Merge 7edaf8b64cb2d59ada22042fee62a417e52368b8 into 5f4422d68dc3530c353af1f87499de1c864b60ad

3301d2cbe8c3b76c97285d75fa59637cb6952d0b qa: Wait for txindex to avoid race condition (Hodlinator)
9bfb0d75ba10591cc6c9620f9fd1ecc0e55e7a48 qa: Remove unnecessary -txindex args (Hodlinator)
7ac281c19cd3d11f316dbbb3308eabf1ad4f26d6 qa: Add missing coverage of corrupt indexes (Hodlinator)

Pull request description:

  - Add synchronization in 3 places where if the Transaction Index happens to be slow, we get rare test failures when querying it for transactions (one such case experienced on Windows, prompting investigation).
  - Remove unnecessary TxIndex initialization in some tests.
  - Add some test coverage where TxIndex aspect could be tested in feature_init.py.

ACKs for top commit:
  fjahr:
    re-ACK 3301d2cbe8c3b76c97285d75fa59637cb6952d0b
  mzumsande:
    Code Review ACK 3301d2cbe8c3b76c97285d75fa59637cb6952d0b
  furszy:
    Code review ACK 3301d2cbe8c3b76c97285d75fa59637cb6952d0b
  Prabhat1308:
    Concept ACK [`3301d2c`](3301d2cbe8)

Tree-SHA512: 7c2019e38455f344856aaf6b381faafbd88d53dc88d13309deb718c1dcfbee4ccca7c7f1b66917395503a6f94c3b216a007ad432cc8b93d0309db9805f38d602

src/bench/CMakeLists.txt

@@ -18,6 +18,7 @@ add_executable(bench_bitcoin
   checkblockindex.cpp
   checkqueue.cpp
   cluster_linearize.cpp
+  connectblock.cpp
   crypto_hash.cpp
   descriptors.cpp
   disconnected_transactions.cpp

