[Issue Backlog] Backlog Analysis - March 11, 2026

[github-actions[bot]]

Executive Summary

• Total open issues analyzed: 139
• Issues recommended for closure: 0
• Potential duplicates / merge candidates: 2 clusters (5 issues total)
• Critical soundness issues: 3 (actively assigned)
• Performance regressions: 4 major issues in v4.15-4.16
• Issues needing maintainer attention: 8

---

Critical Issues Requiring Immediate Attention

Soundness Bugs (Highest Priority)

These issues represent incorrect answers from Z3, which is the most serious category of bugs:

Issue	Title	Status	Priority
#8023	Soundness issue in string logic	8 comments, active	CRITICAL
#8185	Incorrect model in mixed FP/Real logic triggered by string constraint	Under investigation	CRITICAL
#8194	Incorrect model with :produce-proofs true involving String Theory and UF	Under investigation	CRITICAL

Notes:

• All three involve string theory, suggesting a systemic issue in the string solver
• Incorrect model in mixed FP/Real logic triggered by extra string constraint #8185 and Incorrect model with :produce-proofs true involving String Theory and UF #8194 only fail with :produce-proofs true, indicating proof generation may be corrupting models
• All have SMT-LIB2 test cases attached and are reproducible

---

Performance Regressions in v4.15-4.16

Multiple users have reported severe performance degradation in recent releases:

#8893 - String solver performance regression in 4.16

• Impact: 24x slowdown (2.1s → 48.8s) on QF_SLIA benchmark
• Versions affected: 4.16.0 vs 4.15.2
• Component: String solver (regex with length constraints)
• Status: Assigned to @NikolajBjorner, recently opened (Mar 8)

#8890 - z3-solver==4.16.0 finds less solutions for BNH problem

• Impact: Optimization solver finds fewer solutions in 4.16.0 vs 4.8.12
• Versions affected: 4.16.0 regression vs 4.8.12
• Component: Optimization (vZ extension)
• Status: Assigned to @NikolajBjorner and @levnach, opened Mar 8

#8740 - solver.check() returns 'unknown' instead of 'sat'

• Impact: Satisfiable problems now return 'unknown'
• Versions affected: 4.15.8+ (broken), 4.15.4 and earlier (working)
• Component: Integer arithmetic solver
• Status: Regression confirmed by cvc5, opened Feb 23

#8096 - Performance degradation on QF_SLIA with fixed-length strings

• Impact: Exponential degradation (0.16s at x=20 → 2002s at x=60)
• Comparison: cvc5 solves all in 0.02s
• Component: String solver (fixed-length + prefix/suffix)
• Status: Opened Dec 19, v4.15.5

---

Potential Duplicates / Merge Candidates

Cluster 1: JavaScript/Node.js WASM thread non-termination

Issue	Title	Comments	Reactions
#6512	NodeJs process won't stop after using 'solve' or 'solver.check()'	2	👍 8
#6701	Javascript & Typescript API Runs Forever	?	?
#7070	[NodeJS] Z3 doesn't clean up threads/workers after finishing	?	?

Analysis: All three describe the same problem — the z3-solver npm package leaves WASM worker threads alive after solving, preventing Node.js process exit. A comment was already added to #6512 by the previous run linking these.

Recommendation: Keep #6512 as canonical (most reactions), close others as duplicates once fixed.

Affected code: src/api/js/ (JavaScript/WASM bindings)

Cluster 2: arith.solver=6 performance regressions

Multiple issues report performance problems with the newer arithmetic solver:

• [Consolidated] Performance regressions of arith.solver=6 vs. arith.solver=2 #4934, Performance regression from arith.solver=2 to arith.solver=6 #5614, regression with arith.solver=6 #7502 (from cache)

Recommendation: Investigate whether these are the same root cause and merge tracking.

---

Issues Needing More Information

Issue	Title	Age	Missing Information
#8890	BNH optimization finds fewer solutions	3 days	Need to confirm if this is expected behavior vs bug
#6768	Issues importing z3-solver using Vite	2+ years	Need specific error messages and build config
#7958	Z3 4.15.3 is not on opam	1+ year	Is this still an issue? Latest version?

---

Long-Standing Issues Deserving Attention

These issues are old but still relevant:

Issue	Title	Age (days)	Last Updated	Notes
#6512	NodeJs process won't stop	~1130	Feb 24, 2026	8 reactions, affects JS users
#7800	Memory/compute regression 4.14→4.15	~630	Mar 3, 2026	Performance regression
#6788	Finite sets	~1095	Feb 24, 2026	Feature request, has workarounds

---

Recommendations for Maintainers

1. Prioritize soundness bugs (Soundness issue in string logic #8023, Incorrect model in mixed FP/Real logic triggered by extra string constraint #8185, Incorrect model with :produce-proofs true involving String Theory and UF #8194) — all involve string theory with proofs
2. Investigate 4.16.0 regressions — multiple performance issues reported
3. Fix JavaScript thread cleanup (NodeJs process won't stop after using 'solve' or 'solver.check()' #6512, Javascript & Typescript API Runs Forever #6701, [NodeJS] Z3 doesn't seem to clean up threads/workers after finishing its job #7070) — high user impact, clear bug
4. Consider bisecting v4.15.4 → v4.15.8 to identify Z3 Regression: solver.check() returns 'unknown' instead of 'sat' #8740 regression
5. Profile string solver for performance issues (String solver performance regression in 4.16 #8893, Performance degradation in Z3 on QF_SLIA instances with fixed-length strings and prefix/suffix constraints #8096)

---

Automated by Issue Backlog Processor
Last run: February 24, 2026 | Current run: March 11, 2026 (14 days since last analysis)
Total issues: 139 open | Categories: 11 soundness, 28 performance, 12 API/bindings, 27 bugs

AI generated by Issue Backlog Processor · history

• expires on Mar 18, 2026, 12:19 AM UTC

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-03-18T00:19:11.608Z.

Closed by Workflow