Repository navigation
chore(platform)!: pin GroveDB at develop 89dc7351 - #5942
Conversation
Moves every grovedb crate from fa92511c to 89dc7351, bringing in dashpay/grovedb#1036, #1037, #1038, #1039, #1040, #1041, #1042, #1043, #1044, #1045, #1046, #1047 and #1048. #1037, #1040, #1041 and #1043 are gated on GROVE_V5, which protocol version 15 selects; protocol version 14 stays on GROVE_V4, whose tables only gain new slots set to 0. Cargo.lock changes only the grovedb source lines. finalize_block and the startup flush now warn when a pending prefix drop flush leaves records unresolved (the new unresolved_records and first_unresolved_error fields of PendingPrefixDropsReport from #1045), as agreed in that PR's review. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
No actionable comments were generated in the recent review. Walkthrough
Merge Risk:
|
|
Waiting for bot review — thepastaclaw not yet. Wait for the missing reviews, or a writer can post |
|
✅ Final review complete — no blockers (commit 3dcd608) · triage: normal |
thepastaclaw
left a comment
There was a problem hiding this comment.
Final validation — Phase 1 + Phase 2
The GroveDB pin and Platform integration are consistent, and the unresolved pending-drop logging is correctly outside consensus. One existing completeness test is now too weak for the imported flush semantics because it can pass after reclaiming some records while leaving others unresolved. The Rust workspace CI failure is an unrelated pre-existing formatting issue in an unchanged file.
🟡 1 suggestion(s)
Approved: nothing here blocks the merge. The non-blocking findings are optional: fix them forward, in this PR or a follow-up, without waiting for another review round.
1 finding(s) not shown inline (the lines are not part of this PR's diff)
🟡 Suggestion: Assert that no pending-drop records remain unresolved
packages/rs-drive-abci/src/execution/validation/state_transition/state_transitions/masternode_vote/no_locking_contest_tests.rs:1605-1606
The pinned GroveDB implementation can leave a record unresolved when its liveness check fails, while continuing to reclaim later records in the same pass. Consequently, skipped_live == 0 and reclaimed_records > 0 do not establish the test's stated invariant that all dropped buckets were reclaimed: the test could still pass with unresolved_records > 0. Add an explicit zero assertion so this completeness check covers the new report signal.
assert_eq!(report.skipped_live, 0);
assert_eq!(report.unresolved_records, 0);
assert!(report.reclaimed_records > 0);
source: muse-spark-1.3-contributor (phase1-reviewer: general)
Review provenance
Source: reviewer 1: muse-spark-1.3-contributor (agent: phase1-reviewer, role: general); reviewer 2: gpt-6.1-sol (agent: phase2-reviewer, role: general); reviewer 3: gpt-6.1-sol (agent: phase2-reviewer, role: architecture-layering); reviewer 4: gpt-6.1-sol (agent: phase2-reviewer, role: platform-versioning); reviewer 5: gpt-6.1-sol (agent: phase2-reviewer, role: tenderdash-consensus-security); reviewer 6: gpt-6.1-sol (agent: phase2-reviewer, role: general); reviewer 7: gpt-6.1-sol (agent: phase2-reviewer, role: architecture-layering); reviewer 8: gpt-6.1-sol (agent: phase2-reviewer, role: platform-versioning); reviewer 9: gpt-6.1-sol (agent: phase2-reviewer, role: tenderdash-consensus-security); final verifier: gpt-6.1-sol (agent: sol-verifier, role: final-verifier)
- Triage:
normalbygpt-6.1-sol(effort low) — This is a contained dependency pin and logging change, but the GroveDB update can alter unreleased protocol-version behavior, so it warrants ordinary review without meeting the critical-tier bar. - Phase 1 reviewers:
muse-spark-1.3-contributor— general (completed, effort xhigh); agentphase1-reviewer - Phase 1 model:
muse-spark-1.3-contributor— not quota-gated; passed overgemini-3.8-flash-high(antigravity quota lookup failed),glm-5.3-flash(not used above high effort; tier asks max) - Second Phase-2 pass: on this first full review every Phase-2 reviewer also ran a second, independent time, and the verifier weighed both sets
- Fresh verifier:
gpt-6.1-sol— final-verifier; agentsol-verifier - Phase 2 reviewers:
gpt-6.1-sol— general (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— architecture-layering (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— platform-versioning (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— tenderdash-consensus-security (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— general (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— architecture-layering (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— platform-versioning (completed, effort high); agentphase2-reviewer,gpt-6.1-sol— tenderdash-consensus-security (completed, effort high); agentphase2-reviewer
🤖 Prompt for all review comments with AI agents
These findings are from an automated code review (PastaClaw). How to work through them:
- Verify each finding against the code; fix only what is real, as narrowly as the finding.
- If a finding is wrong, decline it on its thread in one or two sentences; the bot answers once and stops.
