Statediff for full node

cmd/geth/config.go

@@ -155,8 +155,14 @@ func makeFullNode(ctx *cli.Context) *node.Node {
 	if ctx.GlobalIsSet(utils.ConstantinopleOverrideFlag.Name) {
 		cfg.Eth.ConstantinopleOverride = new(big.Int).SetUint64(ctx.GlobalUint64(utils.ConstantinopleOverrideFlag.Name))
 	}
+
 	utils.RegisterEthService(stack, &cfg.Eth)
 
+	if ctx.GlobalBool(utils.StateDiffFlag.Name) {
+		cfg.Eth.StateDiff = true
+		utils.RegisterStateDiffService(stack, ctx)
+	}
+
 	if ctx.GlobalBool(utils.DashboardEnabledFlag.Name) {
 		utils.RegisterDashboardService(stack, &cfg.Dashboard, gitCommit)
 	}
@@ -181,9 +187,6 @@ func makeFullNode(ctx *cli.Context) *node.Node {
 		utils.RegisterEthStatsService(stack, cfg.Ethstats.URL)
 	}
 
-	if ctx.GlobalBool(utils.StateDiffFlag.Name) {
-		utils.RegisterStateDiffService(stack, ctx)
-	}
 	return stack
 }
 

core/blockchain.go

@@ -71,10 +71,11 @@ const (
 // CacheConfig contains the configuration values for the trie caching/pruning
 // that's resident in a blockchain.
 type CacheConfig struct {
-	Disabled       bool          // Whether to disable trie write caching (archive node)
-	TrieCleanLimit int           // Memory allowance (MB) to use for caching trie nodes in memory
-	TrieDirtyLimit int           // Memory limit (MB) at which to start flushing dirty trie nodes to disk
-	TrieTimeLimit  time.Duration // Time limit after which to flush the current in-memory trie to disk
+	Disabled          bool          // Whether to disable trie write caching (archive node)
+	TrieCleanLimit    int           // Memory allowance (MB) to use for caching trie nodes in memory
+	TrieDirtyLimit    int           // Memory limit (MB) at which to start flushing dirty trie nodes to disk
+	TrieTimeLimit     time.Duration // Time limit after which to flush the current in-memory trie to disk
+	ProcessStateDiffs bool
 }
 
 // BlockChain represents the canonical chain given a database with a genesis
@@ -136,6 +137,8 @@ type BlockChain struct {
 
 	badBlocks      *lru.Cache              // Bad block cache
 	shouldPreserve func(*types.Block) bool // Function used to determine whether should preserve the given block.
+
+	stateDiffsProcessed map[common.Hash]int
 }
 
 // NewBlockChain returns a fully initialised block chain using information
@@ -155,24 +158,26 @@ func NewBlockChain(db ethdb.Database, cacheConfig *CacheConfig, chainConfig *par
 	blockCache, _ := lru.New(blockCacheLimit)
 	futureBlocks, _ := lru.New(maxFutureBlocks)
 	badBlocks, _ := lru.New(badBlockLimit)
-
+	stateDiffsProcessed := make(map[common.Hash]int)
 	bc := &BlockChain{
-		chainConfig:    chainConfig,
-		cacheConfig:    cacheConfig,
-		db:             db,
-		triegc:         prque.New(nil),
-		stateCache:     state.NewDatabaseWithCache(db, cacheConfig.TrieCleanLimit),
-		quit:           make(chan struct{}),
-		shouldPreserve: shouldPreserve,
-		bodyCache:      bodyCache,
-		bodyRLPCache:   bodyRLPCache,
-		receiptsCache:  receiptsCache,
-		blockCache:     blockCache,
-		futureBlocks:   futureBlocks,
-		engine:         engine,
-		vmConfig:       vmConfig,
-		badBlocks:      badBlocks,
+		chainConfig:         chainConfig,
+		cacheConfig:         cacheConfig,
+		db:                  db,
+		triegc:              prque.New(nil),
+		stateCache:          state.NewDatabaseWithCache(db, cacheConfig.TrieCleanLimit),
+		quit:                make(chan struct{}),
+		shouldPreserve:      shouldPreserve,
+		bodyCache:           bodyCache,
+		bodyRLPCache:        bodyRLPCache,
+		receiptsCache:       receiptsCache,
+		blockCache:          blockCache,
+		futureBlocks:        futureBlocks,
+		engine:              engine,
+		vmConfig:            vmConfig,
+		badBlocks:           badBlocks,
+		stateDiffsProcessed: stateDiffsProcessed,
 	}
+
 	bc.SetValidator(NewBlockValidator(chainConfig, bc, engine))
 	bc.SetProcessor(NewStateProcessor(chainConfig, bc, engine))
 
