core/state: rework dirty handling to avoid quadratic overhead

core/state/dump.go

@@ -53,7 +53,7 @@ func (self *StateDB) RawDump() Dump {
 			panic(err)
 		}
 
-		obj := newObject(nil, common.BytesToAddress(addr), data, nil)
+		obj := newObject(nil, common.BytesToAddress(addr), data)
 		account := DumpAccount{
 			Balance:  data.Balance.String(),
 			Nonce:    data.Nonce,

core/state/journal.go

@@ -24,9 +24,40 @@ import (
 
 type journalEntry interface {
 	undo(*StateDB)
+	getAccount() *common.Address
 }
 
-type journal []journalEntry
+type journal struct {
+	entries        []journalEntry
+	dirtyOverrides []common.Address
+}
+
+func (j *journal) append(entry journalEntry) {
+	j.entries = append(j.entries, entry)
+}
+
+func (j *journal) flatten() map[common.Address]struct{} {
+
+	dirtyObjects := make(map[common.Address]struct{})
+	for _, journalEntry := range j.entries {
+		if addr := journalEntry.getAccount(); addr != nil {
+			dirtyObjects[*addr] = struct{}{}
+		}
+	}
+	for _, addr := range j.dirtyOverrides {
+		dirtyObjects[addr] = struct{}{}
+	}
+	return dirtyObjects
+}
+
+// Length returns the number of journal entries in the journal
+func (j *journal) Length() int {
+	return len(j.entries)
+}
+
+func (j *journal) dirtyOverride(address common.Address) {
+	j.dirtyOverrides = append(j.dirtyOverrides, address)
+}
 
 type (
 	// Changes to the account trie.
@@ -82,10 +113,18 @@ func (ch createObjectChange) undo(s *StateDB) {
 	delete(s.stateObjectsDirty, *ch.account)
 }
 
+func (ch createObjectChange) getAccount() *common.Address {
+	return ch.account
+}
+
 func (ch resetObjectChange) undo(s *StateDB) {
 	s.setStateObject(ch.prev)
 }
 
+func (ch resetObjectChange) getAccount() *common.Address {
+	return nil
+}
+
 func (ch suicideChange) undo(s *StateDB) {
 	obj := s.getStateObject(*ch.account)
 	if obj != nil {
@@ -93,37 +132,52 @@ func (ch suicideChange) undo(s *StateDB) {
 		obj.setBalance(ch.prevbalance)
 	}
 }
+func (ch suicideChange) getAccount() *common.Address {
+	return ch.account
+}
 
 var ripemd = common.HexToAddress("0000000000000000000000000000000000000003")
 
 func (ch touchChange) undo(s *StateDB) {
-	if !ch.prev && *ch.account != ripemd {
-		s.getStateObject(*ch.account).touched = ch.prev
-		if !ch.prevDirty {
-			delete(s.stateObjectsDirty, *ch.account)
-		}
-	}
+}
+func (ch touchChange) getAccount() *common.Address {
+	return ch.account
 }
 
 func (ch balanceChange) undo(s *StateDB) {
 	s.getStateObject(*ch.account).setBalance(ch.prev)
 }
+func (ch balanceChange) getAccount() *common.Address {
+	return ch.account
+}
 
 func (ch nonceChange) undo(s *StateDB) {
 	s.getStateObject(*ch.account).setNonce(ch.prev)
 }
 
+func (ch nonceChange) getAccount() *common.Address {
+	return ch.account
+}
 func (ch codeChange) undo(s *StateDB) {
 	s.getStateObject(*ch.account).setCode(common.BytesToHash(ch.prevhash), ch.prevcode)
 }
+func (ch codeChange) getAccount() *common.Address {
+	return ch.account
+}
 
 func (ch storageChange) undo(s *StateDB) {
 	s.getStateObject(*ch.account).setState(ch.key, ch.prevalue)
 }
+func (ch storageChange) getAccount() *common.Address {
+	return ch.account
+}
 
 func (ch refundChange) undo(s *StateDB) {
 	s.refund = ch.prev
 }
+func (ch refundChange) getAccount() *common.Address {
+	return nil
+}
 
