This is a really non-trivial wallet refactor "effective value" which has been
in the Core queue for a pretty long time. There were a ton of merge conflicts
but they were mostly mechanical (all the benchmarks) or "take the Core side
of the diff" (most of wallet.cpp).
The two nontrivial things were:
* In `SelectCoinsMinConf` in wallet.cpp I had to slightly move the asset
filtering logic (but I didn't change it)
* In `OutputGroup::Insert' in coinselection.cpp I changed the effective
value computation to only consider policyAsset values
725d7ae049 Use PrecomputedTransactionData in signet check (Pieter Wuille)
497718b467 Treat amount<0 also as missing data for P2WPKH/P2WSH (Pieter Wuille)
3820090bd6 Make all SignatureChecker explicit about missing data (Pieter Wuille)
b77b0cc507 Add MissingDataBehavior and make TransactionSignatureChecker handle it (Pieter Wuille)
Pull request description:
Currently we have 2 levels of potentially-missing data in the transaction signature hashes:
* P2WPKH/P2WSH hashes need the spent amount
* P2TR hashes need all spent outputs (amount + scriptPubKey)
Missing amounts are treated as -1 (thus leading to unexpected signature failures), while missing outputs in P2TR validation cause assertion failure. This is hard to extend for signing support, and also quite ugly in general.
In this PR, an explicit configuration option to {Mutable,}TransactionSignatureChecker is added (MissingDataBehavior enum class) to either select ASSERT_FAIL or FAIL. Validation code passes ASSERT_FAIL (as at validation time all data should always be passed, and anything else is a serious bug in the code), while signing code uses FAIL.
The existence of the ASSERT_FAIL option is really just an abundance of caution. Always using FAIL should be just fine, but if there were for some reason a code path in consensus code was introduced that misses certain data, I think we prefer as assertion failure over silently introducing a consensus change.
Potentially useful follow-ups (not for this PR, in my preference):
* Having an explicit script validation error code for missing data.
* Having a MissingDataBehavior::SUCCEED option as well, for use in script/sign.cpp DataFromTransaction (if a signature is present in a witness, and we don't have enough data to fully validate it, we should probably treat it as valid and not touch it).
ACKs for top commit:
sanket1729:
reACK 725d7ae049
Sjors:
ACK 725d7ae049
achow101:
re-ACK 725d7ae049
benthecarman:
ACK 725d7ae049
fjahr:
Code review ACK 725d7ae049
Tree-SHA512: d67dc51bae9ca7ef6eb9acccefd682529f397830f77d74cd305500a081ef55aede0e9fa380648c3a8dd4857aa7eeb1ab54fe808979d79db0784ac94ceb31b657
Includes a memory leak in the checkqueue unit test (but not in
the actual code). WE really need to switch our checkqueue to use
std::unique_pointer rather than bare pointers. But this would be
invasive enough that I want to do it in a followup PR.
Also pretty-much disable the validation_flush_cache unit test.
This is a stupid and irritating test which tries to unit-test
exact memory usage of std containers. It already has at least
one "remove wrong assumptions" update upstream and after many
tries I was unable to change all the magic numbers in a way
that'd consistently pass CI for Elements.
Also adds a couple ubsan suppressions about perfectly-legitimate
conversions of integer types.
Several conflicts in the C++ code related to the new `flags` parameter
to `CheckSignature` and the corresponding function being renamed upstream
to `CheckSignatureECDSA`.
Several conflicts in the test harness as Steven sorta pulled the new
upstream ECKey module into the Python code, and the actual upstream
code was slightly different. Also needed to update the feature_taproot
code to always use the non-RANGEPROOF sighash since dynafed is not
enabled in the Taproot test.
Also had to pull the `set_wif` method out of `ECKey` and inline it because
otherwise it triggers a "circular inclusion" error between script.py (which
would pull in `base58_to_bytes` from address.py) and address.py (which now
pulls in some taproot EC related stuff from script.py).
Noticed that #960 does not test the "sighash rangeproof flag set but no
witnesses" case.
ebc4ab721b refactor: post Optional<> removal cleanups (fanquake)
57e980d13c scripted-diff: remove Optional & nullopt (fanquake)
Pull request description:
Same rationale & motivation as #21404, which turned out to be quite low in the number of potential conflicts. Lets see what the bot has to say here.
