core/rawdb: don't decode the full block body in ReadTransaction (#32027)

Reading a single transaction out of a block shouldn't need decoding the
entire body

---------

Co-authored-by: Felix Lange <fjl@twurst.com>
Co-authored-by: Gary Rong <garyrong0905@gmail.com>

core/rawdb/accessors_indexes.go

@@ -24,6 +24,7 @@ import (
 
 	"github.com/ethereum/go-ethereum/common"
 	"github.com/ethereum/go-ethereum/core/types"
+	"github.com/ethereum/go-ethereum/crypto"
 	"github.com/ethereum/go-ethereum/ethdb"
 	"github.com/ethereum/go-ethereum/log"
 	"github.com/ethereum/go-ethereum/params"
@@ -128,6 +129,46 @@ func DeleteAllTxLookupEntries(db ethdb.KeyValueStore, condition func(common.Hash
 	}
 }
 
+// findTxInBlockBody traverses the given RLP-encoded block body, searching for
+// the transaction specified by its hash.
+func findTxInBlockBody(blockbody rlp.RawValue, target common.Hash) (*types.Transaction, uint64, error) {
+	txnListRLP, _, err := rlp.SplitList(blockbody)
+	if err != nil {
+		return nil, 0, err
+	}
+	iter, err := rlp.NewListIterator(txnListRLP)
+	if err != nil {
+		return nil, 0, err
+	}
+	txIndex := uint64(0)
+	for iter.Next() {
+		if iter.Err() != nil {
+			return nil, 0, err
+		}
+		// The preimage for the hash calculation of legacy transactions
+		// is just their RLP encoding. For typed (EIP-2718) transactions,
+		// which are encoded as byte arrays, the preimage is the content of
+		// the byte array, so trim their prefix here.
+		txRLP := iter.Value()
+		kind, txHashPayload, _, err := rlp.Split(txRLP)
+		if err != nil {
+			return nil, 0, err
+		}
+		if kind == rlp.List { // Legacy transaction
+			txHashPayload = txRLP
+		}
+		if crypto.Keccak256Hash(txHashPayload) == target {
+			var tx types.Transaction
+			if err := rlp.DecodeBytes(txRLP, &tx); err != nil {
+				return nil, 0, err
+			}
+			return &tx, txIndex, nil
+		}
+		txIndex++
+	}
+	return nil, 0, errors.New("transaction not found")
+}
+
 // ReadTransaction retrieves a specific transaction from the database, along with
 // its added positional metadata.
 func ReadTransaction(db ethdb.Reader, hash common.Hash) (*types.Transaction, common.Hash, uint64, uint64) {
@@ -139,18 +180,17 @@ func ReadTransaction(db ethdb.Reader, hash common.Hash) (*types.Transaction, com
 	if blockHash == (common.Hash{}) {
 		return nil, common.Hash{}, 0, 0
 	}
-	body := ReadBody(db, blockHash, *blockNumber)
+	bodyRLP := ReadBodyRLP(db, blockHash, *blockNumber)
-	if body == nil {
+	if bodyRLP == nil {
 		log.Error("Transaction referenced missing", "number", *blockNumber, "hash", blockHash)
 		return nil, common.Hash{}, 0, 0
 	}
-	for txIndex, tx := range body.Transactions {
+	tx, txIndex, err := findTxInBlockBody(bodyRLP, hash)
-		if tx.Hash() == hash {
+	if err != nil {
-			return tx, blockHash, *blockNumber, uint64(txIndex)
+		log.Error("Transaction not found", "number", *blockNumber, "hash", blockHash, "txhash", hash, "err", err)
-		}
+		return nil, common.Hash{}, 0, 0
 	}
-	log.Error("Transaction not found", "number", *blockNumber, "hash", blockHash, "txhash", hash)
+	return tx, blockHash, *blockNumber, txIndex
-	return nil, common.Hash{}, 0, 0
 }
 
 // ReadReceipt retrieves a specific transaction receipt from the database, along with

core/rawdb/accessors_indexes_test.go

@@ -20,11 +20,13 @@ import (
 	"math/big"
 	"testing"
 
+	"github.com/davecgh/go-spew/spew"
 	"github.com/ethereum/go-ethereum/common"
 	"github.com/ethereum/go-ethereum/core/types"
 	"github.com/ethereum/go-ethereum/ethdb"
 	"github.com/ethereum/go-ethereum/internal/blocktest"
 	"github.com/ethereum/go-ethereum/rlp"
+	"github.com/holiman/uint256"
 )
 
