test: reduce nodes in quorum-independent functional tests

[thepastaclaw]

Summary

• add a focused framework helper for mining single-member test quorums
• use one-member quorums in rpc_quorum.py and rpc_verifychainlock.py
• remove the unused masternode from feature_governance_objects.py

Coverage

The reduced topology is limited to tests where multi-member aggregation is not part of the behavior under test:

• rpc_quorum.py still verifies the address reported for every quorum member
• rpc_verifychainlock.py keeps separate non-masternode nodes, including the isolated-node ChainLock checks, and preserves all assertions
• feature_governance_objects.py only exercises local gobject RPC behavior on node 0

The normal quorum-mining helper waits for inter-member connections and probes that cannot exist for a one-member quorum, so this adds a narrow helper that mines until the new quorum commitment appears and then matures it.

Benchmarks

Local macOS ARM64 runs against binaries built from exact base 2bbf4a4617a:

• rpc_quorum.py: 19.06s -> 11.34s median over 5 runs (40.5% faster)
• rpc_verifychainlock.py: 23.22s -> 19.71s median over 5 runs (15.1% faster)
• feature_governance_objects.py: 10.64s -> 7.03s median over 3 runs (33.9% faster)
• combined medians: 52.9s -> 38.1s (~28% faster)

Validation

• test/lint/lint-python.py
• test/lint/lint-tests.py
• python3 -m py_compile on all changed Python files
• rpc_quorum.py
• rpc_verifychainlock.py
• feature_governance_objects.py
• independent pre-PR review: ship, zero actionable findings

[thepastaclaw]

✅ Review complete (commit 944e487)

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: a7229f1f-f905-44cd-a4a9-84240b5ffc8c

📥 Commits

Reviewing files that changed from the base of the PR and between 1821cef and 944e487.

📒 Files selected for processing (5)

• test/functional/feature_governance_objects.py
• test/functional/feature_llmq_singlenode.py
• test/functional/rpc_quorum.py
• test/functional/rpc_verifychainlock.py
• test/functional/test_framework/test_framework.py

🚧 Files skipped from review as they are similar to previous changes (4)

• test/functional/feature_governance_objects.py
• test/functional/rpc_verifychainlock.py
• test/functional/rpc_quorum.py
• test/functional/feature_llmq_singlenode.py

---

Walkthrough

Functional test configurations are reduced for single-member quorum scenarios. The test framework adds helpers to wait for confirmed masternodes and mine a single-member quorum. Quorum, ChainLock, and single-node tests use the new flow, while the governance objects test updates its node/test parameters.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Sequence Diagram(s)

sequenceDiagram
  participant FunctionalTest
  participant DashTestFramework
  participant Node
  FunctionalTest->>DashTestFramework: Configure single-member quorum
  FunctionalTest->>DashTestFramework: mine_quorum_single_member()
  DashTestFramework->>Node: Generate blocks across DKG cycle
  Node-->>DashTestFramework: Report new quorum
  DashTestFramework-->>FunctionalTest: Return quorum hash
  FunctionalTest->>Node: Query quorum info or verify ChainLock

Loading

Suggested reviewers: knst

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 7.14% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: reducing node counts in quorum-independent functional tests.
Description check	✅ Passed	The description matches the changes and summarizes the reduced topology, new quorum helper, and affected tests.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[PastaPastaPasta]

utACK 18aa0e1; I think the comments are over verbose; and commit history could be cleaner, but change itself looks fine

[thepastaclaw]

Code Review

Test-only PR that speeds up rpc_quorum.py and rpc_verifychainlock.py by converting them to single-member LLMQs via a new mine_quorum_single_member helper, and removes a genuinely unused masternode from feature_governance_objects.py. Verified against the local diff: the helper faithfully mirrors mine_quorum's cycle/maturity constants and the proven single-node DKG flow in feature_llmq_singlenode.py, node topology roles and all assertions are preserved, and the governance test only ever used nodes[0] with gobject list-prepared. No in-scope findings.

Source: reviewers — opus (general, completed), gpt-5.6-sol (general, failed: revoked ACP credential), opus (dash-core-commit-history, completed), gpt-5.6-sol (dash-core-commit-history, failed: revoked ACP credential); verifier — opus.

Note: GitHub does not allow me to approve my own PR, so this clean result is posted as a COMMENT.

[thepastaclaw]

Addressed in ea6794e: trimmed the helper and test comments to the non-obvious rationale, and squashed the four micro-commits into one atomic test commit. All three affected functional tests pass against a fresh exact-branch build; git diff --check and focused flake8 are clean.

[thepastaclaw]

Code Review

This PR reduces node and masternode counts in three quorum-independent functional tests by introducing a mine_quorum_single_member helper and single-member LLMQs. The latest code delta from 18aa0e1 to ea6794e only trims explanatory comments/docstrings, while the rewritten head also squashes the prior clean four-commit stack into one well-described test commit; cumulative behavior and every existing assertion remain unchanged. Prior-finding reconciliation is not applicable because the 18aa0e1 review had zero verified findings, so carried-forward prior findings are none; there are no new latest-delta or additional cumulative findings.

Source: reviewers — opus (general, completed), gpt-5.6-sol (general, failed: revoked ACP credential), opus (dash-core-commit-history, completed), gpt-5.6-sol (dash-core-commit-history, failed: revoked ACP credential); verifier — opus.