ACKs for top commit:
practicalswift:
cr ACK ebc4ab721b: patch looks correct
jnewbery:
utACK ebc4ab721b
laanwj:
Code review ACK ebc4ab721b
Tree-SHA512: 550fbeef09b9d35ddefaa805d1755c18c8fd499c4b0f77ebfece8c20296a7abd1cf6c699e2261f92fe3552deeb7555ec2a2287ffe3ab9e98bb9f8612a4d43be3
f9cd2bfbcc Rename CoinSelectionParams::effective_fee to m_effective_feerate (Andrew Chow)
bdd0c2934b wallet: Move discard feerate fetching to CreateTransaction (Andrew Chow)
448d04b931 wallet: Move long term feerate setting to CreateTransaction (Andrew Chow)
e2f429e6bb wallet: Replace nFeeRateNeeded with effective_fee (Andrew Chow)
1a6a0b0dfb wallet: Use existing feerate instead of getting a new one (Andrew Chow)
Pull request description:
During coin selection, there are various places where we need to have a feerate. We need the feerate for the transaction itself, the discard fee rate, and long term feerate. Fetching these each time we need them can lead to a race condition where two feerates that should be the same are actually different. One particular instance where this can happen is during the loop in `CreateTransactionInternal`. After inputs are chosen, the expected transaction fee is calculated using a newly fetched feerate. If `pick_new_inputs == false`, the loop will go again with the assumption that the fee for the transaction remains the same. However because the feerate is fetched again, it is possible that it actually isn't and this causes coin selection to fail.
Instead of fetching the feerate each time it is needed, we fetch them all at once at the top of `CreateTransactionInternal`, store them in `CoinSelectionParams`, and use them where needed.
While some of these fee rates probably don't need this caching, I've done it for consistency and the guarantee that they remain the same.
Fixes#19229
ACKs for top commit:
glozow:
reACK f9cd2bfbcc
fjahr:
Code review re-ACK f9cd2bfbcc
Xekyo:
tACK https://github.com/bitcoin/bitcoin/pull/21083/commits/f9cd2bfbccb7a2b8ff07cec5f6d2adbeca5f07c3
meshcollider:
Code review + test run ACK f9cd2bfbcc
Tree-SHA512: be83ff64ba473c3cdd3469c812e214659b6e2a9584c22ed2b1595618fce0d4b35d0901e61068cd1069fc1a8fb911db01dd7312d05c3b8cbafbe2504ab7a3e863
Instead of fetching the discard feerate for each SelectCoinsMinConf
iteration, fetch and cache it once during CreateTransaction so that it
is shared for each SelectCoinsMinConf through
coin_selection_params.m_discard_feerate.
Does not change behavior.
Instead of setting the long term feerate for each SelectCoinsMinConf
iteration, set it once during CreateTransaction and let it be shared
with each SelectCoinsMinConf through
coin_selection_params.m_long_term_feerate.
Does not change behavior.
Remove the implicit MissingDataBehavior::ASSERT_FAIL in the
*TransationSignatureChecker constructors, and instead specify
it explicit in all call sites:
* Test code uses ASSERT_FAIL
* Validation uses ASSERT_FAIL (through CachingTransactionSignatureChecker)
(including signet)
* libconsensus uses FAIL, matching the existing behavior of the
non-amount API (and the extended required data for taproot validation
is not available yet)
* Signing code uses FAIL
e3e0a2432c Add benchmark to write JSON into a string (Martin Ankerl)
Pull request description:
The benchmark `BlockToJsonVerbose` only tests generating (and destroying)
the JSON data structure, but serializing into a string is also a
performance critical aspect of the RPC calls.
Extracts test setup into a `struct TestBlockAndIndex`, and uses it in
both `BlockToJsonVerbose` and `BlockToJsonVerboseWrite`.
Also, use `ankerl::nanobench::doNotOptimizeAway` to make sure the compiler
can't optimize the result of the calls away.
