Under normal usage conditions, the `transactions` table is not currently
populated for transactions involving transparent UTXOs, and so this join
was always resulting in transparent UTXO information being filtered out
from the transaction history.
Fixes [zcash/ZcashLightClientKit#1271]
This adds the `data_api::scanning::spanning_tree` module under
a new `unstable-spanning-tree` feature flag, making it available to
other implementations who want to be able to write their own storage
backends without having to reinvent the spanning tree logic.
This fixes the following bug:
Due to complexities related to non-linear scanning, checkpoints are only
added to the wallet's commitment tree in cases where there are notes
discovered within a scanned block. At present, the `shardtree` API only
makes it possible to add multiple checkpoints of the same tree state
when adding checkpoints at the chain tip, and this functionality is not
used by `zcash_client_backend` because we perform checkpoint insertion
in batches that may contain gaps in the case that multiple blocks
contain no Sapling notes. While it would be possible to fix this by
altering the `shardtree` API to permit explicit insertion of multiple
checkpoints of the same tree state at a given note position, this fix
takes a simpler approach.
Instead of ensuring that a checkpoint exists at every block and
computing the required checkpoint depth directly from the minimum number
of confirmations required when attempting a spend, we alter the
`WalletCommitmentTrees` API to allow internal information of the note
commitment tree to be used to determine this checkpoint depth, given the
minimum number of commitments as an argument. This allows us to select a
usable checkpoint from the sparse checkpoint set that resulted from the
sparse insertion of checkpoints described above.
The `v_sapling_shard_scan_ranges` view pairs every scan range with every
shard range, such that each row shows an overlapping pair.
For the complete shards, this is an overlap check between two ranges,
which the previous query was performing correctly (if verbosely).
For the last incomplete shard, we have a half-open range that needs to
be handled separately. The previous query only handled the case where a
scan range was contained within the last shard, and did not handle the
case where the scan range contained the last shard.
This led to a puzzling bug, where `WalletDb::get_wallet_summary` was
sometimes treating any note received within the last shard as part of
the wallet's pending balance. If the wallet's scan queue contained a
range that encompassed the last incomplete shard, the bug in the
`v_sapling_shard_scan_ranges` view meant that it omitted any mention of
the last shard, which translated into these notes being considered
unmined when joining `sapling_received_notes` against the sub-view
`v_sapling_shards_scan_state`.
The bug was made harder to diagnose due to the previous commit's bug
that was causing scan ranges to not be correctly merged; this resulted
in smaller scan ranges that were more likely to be contained within the
last shard, making it visible in `v_sapling_shard_scan_ranges` and
enabling notes to be detected as mined.
The fixed view uses a simpler query that enables us to handle complete
and incomplete shards together.
Time spent investigating and fixing: 4.5 hours
Co-authored-by: Kris Nuttycombe <kris@nutty.land>
In the sandblasting block ranges, shard trees only cover a few hundred
blocks at most. When scanning block ranges much larger than this, it is
likely that when a note is discovered, its parent shard tree is entirely
contained within the scan range. In this situation, `extended_range`
would be set to `range`, and then because an extended range existed,
ranges with `FoundNote` priority would always be created - that in this
case are empty.
In an effectively-linear-scanning wallet situation, this leads to a
`SpanningTree` being constructed with adjacent `Scanned` ranges,
separated by empty `FoundNote` ranges, which it was unable to merge.
We address this by both preventing generation of the empty `FoundNote`
ranges, and by defensively fixing `SpanningTree::into_vec` to skip empty
ranges.
Co-authored-by: Kris Nuttycombe <kris@nutty.land>
This modifies `update_chain_tip` and `scan_complete` to ensure that
newly created scan ranged do not extend below the wallet birthday
height.
Fixes#947
The previous code was checking that the existing shard range was at
least partially inside the proposed insertion range, but this didn't
handle the case where the proposed insertion range was fully contained
by the existing shard range.
The intent of this API is to provide a single API which returns in a
single call:
* per-account balances, including pending values
* wallet sync progress
Fixes#865Fixes#900
This also removes the zcash_client_sqlite-specific database
initialization procedures in favor of a standardized approach using the
methods available via the data access API.
When `force_rescans` is set to `true` in a call to
`replace_queue_entries`, previously scanned ranges will have their
existing priority overwritten by the scan priority for a provided range;
otherwise, the existing scan priority dominance rule continues to be
enforced. This enables us to require previously scanned ranges be
re-scanned without interfering with higher-priority scan operations.
