core/state: track all accounts in canon state

This change introduces a global, per-state cache that keeps
account data in the canon state.

core/state/dump.go

@@ -21,6 +21,7 @@ import (
 	"fmt"
 
 	"github.com/ethereum/go-ethereum/common"
+	"github.com/ethereum/go-ethereum/rlp"
 )
 
 type DumpAccount struct {
@@ -46,20 +47,22 @@ func (self *StateDB) RawDump() Dump {
 	it := self.trie.Iterator()
 	for it.Next() {
 		addr := self.trie.GetKey(it.Key)
-		stateObject, err := DecodeObject(common.BytesToAddress(addr), self.db, it.Value)
-		if err != nil {
+		var data Account
+		if err := rlp.DecodeBytes(it.Value, &data); err != nil {
 			panic(err)
 		}
+
+		obj := NewObject(common.BytesToAddress(addr), data)
 		account := DumpAccount{
-			Balance:  stateObject.Balance().String(),
-			Nonce:    stateObject.Nonce(),
-			Root:     common.Bytes2Hex(stateObject.data.Root[:]),
-			CodeHash: common.Bytes2Hex(stateObject.CodeHash()),
-			Code:     common.Bytes2Hex(stateObject.Code(self.db)),
+			Balance:  data.Balance.String(),
+			Nonce:    data.Nonce,
+			Root:     common.Bytes2Hex(data.Root[:]),
+			CodeHash: common.Bytes2Hex(data.CodeHash),
+			Code:     common.Bytes2Hex(obj.Code(self.db)),
 			Storage:  make(map[string]string),
 		}
-		stateObject.initTrie(self.db)
-		storageIt := stateObject.trie.Iterator()
+		obj.initTrie(self.db)
+		storageIt := obj.trie.Iterator()
 		for storageIt.Next() {
 			account.Storage[common.Bytes2Hex(self.trie.GetKey(storageIt.Key))] = common.Bytes2Hex(storageIt.Value)
 		}

core/state/state_object.go

@@ -65,7 +65,7 @@ func (self Storage) Copy() Storage {
 // Finally, call CommitTrie to write the modified storage trie into a database.
 type StateObject struct {
 	address common.Address // Ethereum address of this account
-	data    accountData
+	data    Account
 
 	// DB error.
 	// State objects are used by the consensus core and VM which are
@@ -87,23 +87,29 @@ type StateObject struct {
 	deleted bool
 }
 
-// accountData is the Ethereum consensus representation of accounts.
+// Account is the Ethereum consensus representation of accounts.
 // These objects are stored in the main account trie.
-type accountData struct {
+type Account struct {
 	Nonce    uint64
 	Balance  *big.Int
 	Root     common.Hash // merkle root of the storage trie
 	CodeHash []byte
 }
 
-func NewStateObject(address common.Address, db trie.Database) *StateObject {
-	object := &StateObject{
-		address: address,
-		data:    accountData{Balance: new(big.Int), CodeHash: emptyCodeHash},
-		storage: make(Storage),
-		dirty:   true,
+// NewObject creates a state object.
+func NewObject(address common.Address, data Account) *StateObject {
+	if data.Balance == nil {
+		data.Balance = new(big.Int)
 	}
-	return object
+	if data.CodeHash == nil {
+		data.CodeHash = emptyCodeHash
+	}
+	return &StateObject{address: address, data: data, storage: make(Storage)}
+}
+
+// EncodeRLP implements rlp.Encoder.
+func (c *StateObject) EncodeRLP(w io.Writer) error {
+	return rlp.Encode(w, c.data)
 }
 
 // setError remembers the first non-nil error it is called with.
@@ -224,8 +230,7 @@ func (c *StateObject) SetBalance(amount *big.Int) {
 func (c *StateObject) ReturnGas(gas, price *big.Int) {}
 
 func (self *StateObject) Copy(db trie.Database) *StateObject {
-	stateObject := NewStateObject(self.address, db)
-	stateObject.data = self.data
+	stateObject := NewObject(self.address, self.data)
 	stateObject.data.Balance.Set(self.data.Balance)
 	stateObject.trie = self.trie
 	stateObject.code = self.code
@@ -233,7 +238,6 @@ func (self *StateObject) Copy(db trie.Database) *StateObject {
 	stateObject.remove = self.remove
 	stateObject.dirty = self.dirty
 	stateObject.deleted = self.deleted
-
 	return stateObject
 }
 