- To defer one, link a tracking issue on its thread (or reply asking the bot to file one).
- Suggestions don't block the merge, but most are worth fixing in this PR; judge each on its merits.
- Keep the PR's scope; don't grow it to satisfy hardening suggestions.
- Carried items were raised in an earlier round: read the thread first, and don't re-fix what you already fixed, declined or deferred there. Disputed and deferred items are already handled and are not listed.
In `packages/rs-drive-abci/src/execution/validation/state_transition/state_transitions/masternode_vote/no_locking_contest_tests.rs`:
- [SUGGESTION] packages/rs-drive-abci/src/execution/validation/state_transition/state_transitions/masternode_vote/no_locking_contest_tests.rs:1605-1606: Assert that no pending-drop records remain unresolved
The pinned GroveDB implementation can leave a record unresolved when its liveness check fails, while continuing to reclaim later records in the same pass. Consequently, `skipped_live == 0` and `reclaimed_records > 0` do not establish the test's stated invariant that all dropped buckets were reclaimed: the test could still pass with `unresolved_records > 0`. Add an explicit zero assertion so this completeness check covers the new report signal.
Out-of-scope follow-up suggestions (1)
These are valid observations, but they are outside this PR's scope and should be handled in separate issues or author/maintainer-requested PRs rather than blocking this review.
- Pre-existing Rust formatting failure blocks the workspace check — The Rust workspace test job fails during
cargo fmt --checkinpackages/rs-drive-abci/src/execution/validation/state_transition/processor/traits/is_allowed.rs. That file is unchanged by this pull request, so the failure is not caused by the GroveDB pin or its logging changes.- Follow-up: Fix the formatting drift in a separate follow-up pull request or update the base branch before relying on the Rust workspace check.
|
Bots are done — your move: post |
Basic explanation
Platform pinned GroveDB at
fa92511c. This moves every grovedb crate to89dc7351, the current head ofdevelop, so protocol version 15 runs on the GroveDB it will ship with. Protocol version 14 stays on GROVE_V4 and behaves as before.Issue being fixed or feature implemented
Brings in dashpay/grovedb#1036, #1037, #1038, #1039, #1040, #1041, #1042, #1043, #1044, #1045, #1046, #1047 and #1048.
#1037, #1040, #1041 and #1043 are gated on GROVE_V5, which protocol version 15 already selects (
DRIVE_VERSION_V10). GROVE_V1 to V4 only gain new version slots set to 0, so protocol version 14 and earlier are unchanged.What was done?
Cargo.toml: grovedb revsfa92511c->89dc7351.Cargo.lock: only the grovedb source lines change.finalize_blockand the startup flush log a warning when a pending prefix drop flush leaves records unresolved, using theunresolved_recordsandfirst_unresolved_errorfields that fix(delete)!: keep draining pending drops past a record that cannot be resolved grovedb#1045 added toPendingPrefixDropsReport. This was agreed in that PR's review. It is logging only and outside consensus.Before (
finalize_block,Ok(report)): only the reclaimed and skipped-live counts were logged, at debug.After: the same, plus
The startup flush gets the same warning (it previously logged only an
Err).How Has This Been Tested?
cargo check --locked -p drive -p drive-abci -p dpp -p drive-proof-verifier -p dash-sdk --all-targets--lib: contract index e2e tests, count, ranked and sum queries,drive::document::,drive::shielded,verify::(1306 passed)--lib:create_genesis_state,perform_events_on_first_block_of_protocol_change,shield,masternode_vote,address_funding_from_asset_lock, and the document batch tests for deletion, ranked group drain, TTL, action fees, gas sponsorship, replacement, transfer, private types, keep history, index only and creation (1034 passed)run_chain_heavy_rotation_deterministic_before_payout,run_chain_proposer_proposes_a_chainlock_that_would_remove_themselves_from_the_list_deterministic,run_chain_insert_one_new_identity_per_block_with_epoch_change(3 passed)cargo test --tests -p dpp -p drive -p drive-abci -p dash-sdk -p drive-proof-verifier --all-features -- shield(all passed, includingnullifier_batch_guard)--test deterministic_root_hash(2 passed)No root hash or fee pin changed at any protocol version.
Breaking Changes
Protocol version 15 consensus behaviour changes where the GROVE_V5-gated GroveDB PRs above apply. Protocol version 15 is unreleased. Protocol version 14 and earlier are unchanged.
In-place changes to shipped generations
None.
Checklist:
structure.rs, regeneratedgrovedb-structure.json, and checked the structure viewer link posted on this pull requestFor repository code-owners and collaborators only
🤖 Generated with Claude Code
PR Hygiene ·
3dcd608/self-reviewedWhen every merge requirement is met, the
PR Hygienecheck passes. Reviewer limits do not block merging; other required GitHub checks and protections still apply.Summary by CodeRabbit