 var newTestHasher = blocktest.NewHasher
@@ -72,7 +74,15 @@ func TestLookupStorage(t *testing.T) {
 			tx1 := types.NewTransaction(1, common.BytesToAddress([]byte{0x11}), big.NewInt(111), 1111, big.NewInt(11111), []byte{0x11, 0x11, 0x11})
 			tx2 := types.NewTransaction(2, common.BytesToAddress([]byte{0x22}), big.NewInt(222), 2222, big.NewInt(22222), []byte{0x22, 0x22, 0x22})
 			tx3 := types.NewTransaction(3, common.BytesToAddress([]byte{0x33}), big.NewInt(333), 3333, big.NewInt(33333), []byte{0x33, 0x33, 0x33})
-			txs := []*types.Transaction{tx1, tx2, tx3}
+			tx4 := types.NewTx(&types.DynamicFeeTx{
+				To:        new(common.Address),
+				Nonce:     5,
+				Value:     big.NewInt(5),
+				Gas:       5,
+				GasTipCap: big.NewInt(55),
+				GasFeeCap: big.NewInt(1055),
+			})
+			txs := []*types.Transaction{tx1, tx2, tx3, tx4}
 
 			block := types.NewBlock(&types.Header{Number: big.NewInt(314)}, &types.Body{Transactions: txs}, nil, newTestHasher())
 
@@ -109,3 +119,103 @@ func TestLookupStorage(t *testing.T) {
 		})
 	}
 }
+
+func TestFindTxInBlockBody(t *testing.T) {
+	tx1 := types.NewTx(&types.LegacyTx{
+		Nonce:    1,
+		GasPrice: big.NewInt(1),
+		Gas:      1,
+		To:       new(common.Address),
+		Value:    big.NewInt(5),
+		Data:     []byte{0x11, 0x11, 0x11},
+	})
+	tx2 := types.NewTx(&types.AccessListTx{
+		Nonce:    1,
+		GasPrice: big.NewInt(1),
+		Gas:      1,
+		To:       new(common.Address),
+		Value:    big.NewInt(5),
+		Data:     []byte{0x11, 0x11, 0x11},
+		AccessList: []types.AccessTuple{
+			{
+				Address:     common.Address{0x1},
+				StorageKeys: []common.Hash{{0x1}, {0x2}},
+			},
+		},
+	})
+	tx3 := types.NewTx(&types.DynamicFeeTx{
+		Nonce:     1,
+		Gas:       1,
+		To:        new(common.Address),
+		Value:     big.NewInt(5),
+		Data:      []byte{0x11, 0x11, 0x11},
+		GasTipCap: big.NewInt(55),
+		GasFeeCap: big.NewInt(1055),
+		AccessList: []types.AccessTuple{
+			{
+				Address:     common.Address{0x1},
+				StorageKeys: []common.Hash{{0x1}, {0x2}},
+			},
+		},
+	})
+	tx4 := types.NewTx(&types.BlobTx{
+		Nonce:     1,
+		Gas:       1,
+		To:        common.Address{0x1},
+		Value:     uint256.NewInt(5),
+		Data:      []byte{0x11, 0x11, 0x11},
+		GasTipCap: uint256.NewInt(55),
+		GasFeeCap: uint256.NewInt(1055),
+		AccessList: []types.AccessTuple{
+			{
+				Address:     common.Address{0x1},
+				StorageKeys: []common.Hash{{0x1}, {0x2}},
+			},
+		},
+		BlobFeeCap: uint256.NewInt(1),
+		BlobHashes: []common.Hash{{0x1}, {0x2}},
+	})
+	tx5 := types.NewTx(&types.SetCodeTx{
+		Nonce:     1,
+		Gas:       1,
+		To:        common.Address{0x1},
+		Value:     uint256.NewInt(5),
+		Data:      []byte{0x11, 0x11, 0x11},
+		GasTipCap: uint256.NewInt(55),
+		GasFeeCap: uint256.NewInt(1055),
+		AccessList: []types.AccessTuple{
+			{
+				Address:     common.Address{0x1},
+				StorageKeys: []common.Hash{{0x1}, {0x2}},
+			},
+		},
+		AuthList: []types.SetCodeAuthorization{
+			{
+				ChainID: uint256.Int{1},
+				Address: common.Address{0x1},
+			},
+		},
+	})
+
+	txs := []*types.Transaction{tx1, tx2, tx3, tx4, tx5}
+
+	block := types.NewBlock(&types.Header{Number: big.NewInt(314)}, &types.Body{Transactions: txs}, nil, newTestHasher())
+	db := NewMemoryDatabase()
+	WriteBlock(db, block)
+
+	rlp := ReadBodyRLP(db, block.Hash(), block.NumberU64())
+	for i := 0; i < len(txs); i++ {
+		tx, txIndex, err := findTxInBlockBody(rlp, txs[i].Hash())
+		if err != nil {
+			t.Fatalf("Failed to retrieve tx, err: %v", err)
+		}
+		if txIndex != uint64(i) {
+			t.Fatalf("Unexpected transaction index, want: %d, got: %d", i, txIndex)
+		}
+		if tx.Hash() != txs[i].Hash() {
+			want := spew.Sdump(txs[i])
+			got := spew.Sdump(tx)
+			t.Fatalf("Unexpected transaction, want: %s, got: %s", want, got)
+		}
+	}
+}