@@ -307,15 +311,3 @@ func (self *StateObject) ForEachStorage(cb func(key, value common.Hash) bool) {
 		}
 	}
 }
-
-// EncodeRLP implements rlp.Encoder.
-func (c *StateObject) EncodeRLP(w io.Writer) error {
-	return rlp.Encode(w, c.data)
-}
-
-// DecodeObject decodes an RLP-encoded state object.
-func DecodeObject(address common.Address, db trie.Database, data []byte) (*StateObject, error) {
-	obj := StateObject{address: address, storage: make(Storage)}
-	err := rlp.DecodeBytes(data, &obj.data)
-	return &obj, err
-}

core/state/statedb.go

@@ -43,8 +43,14 @@ type StateDB struct {
 	db   ethdb.Database
 	trie *trie.SecureTrie
 
+	// This map caches canon state accounts.
+	all map[common.Address]Account
+
+	// This map holds 'live' objects, which will get modified
+	// while processing a state transition.
 	stateObjects map[common.Address]*StateObject
 
+	// The refund counter, also used by state transitioning.
 	refund *big.Int
 
 	thash, bhash common.Hash
@@ -62,6 +68,7 @@ func New(root common.Hash, db ethdb.Database) (*StateDB, error) {
 	return &StateDB{
 		db:           db,
 		trie:         tr,
+		all:          make(map[common.Address]Account),
 		stateObjects: make(map[common.Address]*StateObject),
 		refund:       new(big.Int),
 		logs:         make(map[common.Hash]vm.Logs),
@@ -74,15 +81,19 @@ func (self *StateDB) Reset(root common.Hash) error {
 	var (
 		err error
 		tr  = self.trie
+		all = self.all
 	)
 	if self.trie.Hash() != root {
+		// The root has changed, invalidate canon state.
 		if tr, err = trie.NewSecure(root, self.db); err != nil {
 			return err
 		}
+		all = make(map[common.Address]Account)
 	}
 	*self = StateDB{
 		db:           self.db,
 		trie:         tr,
+		all:          all,
 		stateObjects: make(map[common.Address]*StateObject),
 		refund:       new(big.Int),
 		logs:         make(map[common.Hash]vm.Logs),
@@ -244,26 +255,37 @@ func (self *StateDB) DeleteStateObject(stateObject *StateObject) {
 
 // Retrieve a state object given my the address. Returns nil if not found.
 func (self *StateDB) GetStateObject(addr common.Address) (stateObject *StateObject) {
-	stateObject = self.stateObjects[addr]
-	if stateObject != nil {
-		if stateObject.deleted {
-			stateObject = nil
-		}
-
-		return stateObject
-	}
-
-	data := self.trie.Get(addr[:])
-	if len(data) == 0 {
+	// Prefer 'live' objects.
+	if obj := self.stateObjects[addr]; obj != nil {
+		if obj.deleted {
 			return nil
 		}
-	stateObject, err := DecodeObject(addr, self.db, data)
-	if err != nil {
+		return obj
+	}
+
+	// Use cached account data from the canon state if possible.
+	if data, ok := self.all[addr]; ok {
+		return NewObject(addr, data)
+	}
+
+	// Load the object from the database.
+	enc := self.trie.Get(addr[:])
+	if len(enc) == 0 {
+		return nil
+	}
+	var data Account
+	if err := rlp.DecodeBytes(enc, &data); err != nil {
 		glog.Errorf("can't decode object at %x: %v", addr[:], err)
 		return nil
 	}
-	self.SetStateObject(stateObject)
-	return stateObject
+	// Update the all cache. Content in DB always corresponds
+	// to the current head state so this is ok to do here.
+	// The object we just loaded has no storage trie and code yet.
+	self.all[addr] = data
+	// Insert into the live set.
+	obj := NewObject(addr, data)
+	self.SetStateObject(obj)
+	return obj
 }
 
