Fix several interacting issues that prevented partial state nodes from
syncing and following the chain on bal-devnet-2:
1. Stale pivot deadlock: Replace unconditional pivot suppression with
rate-limited advances (2-minute cooldown). This prevents the restart
loop bug while allowing recovery when the initial pivot is too stale
for peers to serve.
2. Storage root resolution: Add snap-based resolver that queries peers
for untracked contracts' storage roots during BAL processing. This
lets the computed state root converge toward the header root.
3. SetCanonical for partial state: When the computed root differs from
the header root (expected when untracked contracts have unresolved
storage roots), check HasState(partialState.Root()) instead of only
HasState(block.Root()). Guard against zero root during snap sync.
4. Canonical hash backfill: AdvancePartialHead now writes canonical
hashes for all blocks between the pivot and snap head, fixing the
"final block not in canonical chain" error caused by
InsertReceiptChain skipping blocks whose bodies already exist.
5. Gap block processing: After snap sync completes, process accumulated
blocks between the sync head and chain tip using their persisted BALs
before entering steady-state chain following.
6. Computed root chaining: Use partialState.Root() (actual computed root)
as parentRoot for subsequent blocks, not the header root. This ensures
correct trie chaining when computed != header root.
Tested end-to-end on bal-devnet-2: snap sync completes, gap blocks
processed, canonical head advances at chain tip (~1 block/12s).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Freeze the pivot header for partial state nodes to ensure stable state
sync progress:
- Suppress pivot movement in fetchHeaders() (beaconsync.go)
- Suppress pivot movement in processSnapSyncContent() (downloader.go)
- Reuse existing pivot across sync cycle restarts in syncToHead()
After initial snap sync completes, bridge the gap from pivot to HEAD:
- Import post-pivot blocks with receipts (no execution needed since
untracked contracts have empty storage tries)
- Run second state sync to download HEAD state root
- Add AdvancePartialHead to update currentBlock without re-execution
Guard the backfiller for partial state mode:
- suspend() skips Cancel() during active snap sync to prevent
constant cancel/restart cycles from beacon head updates
- resume() skips new sync cycles after partial sync completes
Add chain retention for partial state mode: only the most recent N blocks
(default 1024) retain bodies and receipts. During sync, older blocks are
skipped entirely. After sync, the freezer enforces a rolling window.
Add engine API support for Block Access Lists (EIP-7928): NewPayloadV5
accepts BAL data alongside execution payloads, enabling partial state
nodes to receive per-block storage access information from the CL.
Fix beacon backfilling failure caused by dynamic chain cutoff not
clearing the cutoff hash (which remained at the genesis hash).
Add partial state awareness to eth_call/eth_estimateGas to return clear
errors when accessing untracked contract storage.
In this PR, two things have been fixed:
---
(a) truncate the stale beacon headers with latest snap block
Originally, b.filled is used as the indicator for deleting stale beacon headers.
This field is set only after synchronization has been scheduled, under the
assumption that the skeleton chain is already linked to the local chain.
However, the local chain can be mutated via `debug_setHead`, which may
cause `b.filled` outdated. For instance, `b.filled` refers to the last head snap block
in the last sync cycle while after `debug_setHead`, the head snap block has been
rewounded to 1.
As a result, Geth can enter an unintended loop: it repeatedly downloads
the missing beacon headers for the skeleton chain and attempts to schedule the
actual synchronization, but in the final step, all recently fetched headers are removed
by `cleanStales` due to the stale `b.filled` value.
This issue is addressed by always using the latest snap block as the indicator,
without relying on any cached value. However, note that before the skeleton
chain is linked to the local chain, the latest snap block will always be below
skeleton.tail, and this condition should not be treated as an error.
---
(b) merge the subchains once the skeleton chain links to local chain
Once the skeleton chain links with local one, it will try to schedule the
synchronization by fetching the missing blocks and import them then.
It's possible the last subchain already overwrites the previous subchain and
results in having two subchains leftover. As a result, an error log will printed
https://github.com/ethereum/go-ethereum/blob/master/eth/downloader/skeleton.go#L1074
This moves the tracking of the current syncmode into the downloader, fixing an
issue where the syncmode being requested through the engine API could go
out-of-sync with the actual mode being performed by downloader.
