_eq::find to use std::optional (#8300) 2026-02-18 20:57:53 -08:00
seq_regex.cpp Derive with ranges (#9965) 2026-06-26 08:44:13 -06:00
seq_regex.h Derive with ranges (#9965) 2026-06-26 08:44:13 -06:00
smt2_extra_cmds.cpp
smt2_extra_cmds.h booyah 2020-07-04 15:56:30 -07:00
smt_almost_cg_table.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
smt_almost_cg_table.h booyah 2020-07-04 15:56:30 -07:00
smt_arith_value.cpp remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
smt_arith_value.h Merge branch 'master' into finite-sets 2026-02-18 21:11:44 -08:00
smt_b_justification.h remove unneeded constructors (last round) 2020-07-12 17:41:57 +01:00
smt_bool_var_data.h booyah 2020-07-04 15:56:30 -07:00
smt_case_split_queue.cpp Update generation number of already-internalized enodes (#9628) 2026-05-26 15:08:11 -07:00
smt_case_split_queue.h move smt params to params directory, update release.yml 2025-06-09 10:47:22 -07:00
smt_cg_table.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_cg_table.h booyah 2020-07-04 15:56:30 -07:00
smt_checker.cpp control recursion depth for check function 2026-05-27 14:29:53 -07:00
smt_checker.h control recursion depth for check function 2026-05-27 14:29:53 -07:00
smt_clause.cpp Refactor mk_and/mk_or call sites to use vector overloads (#8286) 2026-02-18 20:57:52 -08:00
smt_clause.h Use = default for virtual constructors. 2022-08-05 18:11:46 +03:00
smt_clause_proof.cpp fixup proof log annotations of rules 2025-10-19 10:04:18 +02:00
smt_clause_proof.h fixup proof log annotations of rules 2025-10-19 10:04:18 +02:00
smt_conflict_resolution.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
smt_conflict_resolution.h Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_consequences.cpp fix crash 2026-02-18 21:02:17 -08:00
smt_context.cpp Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
smt_context.h Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
smt_context_inv.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_context_pp.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
smt_context_stat.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_enode.cpp disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
smt_enode.h disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
smt_eq_justification.h booyah 2020-07-04 15:56:30 -07:00
smt_failure.h fix crash reported by Nikhil on F* due to unhandled exception while using the rewriter during search 2025-01-28 16:27:28 -08:00
smt_farkas_util.cpp Centralize and document TRACE tags using X-macros (#7657) 2025-05-28 14:31:25 +01:00
smt_farkas_util.h booyah 2020-07-04 15:56:30 -07:00
smt_for_each_relevant_expr.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_for_each_relevant_expr.h Use = default for virtual constructors. 2022-08-05 18:11:46 +03:00
smt_implied_equalities.cpp Refactor mk_and/mk_or call sites to use vector overloads (#8286) 2026-02-18 20:57:52 -08:00
smt_implied_equalities.h booyah 2020-07-04 15:56:30 -07:00
smt_induction.cpp.disabled Centralize and document TRACE tags using X-macros (#7657) 2025-05-28 14:31:25 +01:00
smt_induction.h.disabled remove dependencies on stale component 2021-08-16 17:52:36 -07:00
smt_internalizer.cpp disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
smt_justification.cpp fix build 2026-02-18 20:58:00 -08:00
smt_justification.h add options for logging learned lemmas and theory axioms 2022-08-08 11:18:56 +03:00
smt_kernel.cpp Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
smt_kernel.h Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
smt_literal.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_literal.h consolidate literals 2021-05-20 12:58:27 -07:00
smt_lookahead.cpp disable pre-processing during cubing 2025-07-31 20:58:58 -07:00
smt_lookahead.h Revert "Parallel solving (#7775)" (#7777) 2025-08-14 18:16:35 -07:00
smt_model_checker.cpp fix build warnings 2026-06-22 18:20:23 -07:00
smt_model_checker.h Remove unused defined_names artifacts and simplify fingerprint_set::contains (#9702) 2026-06-03 08:16:46 -07:00
smt_model_finder.cpp disregard skolems in instantiation set? 2026-06-28 12:04:56 -07:00
smt_model_finder.h Term enumeration (#9908) 2026-06-20 18:14:44 -06:00
smt_model_generator.cpp disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
smt_model_generator.h fix #7584 2025-03-15 13:33:08 -07:00
smt_parallel.cpp [snapshot-regression-fix] smt_parallel: report unknown on theory-incomplete cubes instead of hanging (#9999) 2026-06-29 06:55:24 -07:00
smt_parallel.h [snapshot-regression-fix] smt_parallel: report unknown on theory-incomplete cubes instead of hanging (#9999) 2026-06-29 06:55:24 -07:00
smt_quantifier.cpp remove side definitions 2026-06-02 21:43:55 -07:00
smt_quantifier.h remove side definitions 2026-06-02 21:43:55 -07:00
smt_quantifier_instances.h booyah 2020-07-04 15:56:30 -07:00
smt_quick_checker.cpp remove side definitions 2026-06-02 21:43:55 -07:00
smt_quick_checker.h booyah 2020-07-04 15:56:30 -07:00
smt_relevancy.cpp Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
smt_relevancy.h allow for internalize implies 2025-07-31 20:48:15 -07:00
smt_setup.cpp fix indentation 2026-02-18 20:57:01 -08:00
smt_setup.h Complete theory_finite_set.h header and implement finite set theory solver (#7976) 2025-10-15 14:55:08 +02:00
smt_solver.cpp Parallel tactic (#9824) (#9825) 2026-06-26 10:36:15 -06:00
smt_solver.h booyah 2020-07-04 15:56:30 -07:00
smt_statistics.cpp