[thepastaclaw]

Code Review

PR #7441 is a clean test-only refactor that reduces functional-test node/masternode counts via a shared single-member quorum helper. The code is correct: node-index mapping and assertions are preserved, and the batched full-cycle mining loop is sound. The only kept finding is a non-blocking commit-hygiene suggestion.

Prior finding reconciliation

The prior review at ea6794ed had no verified findings, so reconciliation is NOT APPLICABLE and carried-forward prior findings are none.

New findings in the latest delta

One suggestion: fold the helper corrections in 1821ceff into the introducing commit ea6794ed, while keeping the feature_llmq_singlenode.py reuse as its own clean commit. Additional cumulative current-head findings: none.

Source: reviewers — opus (general, completed), gpt-5.6-sol (general, failed: revoked ACP credential), opus (dash-core-commit-history, completed), gpt-5.6-sol (dash-core-commit-history, failed: revoked ACP credential); verifier — opus (completed).

🟡 1 suggestion(s)

🤖 Prompt for all review comments with AI agents

These findings are from an automated code review. Verify each finding against the current code and only fix it if needed.

In `test/functional/test_framework/test_framework.py`:
- [SUGGESTION] test/functional/test_framework/test_framework.py:2232: Squash helper corrections from 1821ceff into ea6794ed; keep only the reuse as its own commit
  Commit 1821ceff96c is subject-lined "test: reuse single-member quorum helper in feature_llmq_singlenode", but its body opens with "Address review feedback on mine_quorum_single_member" and most of its changes rework the mine_quorum_single_member helper that commit ea6794ed48d added one commit earlier in this same, still-unmerged PR. Verified against the delta: it drops the skip_maturity parameter that ea6794ed introduced with no caller, rewrites ea6794ed's block-by-block `generate(node, 1)` loop into a full-cycle `generate(node, llmq_cycle_len)` batch, and trims the quorumIndex/is_mature log line ea6794ed added. Because Dash merges without squashing, permanent develop history would show ea6794ed adding a dead skip_maturity param and a block-by-block loop, then the very next commit removing/rewriting both — churn and git-blame noise on code that never shipped. Fold these three helper-only corrections into ea6794ed48d so the helper is introduced in its final form. The remaining change in 1821ceff — replacing feature_llmq_singlenode.py's local mine_single_node_quorum with the shared helper and dropping the now-unused time import — has durable standalone value and should stay as its own commit; once the helper fixups are squashed away, its current subject accurately describes it. Use git rebase to fixup the helper edits into ea6794ed and reword the residual commit body to drop the "Address review feedback" framing.

[thepastaclaw]

Code Review

Source: reviewers codex/general=gpt-5.6-sol (completed; parseable output with nonzero wrapper exit), sonnet5/general=claude-sonnet-5 (completed), codex/dash-core-commit-history=gpt-5.6-sol (completed), sonnet5/dash-core-commit-history=claude-sonnet-5 (completed); verifier=claude-sonnet-5; orchestrator=openai/gpt-5.6-sol (high reasoning, orchestration-only, not a reviewer/verifier).

The current two-commit stack is clean and the test-only topology reduction preserves the relevant node roles and assertions. The new readiness helper's work-block depth and regtest confirmation assumptions match the current Dash parameters; no new correctness, coverage, or history-hygiene findings remain.

Prior Reconciliation

• prior-1: FIXED — c55c54f3a5c now introduces mine_quorum_single_member in its final form (no unused skip_maturity, corrected full-cycle mining, and no single-member quorumIndex logging). 944e487c431 now touches only feature_llmq_singlenode.py to reuse the helper, so the thread's resolved/outdated state is accepted based on the rewritten commits.

Carried-Forward Prior Findings

None.

New Findings In Latest Delta

None. The literal 1821ceff..944e487c range crosses a rebase and includes unrelated intervening develop changes; range-diff and the current cumulative PR diff isolate the rewritten five-file change.

Additional Cumulative Findings

None.

Source (experiment sonnet-primary-opus-quarter-sample-20260710, cohort sonnet_primary, bucket 1, trigger new_push, prior head 1821cef, specialists dash-core-commit-history): reviewers codex/general=gpt-5.6-sol(completed_parseable_output_with_nonzero_wrapper_exit); sonnet5/general=claude-sonnet-5(completed); codex/dash-core-commit-history=gpt-5.6-sol(completed); sonnet5/dash-core-commit-history=claude-sonnet-5(completed); verifier=verifier-sonnet5-7441-1783823471=claude-sonnet-5; orchestrator=openai/gpt-5.6-sol reasoning=high (orchestration-only, not a reviewer/verifier).

[thepastaclaw]

The linux64_tsan-test failure is unrelated to this PR and does not require a branch change. It is the known intermittent p2p_blocksonly.py:77 relay-permission timeout tracked in #7436: node 0 accepted the transaction and queued its inventory for peer 1, but the test peer did not observe it within the TSAN timeout; no ThreadSanitizer report occurred.

This PR changes only quorum/governance functional tests and adds LLMQ-specific helpers. It does not touch p2p_blocksonly.py, the P2P peer framework, transaction relay, or inventory scheduling. All 318 other TSAN functional tests and every other test matrix job passed, so the PR is proper as-is. I recorded this occurrence on #7436; maintainers can rerun the flaky job if desired.