Fixes#32629
---------
Co-authored-by: Felix Lange <fjl@twurst.com>
This pull request introduces new sync logic for pruning mode. The downloader will now skip
insertion of block bodies and receipts before the configured history cutoff point.
Originally, in snap sync, the header chain and other components (bodies and receipts) were
inserted separately. However, in Proof-of-Stake, this separation is unnecessary since the
sync target is already verified by the CL.
To simplify the process, this pull request modifies `InsertReceiptChain` to insert headers
along with block bodies and receipts together. Besides, `InsertReceiptChain` doesn't have
the notion of reorg, as the common ancestor is always be found before the sync and extra
side chain is truncated at the beginning if they fall in the ancient store. The stale
canonical chain flags will always be rewritten by the new chain. Explicit reorg logic is
no longer required in `InsertReceiptChain`.
Lots of packages depend on eth/downloader just for the SyncMode type.
Since we have a dedicated package for eth protocol configuration, it
makes more sense to define SyncMode there, turning eth/downloader into
more of a leaf package.
This PR fixes an issue in the setMode method of beaconBackfiller where the
log message was not displaying the previous mode correctly. The log message
now shows both the old and new sync modes.
time.After is equivalent to NewTimer(d).C, and does not call Stop if the timer is no longer needed. This can cause memory leaks. This change changes many such occations to use NewTimer instead, and calling Stop once the timer is no longer needed.
* eth/downloader: fix skeleton cleanup
* eth/downloader: short circuit if nothing to delete
* eth/downloader: polish the logic in cleanup
* eth/downloader: address comments
In legacy (pre-merge) sync mode, headers were contiguously downloaded from the network and when no more headers were available, we checked every few seconds whether there are 64 new blocks to move the pivot.
In beacon (post-merge) sync mode, we don't need to check for new skeleton headers non stop, since those re delivered one by one by the engine API. The missing code snippet from the header fetcher was to actually look at the latest head and move the pivot if it was more than 2*64-8 away. This PR adds the missing movement logic.
* eth: support bubbling up bad blocks from sync to the engine API
* eth/catalyst: fix typo
Co-authored-by: Marius van der Wijden <m.vanderwijden@live.de>
* eth/catalyst: fix typo
Co-authored-by: Marius van der Wijden <m.vanderwijden@live.de>
* Update eth/catalyst/api.go
* eth/catalyst: when forgetting bad hashes, also forget descendants
* eth/catalyst: minor bad block tweaks for resilience
Co-authored-by: Marius van der Wijden <m.vanderwijden@live.de>
Co-authored-by: Martin Holst Swende <martin@swende.se>
* eth/downloader: remove stale beacon headers as backfilling progresses
* eth/downloader: remove leftover from a previous design
* eth/downloader: do partial beacon cleanups if chain is large
* eth/downloader: linter != heart
* eth/downloader: implement beacon sync
* eth/downloader: fix a crash if the beacon chain is reduced in length
* eth/downloader: fix beacon sync start/stop thrashing data race
* eth/downloader: use a non-nil pivot even in degenerate sync requests
* eth/downloader: don't touch internal state on beacon Head retrieval
* eth/downloader: fix spelling mistakes
* eth/downloader: fix some typos
* eth: integrate legacy/beacon sync switchover and UX
* eth: handle UX wise being stuck on post-merge TTD
* core, eth: integrate the beacon client with the beacon sync
* eth/catalyst: make some warning messages nicer
* eth/downloader: remove Ethereum 1&2 notions in favor of merge
* core/beacon, eth: clean up engine API returns a bit
* eth/downloader: add skeleton extension tests
* eth/catalyst: keep non-kiln spec, handle mining on ttd
* eth/downloader: add beacon header retrieval tests
* eth: fixed spelling, commented failing tests out
* eth/downloader: review fixes
* eth/downloader: drop peers failing to deliver beacon headers
* core/rawdb: track beacon sync data in db inspect
* eth: fix review concerns
* internal/web3ext: nit
Co-authored-by: Marius van der Wijden <m.vanderwijden@live.de>