src/bench/connectblock.cpp

@@ -0,0 +1,137 @@
+// Copyright (c) 2025 The Bitcoin Core developers
+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#include <addresstype.h>
+#include <bench/bench.h>
+#include <interfaces/chain.h>
+#include <kernel/cs_main.h>
+#include <script/interpreter.h>
+#include <sync.h>
+#include <test/util/setup_common.h>
+#include <validation.h>
+
+#include <cassert>
+#include <vector>
+
+/*
+ * Creates a test block containing transactions with the following properties:
+ * - Each transaction has the same number of inputs and outputs
+ * - All Taproot inputs use simple key path spends (no script path spends)
+ * - All signatures use SIGHASH_ALL (default sighash)
+ * - Each transaction spends all outputs from the previous transaction
+ */
+CBlock CreateTestBlock(
+    TestChain100Setup& test_setup,
+    const std::vector<CKey>& keys,
+    const std::vector<CTxOut>& outputs,
+    int num_txs = 1000)
+{
+    Chainstate& chainstate{test_setup.m_node.chainman->ActiveChainstate()};
+
+    const WitnessV1Taproot coinbase_taproot{XOnlyPubKey(test_setup.coinbaseKey.GetPubKey())};
+
+    // Create the outputs that will be spent in the first transaction of the test block
+    // Doing this in a separate block excludes the validation of its inputs from the benchmark
+    auto& coinbase_to_spend{test_setup.m_coinbase_txns[0]};
+    const auto [first_tx, _]{test_setup.CreateValidTransaction(
+        {coinbase_to_spend},
+        {COutPoint(coinbase_to_spend->GetHash(), 0)},
+        chainstate.m_chain.Height() + 1, keys, outputs, {}, {})};
+    const CScript coinbase_spk{GetScriptForDestination(coinbase_taproot)};
+    test_setup.CreateAndProcessBlock({first_tx}, coinbase_spk, &chainstate);
+
+    std::vector<CMutableTransaction> txs;
+    txs.reserve(num_txs);
+    CTransactionRef tx_to_spend{MakeTransactionRef(first_tx)};
+    for (int i{0}; i < num_txs; i++) {
+        std::vector<COutPoint> inputs;
+        inputs.reserve(outputs.size());
+
+        for (size_t j{0}; j < outputs.size(); j++) {
+            inputs.emplace_back(tx_to_spend->GetHash(), j);
+        }
+
+        const auto [taproot_tx, _]{test_setup.CreateValidTransaction(
+            {tx_to_spend}, inputs, chainstate.m_chain.Height() + 1, keys, outputs, {}, {})};
+        txs.emplace_back(taproot_tx);
+        tx_to_spend = MakeTransactionRef(taproot_tx);
+    }
+
+    // Coinbase output can use any output type as it is not spent and will not change the benchmark
+    return test_setup.CreateBlock(txs, coinbase_spk, chainstate);
+}
+
+/*
+ * Creates key pairs and corresponding outputs for the benchmark transactions.
+ * - For Schnorr signatures: Creates simple key path spendable outputs
+ * - For Ecdsa signatures: Creates P2WPKH (native SegWit v0) outputs
+ * - All outputs have value of 1 BTC
+ */
+std::pair<std::vector<CKey>, std::vector<CTxOut>> CreateKeysAndOutputs(const CKey& coinbaseKey, size_t num_schnorr, size_t num_ecdsa)
+{
+    std::vector<CKey> keys{coinbaseKey};
+    keys.reserve(num_schnorr + num_ecdsa + 1);
+
+    std::vector<CTxOut> outputs;
+    outputs.reserve(num_schnorr + num_ecdsa);
+
+    for (size_t i{0}; i < num_ecdsa; ++i) {
+        keys.emplace_back(GenerateRandomKey());
+        outputs.emplace_back(COIN, GetScriptForDestination(WitnessV0KeyHash{keys.back().GetPubKey()}));
+    }
+
+    for (size_t i{0}; i < num_schnorr; ++i) {
+        keys.emplace_back(GenerateRandomKey());
+        outputs.emplace_back(COIN, GetScriptForDestination(WitnessV1Taproot{XOnlyPubKey(keys.back().GetPubKey())}));
+    }
+
+    return {keys, outputs};
+}
+
+void BenchmarkConnectBlock(benchmark::Bench& bench, std::vector<CKey>& keys, std::vector<CTxOut>& outputs, TestChain100Setup& test_setup)
+{
+    const auto& test_block{CreateTestBlock(test_setup, keys, outputs)};
+    bench.unit("block").run([&] {
+        LOCK(cs_main);
+        auto& chainman{test_setup.m_node.chainman};
+        auto& chainstate{chainman->ActiveChainstate()};
+        BlockValidationState test_block_state;
+        auto* pindex{chainman->m_blockman.AddToBlockIndex(test_block, chainman->m_best_header)}; // Doing this here doesn't impact the benchmark
+        CCoinsViewCache viewNew{&chainstate.CoinsTip()};
+
+        assert(chainstate.ConnectBlock(test_block, test_block_state, pindex, viewNew));
+    });
+}
+
+static void ConnectBlockAllSchnorr(benchmark::Bench& bench)
+{
+    const auto test_setup{MakeNoLogFileContext<TestChain100Setup>()};
+    auto [keys, outputs]{CreateKeysAndOutputs(test_setup->coinbaseKey, /*num_schnorr=*/4, /*num_ecdsa=*/0)};
+    BenchmarkConnectBlock(bench, keys, outputs, *test_setup);
+}
+
+/**
+ * This benchmark is expected to be slower than the AllSchnorr or Ecdsa benchmark
+ * because it uses transactions with both Schnorr and Ecdsa signatures
+ * which requires the transaction to be hashed multiple times for
+ * the different signature algorithms
+ */
+static void ConnectBlockMixedEcdsaSchnorr(benchmark::Bench& bench)
+{
+    const auto test_setup{MakeNoLogFileContext<TestChain100Setup>()};
+    // Blocks in range 848000 to 868000 have a roughly 20 to 80 ratio of schnorr to ecdsa inputs
+    auto [keys, outputs]{CreateKeysAndOutputs(test_setup->coinbaseKey, /*num_schnorr=*/1, /*num_ecdsa=*/4)};
+    BenchmarkConnectBlock(bench, keys, outputs, *test_setup);
+}
+
+static void ConnectBlockAllEcdsa(benchmark::Bench& bench)
+{
+    const auto test_setup{MakeNoLogFileContext<TestChain100Setup>()};
+    auto [keys, outputs]{CreateKeysAndOutputs(test_setup->coinbaseKey, /*num_schnorr=*/0, /*num_ecdsa=*/4)};
+    BenchmarkConnectBlock(bench, keys, outputs, *test_setup);
+}
+
+BENCHMARK(ConnectBlockAllSchnorr, benchmark::PriorityLevel::HIGH);
+BENCHMARK(ConnectBlockMixedEcdsaSchnorr, benchmark::PriorityLevel::HIGH);
+BENCHMARK(ConnectBlockAllEcdsa, benchmark::PriorityLevel::HIGH);