 func (self *StateDB) SetStateObject(object *StateObject) {
@@ -285,15 +307,14 @@ func (self *StateDB) newStateObject(addr common.Address) *StateObject {
 	if glog.V(logger.Core) {
 		glog.Infof("(+) %x\n", addr)
 	}
-
-	stateObject := NewStateObject(addr, self.db)
-	stateObject.SetNonce(StartingNonce)
-	self.stateObjects[addr] = stateObject
-
-	return stateObject
+	obj := NewObject(addr, Account{})
+	obj.dirty = true
+	obj.SetNonce(StartingNonce)
+	self.stateObjects[addr] = obj
+	return obj
 }
 
-// Creates creates a new state object and takes ownership. This is different from "NewStateObject"
+// Creates creates a new state object and takes ownership.
 func (self *StateDB) CreateStateObject(addr common.Address) *StateObject {
 	// Get previous (if any)
 	so := self.GetStateObject(addr)
@@ -320,6 +341,7 @@ func (self *StateDB) Copy() *StateDB {
 	// ignore error - we assume state-to-be-copied always exists
 	state, _ := New(common.Hash{}, self.db)
 	state.trie = self.trie
+	state.all = self.all
 	for addr, stateObject := range self.stateObjects {
 		if stateObject.dirty {
 			state.stateObjects[addr] = stateObject.Copy(self.db)
@@ -340,6 +362,7 @@ func (self *StateDB) Copy() *StateDB {
 func (self *StateDB) Set(state *StateDB) {
 	self.trie = state.trie
 	self.stateObjects = state.stateObjects
+	self.all = state.all
 
 	self.refund = state.refund
 	self.logs = state.logs
@@ -407,11 +430,13 @@ func (s *StateDB) CommitBatch() (root common.Hash, batch ethdb.Batch) {
 func (s *StateDB) commit(dbw trie.DatabaseWriter) (common.Hash, error) {
 	s.refund = new(big.Int)
 
-	for _, stateObject := range s.stateObjects {
+	// Commit objects to the trie.
+	for addr, stateObject := range s.stateObjects {
 		if stateObject.remove {
 			// If the object has been removed, don't bother syncing it
 			// and just mark it for deletion in the trie.
 			s.DeleteStateObject(stateObject)
+			delete(s.all, addr)
 		} else {
 			// Write any contract code associated with the state object
 			if stateObject.code != nil {
@@ -426,10 +451,18 @@ func (s *StateDB) commit(dbw trie.DatabaseWriter) (common.Hash, error) {
 			}
 			// Update the object in the main account trie.
 			s.UpdateStateObject(stateObject)
+			s.all[addr] = stateObject.data
 		}
 		stateObject.dirty = false
 	}
-	return s.trie.CommitTo(dbw)
+
+	// Write trie changes.
+	root, err := s.trie.CommitTo(dbw)
+	if err != nil {
+		// Committing failed, any updates to the canon state are invalid.
+		s.all = make(map[common.Address]Account)
+	}
+	return root, err
 }
 
 func (self *StateDB) Refunds() *big.Int {

tests/util.go

@@ -104,15 +104,16 @@ func (self Log) Topics() [][]byte {
 }
 
 func StateObjectFromAccount(db ethdb.Database, addr string, account Account) *state.StateObject {
-	obj := state.NewStateObject(common.HexToAddress(addr), db)
-	obj.SetBalance(common.Big(account.Balance))
-
 	if common.IsHex(account.Code) {
 		account.Code = account.Code[2:]
 	}
-	obj.SetCode(common.Hex2Bytes(account.Code))
-	obj.SetNonce(common.Big(account.Nonce).Uint64())
-
+	code := common.Hex2Bytes(account.Code)
+	obj := state.NewObject(common.HexToAddress(addr), state.Account{
+		Balance:  common.Big(account.Balance),
+		CodeHash: crypto.Keccak256(code),
+		Nonce:    common.Big(account.Nonce).Uint64(),
+	})
+	obj.SetCode(code)
 	return obj
 }