 func (ch addLogChange) undo(s *StateDB) {
 	logs := s.logs[ch.txhash]
@@ -134,7 +188,14 @@ func (ch addLogChange) undo(s *StateDB) {
 	}
 	s.logSize--
 }
+func (ch addLogChange) getAccount() *common.Address {
+	return nil
+}
 
 func (ch addPreimageChange) undo(s *StateDB) {
 	delete(s.preimages, ch.hash)
 }
+
+func (ch addPreimageChange) getAccount() *common.Address {
+	return nil
+}

core/state/state_object.go

@@ -85,9 +85,7 @@ type stateObject struct {
 	// during the "update" phase of the state transition.
 	dirtyCode bool // true if the code was updated
 	suicided  bool
-	touched   bool
 	deleted   bool
-	onDirty   func(addr common.Address) // Callback method to mark a state object newly dirty
 }
 
 // empty returns whether the account is considered empty.
@@ -105,7 +103,7 @@ type Account struct {
 }
 
 // newObject creates a state object.
-func newObject(db *StateDB, address common.Address, data Account, onDirty func(addr common.Address)) *stateObject {
+func newObject(db *StateDB, address common.Address, data Account) *stateObject {
 	if data.Balance == nil {
 		data.Balance = new(big.Int)
 	}
@@ -119,7 +117,6 @@ func newObject(db *StateDB, address common.Address, data Account, onDirty func(a
 		data:          data,
 		cachedStorage: make(Storage),
 		dirtyStorage:  make(Storage),
-		onDirty:       onDirty,
 	}
 }
 
@@ -137,23 +134,17 @@ func (self *stateObject) setError(err error) {
 
 func (self *stateObject) markSuicided() {
 	self.suicided = true
-	if self.onDirty != nil {
-		self.onDirty(self.Address())
-		self.onDirty = nil
-	}
 }
 
 func (c *stateObject) touch() {
-	c.db.journal = append(c.db.journal, touchChange{
-		account:   &c.address,
-		prev:      c.touched,
-		prevDirty: c.onDirty == nil,
+	c.db.journal.append(touchChange{
+		account: &c.address,
 	})
-	if c.onDirty != nil {
-		c.onDirty(c.Address())
-		c.onDirty = nil
+	if c.address == ripemd {
+		//Explicitly put it in the dirty-cache, which is otherwise
+		// generated from flattened journals
+		c.db.journal.dirtyOverride(c.address)
 	}
-	c.touched = true
 }
 
 func (c *stateObject) getTrie(db Database) Trie {
@@ -195,7 +186,7 @@ func (self *stateObject) GetState(db Database, key common.Hash) common.Hash {
 
 // SetState updates a value in account storage.
 func (self *stateObject) SetState(db Database, key, value common.Hash) {
-	self.db.journal = append(self.db.journal, storageChange{
+	self.db.journal.append(storageChange{
 		account:  &self.address,
 		key:      key,
 		prevalue: self.GetState(db, key),
@@ -207,10 +198,6 @@ func (self *stateObject) setState(key, value common.Hash) {
 	self.cachedStorage[key] = value
 	self.dirtyStorage[key] = value
 
-	if self.onDirty != nil {
-		self.onDirty(self.Address())
-		self.onDirty = nil
-	}
 }
 
 // updateTrie writes cached storage modifications into the object's storage trie.
@@ -274,7 +261,7 @@ func (c *stateObject) SubBalance(amount *big.Int) {
 }
 
 func (self *stateObject) SetBalance(amount *big.Int) {
-	self.db.journal = append(self.db.journal, balanceChange{
+	self.db.journal.append(balanceChange{
 		account: &self.address,
 		prev:    new(big.Int).Set(self.data.Balance),
 	})
@@ -283,17 +270,13 @@ func (self *stateObject) SetBalance(amount *big.Int) {
 
 func (self *stateObject) setBalance(amount *big.Int) {
 	self.data.Balance = amount
-	if self.onDirty != nil {
-		self.onDirty(self.Address())
-		self.onDirty = nil
-	}
 }
 
