Merge pull request #40 from vulcanize/rebase-1.10.23-wip

version updates for geth 1.10.23 rebase

cmd/validateTrie.go

@@ -59,7 +59,8 @@ It can operate at three levels:
 }
 
 func validateTrie() {
-	v, err := newValidator()
+	numWorkers := viper.GetUint("validator.workers")
+	v, err := newValidator(numWorkers)
 	if err != nil {
 		logWithCommand.Fatal(err)
 	}
@@ -104,20 +105,20 @@ func validateTrie() {
 	logWithCommand.Debugf("groupcache stats %+v", stats)
 }
 
-func newValidator() (*validator.Validator, error) {
+func newValidator(workers uint) (*validator.Validator, error) {
 	ipfsPath := viper.GetString("ipfs.path")
 	if ipfsPath == "" {
 		db, err := validator.NewDB()
 		if err != nil {
 			logWithCommand.Fatal(err)
 		}
-		return validator.NewPGIPFSValidator(db), nil
+		return validator.NewPGIPFSValidator(db, workers), nil
 	}
 	bs, err := validator.InitIPFSBlockService(ipfsPath)
 	if err != nil {
 		return nil, err
 	}
-	return validator.NewIPFSValidator(bs), nil
+	return validator.NewIPFSValidator(bs, workers), nil
 }
 
 func init() {
@@ -128,10 +129,12 @@ func init() {
 	validateTrieCmd.PersistentFlags().String("storage-root", "", "Root of the storage trie we wish to validate; for storage validation")
 	validateTrieCmd.PersistentFlags().String("address", "", "Contract address for the storage trie we wish to validate; for storage validation")
 	validateTrieCmd.PersistentFlags().String("ipfs-path", "", "Path to IPFS repository; if provided operations move through the IPFS repo otherwise Postgres connection params are expected in the provided config")
+	validateTrieCmd.PersistentFlags().Int("workers", 4, "number of concurrent workers to use")
 
 	viper.BindPFlag("validator.stateRoot", validateTrieCmd.PersistentFlags().Lookup("state-root"))
 	viper.BindPFlag("validator.type", validateTrieCmd.PersistentFlags().Lookup("type"))
 	viper.BindPFlag("validator.storageRoot", validateTrieCmd.PersistentFlags().Lookup("storage-root"))
 	viper.BindPFlag("validator.address", validateTrieCmd.PersistentFlags().Lookup("address"))
+	viper.BindPFlag("validator.workers", validateTrieCmd.PersistentFlags().Lookup("workers"))
 	viper.BindPFlag("ipfs.path", validateTrieCmd.PersistentFlags().Lookup("ipfs-path"))
 }

go.mod

@@ -3,14 +3,14 @@ module github.com/vulcanize/eth-ipfs-state-validator/v4
 go 1.18
 
 require (
-	github.com/ethereum/go-ethereum v1.10.21
+	github.com/ethereum/go-ethereum v1.10.23
 	github.com/ipfs/go-blockservice v0.4.0
 	github.com/ipfs/go-cid v0.2.0
 	github.com/ipfs/go-ipfs-blockstore v1.2.0
 	github.com/ipfs/go-ipfs-ds-help v1.1.0
 	github.com/ipfs/kubo v0.14.0
 	github.com/jmoiron/sqlx v1.3.5
-	github.com/lib/pq v1.10.5
+	github.com/lib/pq v1.10.6
 	github.com/mailgun/groupcache/v2 v2.3.0
 	github.com/multiformats/go-multihash v0.2.0
 	github.com/onsi/ginkgo v1.16.5
@@ -18,7 +18,7 @@ require (
 	github.com/sirupsen/logrus v1.8.1
 	github.com/spf13/cobra v1.4.0
 	github.com/spf13/viper v1.11.0
-	github.com/vulcanize/ipfs-ethdb/v4 v4.0.5-alpha
+	github.com/vulcanize/ipfs-ethdb/v4 v4.0.7-alpha
 )
 