Here are benchmark results on my Intel i7-8700:
| ns/op | op/s | err% | ins/op | cyc/op | IPC | bra/op | miss% | total | benchmark
|--------------------:|--------------------:|--------:|----------------:|----------------:|-------:|---------------:|--------:|----------:|:----------
| 71,807,017.00 | 13.93 | 0.4% | 555,782,961.00 | 220,788,645.00 | 2.517 | 102,279,341.00 | 0.4% | 0.80 | `BlockToJsonVerbose`
| 27,916,835.00 | 35.82 | 0.1% | 235,084,034.00 | 89,033,525.00 | 2.640 | 42,911,139.00 | 0.3% | 0.32 | `BlockToJsonVerboseWrite`
ACKs for top commit:
laanwj:
Code review ACK e3e0a2432c
Tree-SHA512: bc4d6d1588d47d4bd7af8e7908e44b8561bc391a2d73eccd7c0aa37fc40e8a9ce1fa1f3c29b416eef24a73c6bce3036839c0bbfe1b8dbd6d1bba3718b7ca5383
e829c9afbf refactor: replace sizeof(a)/sizeof(a[0]) by std::size (C++17) (Sebastian Falbesoner)
365539c846 refactor: init vectors via std::{begin,end} to avoid pointer arithmetic (Sebastian Falbesoner)
63d4ee1968 refactor: iterate arrays via C++11 range-based for loops if idx is not needed (Sebastian Falbesoner)
Pull request description:
This refactoring PR picks up the idea of #19626 and replaces all occurences of `sizeof(x)/sizeof(x[0])` (or `sizeof(x)/sizeof(*x)`, respectively) with the now-available C++17 [`std::size`](https://en.cppreference.com/w/cpp/iterator/size) (as [suggested by sipa](https://github.com/bitcoin/bitcoin/pull/19626#issuecomment-666487228)), making the macro `ARRAYLEN` obsolete.
As preparation for this, two other changes are done to eliminate `sizeof(x)/sizeof(x[0])` usage:
* all places where arrays are iterated via an index are changed to use C++11 range-based for loops If the index' only purpose is to access the array element (as [suggested by MarcoFalke](https://github.com/bitcoin/bitcoin/pull/19626#discussion_r463404541)).
* `std::vector` initializations are done via `std::begin` and `std::end` rather than using pointer arithmetic to calculate the end (also [suggested by MarcoFalke](https://github.com/bitcoin/bitcoin/pull/20429#discussion_r567418821)).
ACKs for top commit:
practicalswift:
cr ACK e829c9afbf: patch looks correct
fanquake:
ACK e829c9afbf
MarcoFalke:
review ACK e829c9afbf 🌩
Tree-SHA512: b01d32c04b9e04d562b7717cae00a651ec9a718645047a90761be6959e0cc2adbd67494e058fe894641076711bb09c3b47a047d0275c736f0b2218e1ce0d193d
The benchmark BlockToJsonVerbose only tests generating (and destroying)
the JSON data structure, but serializing into a string is also a
performance critical aspect of the RPC calls.
Also, use ankerl::nanobench::doNotOptimizeAway to make sure the compiler
can't optimize the result of the calls away.
53e716ea11 [refactor] improve style for touched code (gzhao408)
174cb5330a [refactor] const ATMPArgs and non-const Workspace (gzhao408)
f82baf0762 [refactor] return MempoolAcceptResult (gzhao408)
9db10a5506 [refactor] clean up logic in testmempoolaccept (gzhao408)
Pull request description:
This is the first 4 commits of #20833, and does refactoring only. It should be relatively simple to review, and offers a few nice things:
- It makes accessing values that don't make sense (e.g. fee) when the tx is invalid an error.
- Returning `MempoolAcceptResult` from ATMP makes the interface cleaner. The caller can get a const instead of passing in a mutable "out" param.
- We don't have to be iterating through a bunch of lists for package validation, we can just return a `std::vector<MempoolAcceptResult>`.
- We don't have to refactor all ATMP call sites again if/when we want to return more stuff from it.
ACKs for top commit:
MarcoFalke:
ACK 53e716ea11💿
jnewbery:
Code review ACK 53e716ea11
ariard:
Code Review ACK 53e716e, I did tweak a bit the touched paths to see if we had good test coverage. Didn't find holes.