 // Return the gas back to the origin. Used by the Virtual machine or Closures
 func (c *stateObject) ReturnGas(gas *big.Int) {}
 
-func (self *stateObject) deepCopy(db *StateDB, onDirty func(addr common.Address)) *stateObject {
-	stateObject := newObject(db, self.address, self.data, onDirty)
+func (self *stateObject) deepCopy(db *StateDB) *stateObject {
+	stateObject := newObject(db, self.address, self.data)
 	if self.trie != nil {
 		stateObject.trie = db.db.CopyTrie(self.trie)
 	}
@@ -333,7 +316,7 @@ func (self *stateObject) Code(db Database) []byte {
 
 func (self *stateObject) SetCode(codeHash common.Hash, code []byte) {
 	prevcode := self.Code(self.db.db)
-	self.db.journal = append(self.db.journal, codeChange{
+	self.db.journal.append(codeChange{
 		account:  &self.address,
 		prevhash: self.CodeHash(),
 		prevcode: prevcode,
@@ -345,14 +328,10 @@ func (self *stateObject) setCode(codeHash common.Hash, code []byte) {
 	self.code = code
 	self.data.CodeHash = codeHash[:]
 	self.dirtyCode = true
-	if self.onDirty != nil {
-		self.onDirty(self.Address())
-		self.onDirty = nil
-	}
 }
 
 func (self *stateObject) SetNonce(nonce uint64) {
-	self.db.journal = append(self.db.journal, nonceChange{
+	self.db.journal.append(nonceChange{
 		account: &self.address,
 		prev:    self.data.Nonce,
 	})
@@ -361,10 +340,6 @@ func (self *stateObject) SetNonce(nonce uint64) {
 
 func (self *stateObject) setNonce(nonce uint64) {
 	self.data.Nonce = nonce
-	if self.onDirty != nil {
-		self.onDirty(self.Address())
-		self.onDirty = nil
-	}
 }
 
 func (self *stateObject) CodeHash() []byte {

core/state/statedb.go

@@ -131,7 +131,7 @@ func (self *StateDB) Reset(root common.Hash) error {
 }
 
 func (self *StateDB) AddLog(log *types.Log) {
-	self.journal = append(self.journal, addLogChange{txhash: self.thash})
+	self.journal.append(addLogChange{txhash: self.thash})
 
 	log.TxHash = self.thash
 	log.BlockHash = self.bhash
@@ -156,7 +156,7 @@ func (self *StateDB) Logs() []*types.Log {
 // AddPreimage records a SHA3 preimage seen by the VM.
 func (self *StateDB) AddPreimage(hash common.Hash, preimage []byte) {
 	if _, ok := self.preimages[hash]; !ok {
-		self.journal = append(self.journal, addPreimageChange{hash: hash})
+		self.journal.append(addPreimageChange{hash: hash})
 		pi := make([]byte, len(preimage))
 		copy(pi, preimage)
 		self.preimages[hash] = pi
@@ -169,7 +169,7 @@ func (self *StateDB) Preimages() map[common.Hash][]byte {
 }
 
 func (self *StateDB) AddRefund(gas uint64) {
-	self.journal = append(self.journal, refundChange{prev: self.refund})
+	self.journal.append(refundChange{prev: self.refund})
 	self.refund += gas
 }
 
@@ -255,7 +255,7 @@ func (self *StateDB) StorageTrie(addr common.Address) Trie {
 	if stateObject == nil {
 		return nil
 	}
-	cpy := stateObject.deepCopy(self, nil)
+	cpy := stateObject.deepCopy(self)
 	return cpy.updateTrie(self.db)
 }
 