smt_statistics.h Add global backbones to parallel architecture (#9343) 2026-04-20 18:22:07 +02:00
smt_theory.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
smt_theory.h prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
smt_types.h signed 2021-05-21 15:51:27 -07:00
smt_value_sort.cpp refactor get_sort 2021-02-02 04:45:54 -08:00
smt_value_sort.h booyah 2020-07-04 15:56:30 -07:00
spanning_tree.h booyah 2020-07-04 15:56:30 -07:00
spanning_tree_base.h booyah 2020-07-04 15:56:30 -07:00
spanning_tree_def.h Centralize and document TRACE tags using X-macros (#7657) 2025-05-28 14:31:25 +01:00
theory_arith.cpp
theory_arith.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_arith_aux.h prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_arith_core.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_arith_def.h booyah 2020-07-04 15:56:30 -07:00
theory_arith_eq.h Centralize and document TRACE tags using X-macros (#7657) 2025-05-28 14:31:25 +01:00
theory_arith_int.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_arith_inv.h Standardize for-loop increments to prefix form (++i) (#8199) 2026-02-18 20:57:29 -08:00
theory_arith_nl.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_arith_pp.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_array.cpp fix rewriting of complemnt of ranges 2026-06-20 11:08:00 -07:00
theory_array.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_array_base.cpp fix rewriting of complemnt of ranges 2026-06-20 11:08:00 -07:00
theory_array_base.h disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
theory_array_full.cpp disable test in tptp, move to native lambdas 2026-06-02 10:38:51 -07:00
theory_array_full.h move closure conversion to solver internalization 2026-05-30 18:41:37 -07:00
theory_bv.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_bv.h prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_char.cpp Refactor theory_char::get_char_value to use std::optional (#8302) 2026-02-18 20:57:53 -08:00
theory_char.h Refactor theory_char::get_char_value to use std::optional (#8302) 2026-02-18 20:57:53 -08:00
theory_datatype.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_datatype.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_dense_diff_logic.cpp
theory_dense_diff_logic.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_dense_diff_logic_def.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_diff_logic.cpp
theory_diff_logic.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_diff_logic_def.h prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_dl.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_dl.h booyah 2020-07-04 15:56:30 -07:00
theory_dummy.cpp remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_dummy.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_finite_set.cpp fix build warnings 2026-06-22 18:20:23 -07:00
theory_finite_set.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_finite_set_size.cpp build warnings 2026-05-29 10:17:46 -07:00
theory_finite_set_size.h add functions that create unique sets for model construction based on solving cardinality constraints 2025-12-29 11:57:48 -08:00
theory_fpa.cpp Fix FPA incremental soundness for fp.to_* terms (#9022) (#9787) 2026-06-09 10:22:03 -07:00
theory_fpa.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_intblast.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_intblast.h Fix intblast ubv_to_int bug: add bv2int axioms for compound expressions 2026-02-25 00:46:13 +00:00
theory_lra.cpp Lcube (#9858) 2026-06-14 16:25:21 -07:00
theory_lra.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_opt.cpp
theory_opt.h Add and fix a few general compiler warnings. (#5628) 2021-10-29 15:42:32 +02:00
theory_pb.cpp prepare for enodes over lambdas 2026-06-01 13:00:35 -07:00
theory_pb.h Remove redundant default constructors when they're the only constructor (#8461) 2026-02-18 20:58:01 -08:00
theory_polymorphism.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_recfun.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_recfun.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_seq.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_seq.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_seq_empty.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_sls.cpp remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_sls.h Eliminate unused private fields and local variables. (#9875) 2026-06-16 14:55:18 +01:00
theory_special_relations.cpp use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_special_relations.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_user_propagator.cpp remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_user_propagator.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
theory_utvpi.cpp Fix some typos. (#7115) 2024-02-07 23:06:43 -08:00
theory_utvpi.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_utvpi_def.h use expr based access to enodes to allow for storing first-class lambas 2026-05-30 15:13:08 -07:00
theory_wmaxsat.cpp Refactor mk_and/mk_or call sites to use vector overloads (#8286) 2026-02-18 20:57:52 -08:00
theory_wmaxsat.h remove set cardinality operators from array theory. Make final-check use priority levels 2026-02-18 20:56:51 -08:00
uses_theory.cpp AIX compat (#8113) 2026-02-18 20:57:04 -08:00
uses_theory.h booyah 2020-07-04 15:56:30 -07:00
watch_list.cpp fix #4856 2020-12-06 14:06:08 -08:00
watch_list.h enable ranges for bit-vectors 2020-08-13 10:53:37 -07:00