@@ -922,6 +927,17 @@ func (bc *BlockChain) WriteBlockWithoutState(block *types.Block, td *big.Int) (e
 	return nil
 }
 
+func (bc *BlockChain) AddToStateDiffProcessedCollection(hash common.Hash) {
+	count, ok := bc.stateDiffsProcessed[hash]
+
+	if ok {
+		count++
+		bc.stateDiffsProcessed[hash] = count
+	} else {
+		bc.stateDiffsProcessed[hash] = 1
+	}
+}
+
 // WriteBlockWithState writes the block and all associated state to the database.
 func (bc *BlockChain) WriteBlockWithState(block *types.Block, receipts []*types.Receipt, state *state.StateDB) (status WriteStatus, err error) {
 	bc.wg.Add(1)
@@ -994,6 +1010,16 @@ func (bc *BlockChain) WriteBlockWithState(block *types.Block, receipts []*types.
 					bc.triegc.Push(root, number)
 					break
 				}
+
+				if bc.cacheConfig.ProcessStateDiffs {
+					if !bc.allowedRootToBeDereferenced(root.(common.Hash)) {
+						bc.triegc.Push(root, number)
+						break
+					} else {
+						delete(bc.stateDiffsProcessed, root.(common.Hash))
+					}
+				}
+
 				triedb.Dereference(root.(common.Hash))
 			}
 		}
@@ -1048,6 +1074,13 @@ func (bc *BlockChain) WriteBlockWithState(block *types.Block, receipts []*types.
 	return status, nil
 }
 
+//if we haven't processed the statediff for a given state root and it's child, don't dereference it yet
+func (bc *BlockChain) allowedRootToBeDereferenced(root common.Hash) bool {
+	diffProcessedForSelfAndChildCount := 2
+	count := bc.stateDiffsProcessed[root]
+	return count >= diffProcessedForSelfAndChildCount
+}
+
 // addFutureBlock checks if the block is within the max allowed window to get
 // accepted for future processing, and returns an error if the block is too far
 // ahead and was not added.