This change modifies the implementation of `get_spendable_sapling_notes`
and `select_spendable_sapling_notes` to only return notes at positions
where the associated note commitment tree shard has been fully scanned.
This is slightly more conservative than it needs to be, because
there could be cases where witnesses imported into the tree in the
`shardtree_support` migration cover the complete range of a subtree (and
hence that subtree doesn't need to be re-scanned). However, we can't
detect or depend upon that condition in general without attempting to
create a witness for each note retrieved.
A possible alternative to this approach would be to not bound our query
results on the requested total, and instead attempt to construct a
witness for each note we retrieve, skipping the note if we cannot
construct a witness. However, given that accessing the note commitment
tree can be a costly operation requiring nontrivial deserialization
costs, the more conservative database-oriented approach is perhaps
better.
Prior to the scan-before-sync changes, the wallet was able to assume
that the maximum scanned block height at the time of the spend was
within a few blocks of the chain tip. However, under linear scanning
after the spend-before-sync changes this invariant no longer holds,
resulting in a situation where in linear sync conditions the wallet
could attempt to create transactions with already-past expiry heights.
This change separates the notion of "chain tip" from "max scanned
height", relying upon the `scan_queue` table to maintain the wallet's
view of the consensus chain height and using information from the
`blocks` table only in situations where the latest and/or earliest
scanned height is required.
As part of this change, the `WalletRead` interface is also modified to
disambiguate these concepts.
Previously this was not clearly specified, and the implementations in
`zcash_client_sqlite` behaved similarly to when `from_height = None`.
Closeszcash/librustzcash#892.
In general, it is preferable to use globally relevant identifiers where
possible. This PR removes the `WalletRead::TxRef` associated type in
favor of using `TxId` directly for the transaction identifier, and
restricts the use of the `NoteRef` type to those scenarios where the
result of one query is intended to be used directly as the input to
another query.
Closes#834
The `add_checkpoint` method is intended to be idempotent. In the case
that we add a checkpoint at an already-checkpointed block height, we
should only raise an error in the case that the note commitment tree
position or the set of notes spent in the checkpointed block has
changed.
The `shardtree` migration is applied to a database state that was
created via linear scanning, so we have complete wallet information for
those blocks.
We only need to load frontiers into the ShardTree that are close enough
to the wallet's known chain tip to fill `PRUNING_DEPTH` checkpoints, so
that ShardTree's witness generation will be able to correctly handle
anchor depths. Loading frontiers further back than this doesn't add any
useful nodes to the ShardTree (as we don't support rollbacks beyond
`PRUNING_DEPTH`, and we won't be finding notes in earlier blocks), and
hurts performance (as frontier importing has a significant Merkle tree
hashing cost).
Closeszcash/librustzcash#877.
Previously `extended_range` only covered the extent of the leaves of
all subtrees in which notes were found during a scan. When the scanned
range was large, this was not guaranteed to be contained within the
subtree leaves, causing an assertion failure when an invalid `ScanRange`
was constructed.
The Merkle hashes used for the note commitment trees are domain
separated by level, so when pretending that the subtree roots are leaves
of the cap tree, we need to adjust for their level not being zero.
Closeszcash/librustzcash#874.
Co-authored-by: Sean Bowe <ewillbefull@gmail.com>
This implements a priority queue backed by the wallet database for scan
range ordering. The scan queue is updated on each call to `put_blocks`
or to `update_chain_tip`.
Instead of calling `put_block` for each block scanned,
`scan_cached_blocks` will now defer the block writes until the scan of a
batch is complete and will perform the block writes and note commitment
tree updates all within a single transaction.
This should ordinarily be fine in terms of memory consumption, because
the block data being saved is pruned to only that spend an output
information that is related to transactions in the wallet, which will
normally be sparse enough that the block range size that is appropriate
for a given platform to run within a batch scanner adequately bounds the
memory consumption of this pruned representation.
There are cases where we wish to return informaiton that is relevant to
a specific shielded protocol and `Transparent` is an invalid case. This
is a minor preparatory refactoring that makes this distinction
expressible.
In preparation for out-of-order range-based scanning, it is necessary
to ensure that the size of the Sapling note commitment tree is carried
along through the scan process and that stored blocks are always
persisted with the updated note commitment tree size.
Local chain validation will be performed internal to
`scan_cached_blocks`, and as handling of chain reorgs will need to
change to support out-of-order scanning, the `validate_chain` method
will be superfluous. It is removed in advance of other changes in order
to avoid updating it to reflect the forthcoming changes.
We move thes fields out into a separate BlockMetadata struct to ensure
that future additions to block metadata are structurally separated from
future additions to block data.