Tree-SHA512: fa6ec324a08ad9e6e55948615cda324cba176255708bf0a0a0f37cedb7a75311aa334ac6f223be7d8df3c7379502b1081102b9589f9a9afa1713ad3d9ab3c24f
This creates a cleaner interface with ATMP, allows us to make results const,
and makes accessing values that don't make sense (e.g. fee when tx is
invalid) an error.
fa29272459 Remove redundant MakeUCharSpan wrappers (MarcoFalke)
faf4aa2f47 Remove CDataStream::Init in favor of C++11 member initialization (MarcoFalke)
fada14b948 Treat CDataStream bytes as uint8_t (MarcoFalke)
fa8bdb048e refactor: Drop CDataStream constructors in favor of one taking a Span of bytes (MarcoFalke)
faa96f841f Remove unused CDataStream methods (MarcoFalke)
Pull request description:
Using `uint8_t` for raw bytes has a style benefit:
* The signedness is clear from reading the code, as it does not depend on the architecture
Other clean-ups in this pull include:
* Remove unused methods
* Constructor is simplified with `Span`
* Remove `Init()` member in favor of C++11 member initialization
ACKs for top commit:
laanwj:
code review ACK fa29272459
theStack:
ACK fa29272459🍾
Tree-SHA512: 931ee28bd99843d7e894b48e90e1187ffb0278677c267044b3c0c255069d9bbd9298ab2e539b1002a30b543d240450eaec718ef4ee95a7fd4be0a295e926343f
5d4597666d Rewrite OutputGroups to be clearer and to use scriptPubKeys (Andrew Chow)
f6b3052739 Explicitly filter out partial groups when we don't want them (Andrew Chow)
416d74fb16 Move OutputGroup positive only filtering into Insert (Andrew Chow)
d895e98b59 Move EligibleForSpending into GroupOutputs (Andrew Chow)
99b399aba5 Move fee setting of OutputGroup to Insert (Andrew Chow)
6148a8acda Move GroupOutputs into SelectCoinsMinConf (Andrew Chow)
2acad03657 Remove OutputGroup non-default constructors (Andrew Chow)
Pull request description:
Even after #17458, we still deal with setting fees of an `OutputGroup` and filtering the `OutputGroup` outside of the struct. We currently make all of the `OutputGroup`s in `SelectCoins` and then copy and modify them within each `SelectCoinsMinConf` scenario. This PR changes this to constructing the `OutputGroup`s within the `SelectCoinsMinConf` so that the scenario can be taken into account during the group construction. Furthermore, setting of fees and filtering for effective value is moved into `OutputGroup::Insert` itself so that we don't add undesirable outputs to an `OutputGroup` rather than deleting them afterwards.
To facilitate fee calculation and effective value filtering during `OutputGroup::Insert`, `OutputGroup` now takes the feerates in its constructor and computes the fees and effective value for each output during `Insert`.
While removing `OutputGroup`s in accordance with the `CoinEligibilityFilter` still requires creating the `OutputGroup`s first, we can do that within the function that makes them - `GroupOutput`s.
ACKs for top commit:
Xekyo:
Code review ACK: https://github.com/bitcoin/bitcoin/pull/20040/commits/5d4597666d589e39354e0d8dd5b2afbe1a5d7d8e
fjahr:
Code review ACK 5d4597666d
meshcollider:
Light utACK 5d4597666d
Tree-SHA512: 35965b6d49a87f4ebb366ec4f00aafaaf78e9282481ae2c9682b515a3a9f2cbcd3cd6e202fee29489d48fe7f3a7cede4270796f5e72bbaff76da647138fb3059
bb6fcc75d1 refactor: Drop boost::thread stuff in CCheckQueue (Hennadii Stepanov)
6784ac471b bench: Use CCheckQueue local thread pool (Hennadii Stepanov)
dba30695fc test: Use CCheckQueue local thread pool (Hennadii Stepanov)
01511776ac Add local thread pool to CCheckQueue (Hennadii Stepanov)
0ef938685b refactor: Use member initializers in CCheckQueue (Hennadii Stepanov)
Pull request description:
This PR:
- gets rid of `boost::thread_group` in the `CCheckQueue` class
- allows thread safety annotation usage in the `CCheckQueue` class
- is alternative to #14464 (https://github.com/bitcoin/bitcoin/pull/18710#issuecomment-616618525, https://github.com/bitcoin/bitcoin/pull/18710#issuecomment-617291612)
Also, with this PR (I hope) it could be easier to resurrect a bunch of brilliant ideas from #9938.
Related: #17307
ACKs for top commit:
laanwj:
Code review ACK bb6fcc75d1
LarryRuane:
ACK bb6fcc75d1
jonatack:
Code review ACK bb6fcc75d1 and verified rebase to master builds cleanly with unit/functional tests green
Tree-SHA512: fddeb720d5a391b48bb4c6fa58ed34ccc3f57862fdb8e641745c021841c8340e35c5126338271446cbd98f40bd5484f27926aa6c3e76fa478ba1efafe72e73c1