@@ -325,7 +325,7 @@ func (self *StateDB) Suicide(addr common.Address) bool {
 	if stateObject == nil {
 		return false
 	}
-	self.journal = append(self.journal, suicideChange{
+	self.journal.append(suicideChange{
 		account:     &addr,
 		prev:        stateObject.suicided,
 		prevbalance: new(big.Int).Set(stateObject.Balance()),
@@ -379,7 +379,7 @@ func (self *StateDB) getStateObject(addr common.Address) (stateObject *stateObje
 		return nil
 	}
 	// Insert into the live set.
-	obj := newObject(self, addr, data, self.MarkStateObjectDirty)
+	obj := newObject(self, addr, data)
 	self.setStateObject(obj)
 	return obj
 }
@@ -397,22 +397,16 @@ func (self *StateDB) GetOrNewStateObject(addr common.Address) *stateObject {
 	return stateObject
 }
 
-// MarkStateObjectDirty adds the specified object to the dirty map to avoid costly
-// state object cache iteration to find a handful of modified ones.
-func (self *StateDB) MarkStateObjectDirty(addr common.Address) {
-	self.stateObjectsDirty[addr] = struct{}{}
-}
-
 // createObject creates a new state object. If there is an existing account with
 // the given address, it is overwritten and returned as the second return value.
 func (self *StateDB) createObject(addr common.Address) (newobj, prev *stateObject) {
 	prev = self.getStateObject(addr)
-	newobj = newObject(self, addr, Account{}, self.MarkStateObjectDirty)
+	newobj = newObject(self, addr, Account{})
 	newobj.setNonce(0) // sets the object to dirty
 	if prev == nil {
-		self.journal = append(self.journal, createObjectChange{account: &addr})
+		self.journal.append(createObjectChange{account: &addr})
 	} else {
-		self.journal = append(self.journal, resetObjectChange{prev: prev})
+		self.journal.append(resetObjectChange{prev: prev})
 	}
 	self.setStateObject(newobj)
 	return newobj, prev
@@ -462,20 +456,22 @@ func (self *StateDB) Copy() *StateDB {
 	self.lock.Lock()
 	defer self.lock.Unlock()
 
+	dirtyObjects := self.journal.flatten()
+
 	// Copy all the basic fields, initialize the memory ones
 	state := &StateDB{
 		db:                self.db,
 		trie:              self.db.CopyTrie(self.trie),
-		stateObjects:      make(map[common.Address]*stateObject, len(self.stateObjectsDirty)),
-		stateObjectsDirty: make(map[common.Address]struct{}, len(self.stateObjectsDirty)),
+		stateObjects:      make(map[common.Address]*stateObject, len(dirtyObjects)),
+		stateObjectsDirty: make(map[common.Address]struct{}, len(dirtyObjects)),
 		refund:            self.refund,
 		logs:              make(map[common.Hash][]*types.Log, len(self.logs)),
 		logSize:           self.logSize,
 		preimages:         make(map[common.Hash][]byte),
 	}
 	// Copy the dirty states, logs, and preimages
-	for addr := range self.stateObjectsDirty {
-		state.stateObjects[addr] = self.stateObjects[addr].deepCopy(state, state.MarkStateObjectDirty)
+	for addr := range dirtyObjects {
+		state.stateObjects[addr] = self.stateObjects[addr].deepCopy(state)
 		state.stateObjectsDirty[addr] = struct{}{}
 	}
 	for hash, logs := range self.logs {
@@ -492,7 +488,7 @@ func (self *StateDB) Copy() *StateDB {
 func (self *StateDB) Snapshot() int {
 	id := self.nextRevisionId
 	self.nextRevisionId++
-	self.validRevisions = append(self.validRevisions, revision{id, len(self.journal)})
+	self.validRevisions = append(self.validRevisions, revision{id, self.journal.Length()})
 	return id
 }
 
@@ -508,10 +504,10 @@ func (self *StateDB) RevertToSnapshot(revid int) {
 	snapshot := self.validRevisions[idx].journalIndex
 
 	// Replay the journal to undo changes.
-	for i := len(self.journal) - 1; i >= snapshot; i-- {
-		self.journal[i].undo(self)
+	for i := self.journal.Length() - 1; i >= snapshot; i-- {
+		self.journal.entries[i].undo(self)
 	}
-	self.journal = self.journal[:snapshot]
+	self.journal.entries = self.journal.entries[:snapshot]
 
