fix(core): clean up dangling/orphan blocks on SetHead by gzliudan · Pull Request #2442 · XinFinOrg/XDPoSChain

gzliudan · 2026-07-07T03:41:22Z

Proposed changes

SetHead rewinds the chain to a lower block height. Without comprehensive cleanup,
the downloader can jump to stale blocks instead of resyncing intermediate ranges.

SetHead previously only deleted the canonical chain segment, leaving orphaned/fork
blocks intact. The downloader's HasBlock() check would hit these orphaned entries,
incorrectly reporting ancestors found, skipping intermediate resync.

Root cause: Original implementation used a contiguous upward scan (checking
head+1, head+2, ...) that stopped at the first height with no header, leaving
orphaned segments above gaps intact.

Solution: Replace contiguous scan with ReadAllDanglingHashes(), which walks
the header keyspace directly via iterator and returns ALL headers above head
regardless of gaps. SetHead now performs post-loop sweep deleting all header/td/
body/receipt/canonical data above the new head.

For bad blocks: Only delete records strictly above the new head, preserving
debugging data for blocks at/below head for operator review (respects
DeleteAllBadBlocks config).

Optimization: Use two-phase batch commits to avoid duplicate deletes.

Changes:

Add ReadAllDanglingHashes() for iterator-based orphan detection
Add DeleteBadBlocksAbove() for selective bad block cleanup
Refactor SetHead() with two-phase batch writes and iterator-based sweep
Add TestSetHeadCleansDanglingBlocksAcrossGap for gap-handling validation
Add TestSetHeadCleansOnlyBadBlocksAboveHead for bad block preservation

Verification: 5/5 unit tests pass, remote testnet verified

Types of changes

What types of changes does your code introduce to XDC network?
Put an ✅ in the boxes that apply

Impacted Components

Which parts of the codebase does this PR touch?
Put an ✅ in the boxes that apply

Checklist

Put an ✅ in the boxes once you have confirmed below actions (or provide reasons on not doing so) that

This PR has sufficient test coverage (unit/integration test) OR I have provided reason in the PR description for not having test coverage
Tested on a private network from the genesis block and monitored the chain operating correctly for multiple epochs.
Provide an end-to-end test plan in the PR description on how to manually test it on the devnet/testnet.
Tested the backwards compatibility.
Tested with XDC nodes running this version co-exist with those running the previous version.
Relevant documentation has been updated as part of this PR
N/A

coderabbitai · 2026-07-07T03:41:30Z

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 505f306a-dc0c-4f0e-b838-3374e03f8e28

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

✨ 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

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

Copilot

Pull request overview

This PR ports upstream geth HeaderChain.SetHead cleanup behavior into XDPoSChain to ensure rewinds fully delete orphan/dangling block data above the rollback target, preventing stale hash-keyed block data from misleading the downloader’s ancestor search after restarts.

Changes:

Add rawdb.ReadAllHashes to enumerate all block hashes stored at a given height.
Extend core/HeaderChain.SetHead to also delete data for all hashes at each rewound height and to additionally wipe “leftover” heights above the current head on the first iteration.
Add core/blockchain_sethead_test.go to verify dangling blocks above head are fully wiped by SetHead.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 4 comments.

File	Description
core/rawdb/accessors_chain.go	Adds `ReadAllHashes` helper to enumerate hashes stored at a specific block height.
core/headerchain.go	Updates `HeaderChain.SetHead` to remove side-fork/orphan data and leftover heights above head during rewinds.
core/blockchain_sethead_test.go	Adds regression test ensuring dangling block data above head is removed after `SetHead`.