test/functional/feature_init.py

@@ -88,7 +88,7 @@ class InitTest(BitcoinTestFramework):
 
         args = ['-txindex=1', '-blockfilterindex=1', '-coinstatsindex=1']
         for terminate_line in lines_to_terminate_after:
-            self.log.info(f"Starting node and will exit after line {terminate_line}")
+            self.log.info(f"Starting node and will terminate after line {terminate_line}")
             with node.busy_wait_for_debug_log([terminate_line]):
                 if platform.system() == 'Windows':
                     # CREATE_NEW_PROCESS_GROUP is required in order to be able
@@ -108,12 +108,22 @@ class InitTest(BitcoinTestFramework):
             'blocks/index/*.ldb': 'Error opening block database.',
             'chainstate/*.ldb': 'Error opening coins database.',
             'blocks/blk*.dat': 'Error loading block database.',
+            'indexes/txindex/MANIFEST*': 'LevelDB error: Corruption: CURRENT points to a non-existent file',
+            # Removing these files does not result in a startup error:
+            # 'indexes/blockfilter/basic/*.dat', 'indexes/blockfilter/basic/db/*.*', 'indexes/coinstats/db/*.*',
+            # 'indexes/txindex/*.log', 'indexes/txindex/CURRENT', 'indexes/txindex/LOCK'
         }
 
         files_to_perturb = {
             'blocks/index/*.ldb': 'Error loading block database.',
             'chainstate/*.ldb': 'Error opening coins database.',
             'blocks/blk*.dat': 'Corrupted block database detected.',
+            'indexes/blockfilter/basic/db/*.*': 'LevelDB error: Corruption',
+            'indexes/coinstats/db/*.*': 'LevelDB error: Corruption',
+            'indexes/txindex/*.log': 'LevelDB error: Corruption',
+            'indexes/txindex/CURRENT': 'LevelDB error: Corruption',
+            # Perturbing these files does not result in a startup error:
+            # 'indexes/blockfilter/basic/*.dat', 'indexes/txindex/MANIFEST*', 'indexes/txindex/LOCK'
         }
 
         for file_patt, err_fragment in files_to_delete.items():
@@ -135,9 +145,10 @@ class InitTest(BitcoinTestFramework):
             self.stop_node(0)
 
         self.log.info("Test startup errors after perturbing certain essential files")
+        dirs = ["blocks", "chainstate", "indexes"]
         for file_patt, err_fragment in files_to_perturb.items():
-            shutil.copytree(node.chain_path / "blocks", node.chain_path / "blocks_bak")
-            shutil.copytree(node.chain_path / "chainstate", node.chain_path / "chainstate_bak")
+            for dir in dirs:
+                shutil.copytree(node.chain_path / dir, node.chain_path / f"{dir}_bak")
             target_files = list(node.chain_path.glob(file_patt))
 
             for target_file in target_files:
@@ -151,10 +162,9 @@ class InitTest(BitcoinTestFramework):
 
             start_expecting_error(err_fragment)
 