 	// Remove invalidated snapshots from the stack.
 	self.validRevisions = self.validRevisions[:idx]
@@ -525,7 +521,8 @@ func (self *StateDB) GetRefund() uint64 {
 // Finalise finalises the state by removing the self destructed objects
 // and clears the journal as well as the refunds.
 func (s *StateDB) Finalise(deleteEmptyObjects bool) {
-	for addr := range s.stateObjectsDirty {
+
+	for addr, v := range s.journal.flatten() {
 		stateObject := s.stateObjects[addr]
 		if stateObject.suicided || (deleteEmptyObjects && stateObject.empty()) {
 			s.deleteStateObject(stateObject)
@@ -533,6 +530,7 @@ func (s *StateDB) Finalise(deleteEmptyObjects bool) {
 			stateObject.updateRoot(s.db)
 			s.updateStateObject(stateObject)
 		}
+		s.stateObjectsDirty[addr] = v
 	}
 	// Invalidate journal because reverting across transactions is not allowed.
 	s.clearJournalAndRefund()
@@ -576,7 +574,7 @@ func (s *StateDB) DeleteSuicides() {
 }
 
 func (s *StateDB) clearJournalAndRefund() {
-	s.journal = nil
+	s.journal = journal{}
 	s.validRevisions = s.validRevisions[:0]
 	s.refund = 0
 }
@@ -585,6 +583,10 @@ func (s *StateDB) clearJournalAndRefund() {
 func (s *StateDB) Commit(deleteEmptyObjects bool) (root common.Hash, err error) {
 	defer s.clearJournalAndRefund()
 
+	for addr, v := range s.journal.flatten() {
+		s.stateObjectsDirty[addr] = v
+	}
+
 	// Commit objects to the trie.
 	for addr, stateObject := range s.stateObjects {
 		_, isDirty := s.stateObjectsDirty[addr]

core/state/statedb_test.go

@@ -413,11 +413,12 @@ func (s *StateSuite) TestTouchDelete(c *check.C) {
 
 	snapshot := s.state.Snapshot()
 	s.state.AddBalance(common.Address{}, new(big.Int))
-	if len(s.state.stateObjectsDirty) != 1 {
+
+	if len(s.state.journal.flatten()) != 1 {
 		c.Fatal("expected one dirty state object")
 	}
 	s.state.RevertToSnapshot(snapshot)
-	if len(s.state.stateObjectsDirty) != 0 {
+	if len(s.state.journal.flatten()) != 0 {
 		c.Fatal("expected no dirty state object")
 	}
 }

tests/state_test.go

@@ -36,14 +36,8 @@ func TestState(t *testing.T) {
 	st.skipLoad(`^stTransactionTest/zeroSigTransa[^/]*\.json`) // EIP-86 is not supported yet
 	// Expected failures:
 	st.fails(`^stRevertTest/RevertPrecompiledTouch\.json/EIP158`, "bug in test")
-	st.fails(`^stRevertTest/RevertPrefoundEmptyOOG\.json/EIP158`, "bug in test")
 	st.fails(`^stRevertTest/RevertPrecompiledTouch\.json/Byzantium`, "bug in test")
-	st.fails(`^stRevertTest/RevertPrefoundEmptyOOG\.json/Byzantium`, "bug in test")
 	st.fails(`^stRandom2/randomStatetest64[45]\.json/(EIP150|Frontier|Homestead)/.*`, "known bug #15119")
-	st.fails(`^stCreateTest/TransactionCollisionToEmpty\.json/EIP158/2`, "known bug ")
-	st.fails(`^stCreateTest/TransactionCollisionToEmpty\.json/EIP158/3`, "known bug ")
-	st.fails(`^stCreateTest/TransactionCollisionToEmpty\.json/Byzantium/2`, "known bug ")
-	st.fails(`^stCreateTest/TransactionCollisionToEmpty\.json/Byzantium/3`, "known bug ")
 
 	st.walk(t, stateTestDir, func(t *testing.T, name string, test *StateTest) {
 		for _, subtest := range test.Subtests() {