core/blockchain_test.go

@@ -1483,3 +1483,84 @@ func BenchmarkBlockChain_1x1000Executions(b *testing.B) {
 
 	benchmarkLargeNumberOfValueToNonexisting(b, numTxs, numBlocks, recipientFn, dataFn)
 }
+
+func TestProcessingStateDiffs(t *testing.T) {
+	defaultTrieCleanCache := 256
+	defaultTrieDirtyCache := 256
+	defaultTrieTimeout := 60 * time.Minute
+	cacheConfig := &CacheConfig{
+		Disabled:          false,
+		TrieCleanLimit:    defaultTrieCleanCache,
+		TrieDirtyLimit:    defaultTrieDirtyCache,
+		TrieTimeLimit:     defaultTrieTimeout,
+		ProcessStateDiffs: true,
+	}
+	db := ethdb.NewMemDatabase()
+	genesis := new(Genesis).MustCommit(db)
+	numberOfBlocks := triesInMemory
+	engine := ethash.NewFaker()
+	blockchain, _ := NewBlockChain(db, cacheConfig, params.AllEthashProtocolChanges, engine, vm.Config{}, nil)
+	blocks := makeBlockChain(genesis, numberOfBlocks+1, engine, db, canonicalSeed)
+	_, err := blockchain.InsertChain(blocks)
+	if err != nil {
+		t.Fatalf("failed to create pristine chain: %v", err)
+	}
+	defer blockchain.Stop()
+
+	//when adding a root hash to the collection, it will increment the count
+	firstStateRoot := blocks[0].Root()
+	blockchain.AddToStateDiffProcessedCollection(firstStateRoot)
+	value, ok := blockchain.stateDiffsProcessed[firstStateRoot]
+	if !ok {
+		t.Error("state root not found in collection")
+	}
+	if value != 1 {
+		t.Error("state root count not correct", "want", 1, "got", value)
+	}
+
+	blockchain.AddToStateDiffProcessedCollection(firstStateRoot)
+	value, ok = blockchain.stateDiffsProcessed[firstStateRoot]
+	if !ok {
+		t.Error("state root not found in collection")
+	}
+	if value != 2 {
+		t.Error("state root count not correct", "want", 2, "got", value)
+	}
+
+	moreBlocks := makeBlockChain(blocks[len(blocks)-1], 1, engine, db, canonicalSeed)
+	_, err = blockchain.InsertChain(moreBlocks)
+
+	//a root hash can be dereferenced when it's state diff and it's child's state diff have been processed
+	//(i.e. it has a count of 2 in stateDiffsProcessed)
+	nodes := blockchain.stateCache.TrieDB().Nodes()
+	if containsRootHash(nodes, firstStateRoot) {
+		t.Errorf("stateRoot %s in nodes, want: %t, got: %t", firstStateRoot.Hex(), false, true)
+	}
+
+	//a root hash should still be in the in-mem db if it's child's state diff hasn't yet been processed
+	//(i.e. it has a count of 1 stateDiffsProcessed)
+	secondStateRoot := blocks[1].Root()
+	blockchain.AddToStateDiffProcessedCollection(secondStateRoot)
+	if !containsRootHash(nodes, secondStateRoot) {
+		t.Errorf("stateRoot %s in nodes, want: %t, got: %t", secondStateRoot.Hex(), true, false)
+	}
+
+	//the stateDiffsProcessed collection is cleaned up once a hash has been dereferenced
+	_, ok = blockchain.stateDiffsProcessed[firstStateRoot]
+	if ok {
+		t.Errorf("stateRoot %s in stateDiffsProcessed collection, want: %t, got: %t",
+			firstStateRoot.Hex(),
+			false,
+			ok,
+		)
+	}
+}
+
+func containsRootHash(collection []common.Hash, hash common.Hash) bool {
+	for _, n := range collection {
+		if n == hash {
+			return true
+		}
+	}
+	return false
+}

eth/backend.go

@@ -157,10 +157,11 @@ func New(ctx *node.ServiceContext, config *Config) (*Ethereum, error) {
 			EVMInterpreter:          config.EVMInterpreter,
 		}
 		cacheConfig = &core.CacheConfig{
-			Disabled:       config.NoPruning,
-			TrieCleanLimit: config.TrieCleanCache,
-			TrieDirtyLimit: config.TrieDirtyCache,
-			TrieTimeLimit:  config.TrieTimeout,
+			Disabled:          config.NoPruning,
+			TrieCleanLimit:    config.TrieCleanCache,
+			TrieDirtyLimit:    config.TrieDirtyCache,
+			TrieTimeLimit:     config.TrieTimeout,
+			ProcessStateDiffs: config.StateDiff,
 		}
 	)
 	eth.blockchain, err = core.NewBlockChain(chainDb, cacheConfig, eth.chainConfig, eth.engine, vmConfig, eth.shouldPreserve)

eth/config.go

@@ -59,6 +59,8 @@ var DefaultConfig = Config{
 		Blocks:     20,
 		Percentile: 60,
 	},
+
+	StateDiff: false,
 }
 
 func init() {
@@ -135,6 +137,8 @@ type Config struct {
 
 	// Constantinople block override (TODO: remove after the fork)
 	ConstantinopleOverride *big.Int
+
+	StateDiff bool
 }
 
 type configMarshaling struct {