-            shutil.rmtree(node.chain_path / "blocks")
-            shutil.rmtree(node.chain_path / "chainstate")
-            shutil.move(node.chain_path / "blocks_bak", node.chain_path / "blocks")
-            shutil.move(node.chain_path / "chainstate_bak", node.chain_path / "chainstate")
+            for dir in dirs:
+                shutil.rmtree(node.chain_path / dir)
+                shutil.move(node.chain_path / f"{dir}_bak", node.chain_path / dir)
 
     def init_pid_test(self):
         BITCOIN_PID_FILENAME_CUSTOM = "my_fancy_bitcoin_pid_file.foobar"

test/functional/mempool_accept.py

@@ -45,6 +45,7 @@ from test_framework.util import (
     assert_equal,
     assert_greater_than,
     assert_raises_rpc_error,
+    sync_txindex,
 )
 from test_framework.wallet import MiniWallet
 from test_framework.wallet_util import generate_keypair
@@ -270,6 +271,7 @@ class MempoolAcceptanceTest(BitcoinTestFramework):
 
         self.log.info('A coinbase transaction')
         # Pick the input of the first tx we created, so it has to be a coinbase tx
+        sync_txindex(self, node)
         raw_tx_coinbase_spent = node.getrawtransaction(txid=node.decoderawtransaction(hexstring=raw_tx_in_block)['vin'][0]['txid'])
         tx = tx_from_hex(raw_tx_coinbase_spent)
         self.check_mempool_result(

test/functional/rpc_rawtransaction.py

@@ -34,6 +34,7 @@ from test_framework.util import (
     assert_equal,
     assert_greater_than,
     assert_raises_rpc_error,
+    sync_txindex,
 )
 from test_framework.wallet import (
     getnewdestination,
@@ -70,7 +71,7 @@ class RawTransactionsTest(BitcoinTestFramework):
         self.num_nodes = 3
         self.extra_args = [
             ["-txindex"],
-            ["-txindex"],
+            [],
             ["-fastprune", "-prune=1"],
         ]
         # whitelist peers to speed up tx relay / mempool sync
@@ -109,6 +110,7 @@ class RawTransactionsTest(BitcoinTestFramework):
             self.log.info(f"Test getrawtransaction {'with' if n == 0 else 'without'} -txindex")
 
             if n == 0:
+                sync_txindex(self, self.nodes[n])
                 # With -txindex.
                 # 1. valid parameters - only supply txid
                 assert_equal(self.nodes[n].getrawtransaction(txId), tx['hex'])

test/functional/rpc_txoutproof.py

@@ -12,6 +12,7 @@ from test_framework.test_framework import BitcoinTestFramework
 from test_framework.util import (
     assert_equal,
     assert_raises_rpc_error,
+    sync_txindex,
 )
 from test_framework.wallet import MiniWallet
 
@@ -77,6 +78,7 @@ class MerkleBlockTest(BitcoinTestFramework):
         assert_equal(sorted(self.nodes[0].verifytxoutproof(self.nodes[0].gettxoutproof([txid1, txid2]))), sorted(txlist))
         assert_equal(sorted(self.nodes[0].verifytxoutproof(self.nodes[0].gettxoutproof([txid2, txid1]))), sorted(txlist))
         # We can always get a proof if we have a -txindex
+        sync_txindex(self, self.nodes[1])
         assert_equal(self.nodes[0].verifytxoutproof(self.nodes[1].gettxoutproof([txid_spent])), [txid_spent])
         # We can't get a proof if we specify transactions from different blocks
         assert_raises_rpc_error(-5, "Not all transactions found in specified or retrieved block", self.nodes[0].gettxoutproof, [txid1, txid3])

test/functional/test_framework/util.py

@@ -592,3 +592,10 @@ def find_vout_for_address(node, txid, addr):
         if addr == tx["vout"][i]["scriptPubKey"]["address"]:
             return i
     raise RuntimeError("Vout not found for address: txid=%s, addr=%s" % (txid, addr))
+
+
+def sync_txindex(test_framework, node):
+    test_framework.log.debug("Waiting for node txindex to sync")
+    sync_start = int(time.time())
+    test_framework.wait_until(lambda: node.getindexinfo("txindex")["txindex"]["synced"])
+    test_framework.log.debug(f"Synced in {time.time() - sync_start} seconds")

test/functional/wallet_avoid_mixing_output_types.py

@@ -117,7 +117,6 @@ class AddressInputTypeGrouping(BitcoinTestFramework):
         self.extra_args = [
             [
                 "-addresstype=bech32",
-                "-txindex",
             ],
             [
                 "-addresstype=p2sh-segwit",