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

Copilot

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

SetHead rewinds the chain to a lower block height. Without comprehensive cleanup, the downloader can jump to stale blocks instead of resyncing intermediate ranges. SetHead previously only deleted the canonical chain segment, leaving orphaned/fork blocks intact. The downloader's HasBlock() check would hit these orphaned entries, incorrectly reporting ancestors found, skipping intermediate resync. Root cause: Original implementation used a contiguous upward scan (checking head+1, head+2, ...) that stopped at the first height with no header, leaving orphaned segments above gaps intact. Real-world data showed orphaned fork segments can sit ABOVE gaps (e.g., head at 83,531,000 with no markers at 83,531,501–83,532,999 but orphaned data at 83,533,000–83,599,999). Solution: Replace contiguous scan with ReadAllDanglingHashes(), which walks the header keyspace directly via iterator and returns ALL headers above head regardless of gaps. SetHead now performs post-loop sweep deleting all header/td/ body/receipt/canonical data above the new head. For bad blocks: Only delete records strictly above the new head, preserving debugging data for blocks at/below head for operator review (respects DeleteAllBadBlocks config). Optimization: Use two-phase batch commits to avoid duplicate deletes. Flush rewind deletions to database before scanning for dangling data, preventing ReadAllDanglingHashes from re-observing already-deleted segments. Changes: - Add ReadAllDanglingHashes(db, head) for iterator-based orphan detection - Add DeleteBadBlocksAbove(db, head) for selective bad block cleanup - Refactor SetHead() with two-phase batch writes and iterator-based sweep - Add TestSetHeadCleansDanglingBlocksAcrossGap for gap-handling validation - Add TestSetHeadCleansOnlyBadBlocksAboveHead for bad block preservation Verification: - 5/5 SetHead unit tests pass - Remote testnet: rollback 83,531,500→83,531,000 cleaned 69,000 heights, sequential block import without bad block jumps - Full core test suite passes

gzliudan requested review from AnilChinchawale, anunay-xin, benjamin202410, Copilot, liam-lai and wanwiset25 July 7, 2026 03:41

Copilot started reviewing on behalf of gzliudan July 7, 2026 03:42 View session

Copilot AI reviewed Jul 7, 2026

View reviewed changes

Comment thread core/headerchain.go Outdated

Comment thread core/headerchain.go Outdated

Comment thread core/rawdb/accessors_chain.go

Comment thread core/rawdb/accessors_chain.go Outdated

gzliudan force-pushed the fix-set-head branch from 01c637e to 001214b Compare July 7, 2026 05:03

gzliudan requested a review from Copilot July 7, 2026 05:08

Copilot started reviewing on behalf of gzliudan July 7, 2026 05:09 View session

Copilot AI reviewed Jul 7, 2026

View reviewed changes

Comment thread core/headerchain.go Outdated

gzliudan force-pushed the fix-set-head branch 3 times, most recently from 2d3c33e to c14dd35 Compare July 7, 2026 06:31

gzliudan changed the title ~~fix(core): wipe dangling/orphan blocks above head on SetHead~~ fix(core): clean up dangling/orphan/bad blocks on SetHead Jul 7, 2026

gzliudan requested a review from Copilot July 7, 2026 08:01

Copilot started reviewing on behalf of gzliudan July 7, 2026 08:01 View session

gzliudan changed the title ~~fix(core): clean up dangling/orphan/bad blocks on SetHead~~ fix(core): wipe orphaned blocks above head across gaps on SetHead Jul 7, 2026

Copilot AI reviewed Jul 7, 2026

View reviewed changes

Comment thread core/headerchain.go

Comment thread core/headerchain.go Outdated

gzliudan force-pushed the fix-set-head branch from 3654ffa to c3e4cef Compare July 7, 2026 08:41

gzliudan changed the title ~~fix(core): wipe orphaned blocks above head across gaps on SetHead~~ fix(core): clean up dangling/orphan blocks on SetHead Jul 7, 2026

gzliudan requested a review from Copilot July 7, 2026 08:42

Copilot started reviewing on behalf of gzliudan July 7, 2026 08:43 View session

Copilot AI reviewed Jul 7, 2026

View reviewed changes

Comment thread core/rawdb/accessors_chain.go Outdated

gzliudan force-pushed the fix-set-head branch from c3e4cef to 987accd Compare July 7, 2026 09:00

gzliudan requested a review from Copilot July 7, 2026 09:00

Copilot started reviewing on behalf of gzliudan July 7, 2026 09:00 View session

Copilot AI reviewed Jul 7, 2026

View reviewed changes

Comment thread core/headerchain.go

gzliudan force-pushed the fix-set-head branch from 987accd to 424c179 Compare July 7, 2026 09:21

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix(core): clean up dangling/orphan blocks on SetHead#2442

fix(core): clean up dangling/orphan blocks on SetHead#2442
gzliudan wants to merge 1 commit into
XinFinOrg:dev-upgradefrom
gzliudan:fix-set-head

gzliudan commented Jul 7, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Jul 7, 2026 •

edited

Loading

Review skipped

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

gzliudan commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Proposed changes

Types of changes

Impacted Components

Checklist

Uh oh!

coderabbitai Bot commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

gzliudan commented Jul 7, 2026 •

edited

Loading

coderabbitai Bot commented Jul 7, 2026 •

edited

Loading