 require (
@@ -202,6 +202,7 @@ require (
 	github.com/tidwall/pretty v1.2.0 // indirect
 	github.com/tklauser/go-sysconf v0.3.5 // indirect
 	github.com/tklauser/numcpus v0.2.2 // indirect
+	github.com/vulcanize/go-eth-state-node-iterator v1.1.4
 	github.com/wI2L/jsondiff v0.2.0 // indirect
 	github.com/whyrusleeping/base32 v0.0.0-20170828182744-c30ac30633cc // indirect
 	github.com/whyrusleeping/cbor-gen v0.0.0-20210219115102-f37d292932f2 // indirect
@@ -219,7 +220,7 @@ require (
 	go.uber.org/multierr v1.8.0 // indirect
 	go.uber.org/zap v1.21.0 // indirect
 	go4.org v0.0.0-20200411211856-f5505b9728dd // indirect
-	golang.org/x/crypto v0.0.0-20220525230936-793ad666bf5e // indirect
+	golang.org/x/crypto v0.0.0-20220829220503-c86fa9a7ed90 // indirect
 	golang.org/x/mod v0.6.0-dev.0.20220106191415-9b9b3d81d5e3 // indirect
 	golang.org/x/net v0.0.0-20220607020251-c690dde0001d // indirect
 	golang.org/x/sync v0.0.0-20210220032951-036812b2e83c // indirect
@@ -235,4 +236,4 @@ require (
 	lukechampine.com/blake3 v1.1.7 // indirect
 )
 
-replace github.com/ethereum/go-ethereum v1.10.21 => github.com/vulcanize/go-ethereum v1.10.21-statediff-4.1.2-alpha
+replace github.com/ethereum/go-ethereum v1.10.23 => github.com/vulcanize/go-ethereum v1.10.23-statediff-4.2.0-alpha

go.sum

@@ -732,8 +732,8 @@ github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI=
 github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/lib/pq v1.2.0/go.mod h1:5WUZQaWbwv1U+lTReE5YruASi9Al49XbQIvNi/34Woo=
-github.com/lib/pq v1.10.5 h1:J+gdV2cUmX7ZqL2B0lFcW0m+egaHC2V3lpO8nWxyYiQ=
-github.com/lib/pq v1.10.5/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
+github.com/lib/pq v1.10.6 h1:jbk+ZieJ0D7EVGJYpL9QTz7/YW6UHbmdnZWYyK5cdBs=
+github.com/lib/pq v1.10.6/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/libp2p/go-addr-util v0.0.1/go.mod h1:4ac6O7n9rIAKB1dnd+s8IbbMXkt+oBpzX4/+RACcnlQ=
 github.com/libp2p/go-addr-util v0.0.2/go.mod h1:Ecd6Fb3yIuLzq4bD7VcywcVSBtefcAwnUISBM3WG15E=
 github.com/libp2p/go-addr-util v0.1.0/go.mod h1:6I3ZYuFr2O/9D+SoyM0zEw0EF3YkldtTX406BpdQMqw=
@@ -1519,10 +1519,12 @@ github.com/urfave/cli v1.22.2/go.mod h1:Gos4lmkARVdJ6EkW0WaNv/tZAAMe9V7XWyB60NtX
 github.com/urfave/cli/v2 v2.0.0/go.mod h1:SE9GqnLQmjVa0iPEY0f1w3ygNIYcIJ0OKPMoW2caLfQ=
 github.com/viant/assertly v0.4.8/go.mod h1:aGifi++jvCrUaklKEKT0BU95igDNaqkvz+49uaYMPRU=
 github.com/viant/toolbox v0.24.0/go.mod h1:OxMCG57V0PXuIP2HNQrtJf2CjqdmbrOx5EkMILuUhzM=
-github.com/vulcanize/go-ethereum v1.10.21-statediff-4.1.2-alpha h1:ct+8FGuQnHA6SOGOQoAMINWdeexuSF40+IjF48J094A=
-github.com/vulcanize/go-ethereum v1.10.21-statediff-4.1.2-alpha/go.mod h1:dNJkmCSbaasX0zfQM6pm1g3rWlW3EGhLOEZMScyrRAs=
-github.com/vulcanize/ipfs-ethdb/v4 v4.0.5-alpha h1:NFRwWeMB3Q+QqLM9qdcHvfvWBxOk0lPwhOqXJpkIg30=
-github.com/vulcanize/ipfs-ethdb/v4 v4.0.5-alpha/go.mod h1:WvYj0m0cLPAtoytTbcbE2nZ3Hg9iuuF+lY14dBVRWZQ=
+github.com/vulcanize/go-eth-state-node-iterator v1.1.4 h1:2jIbgZ/yxrNZt5zmsXtvUPSxT0sP5k1PR5/O1GN/Ma0=
+github.com/vulcanize/go-eth-state-node-iterator v1.1.4/go.mod h1:kFzlSi7rUB9mJWZf08xIEiiGuXnnUEn48It+WMHzxPc=
+github.com/vulcanize/go-ethereum v1.10.23-statediff-4.2.0-alpha h1:iy1uhdazPIlUJcCu6Kal+hFQiZblTDwSe2i644o167Y=
+github.com/vulcanize/go-ethereum v1.10.23-statediff-4.2.0-alpha/go.mod h1:lKBVBWksSwBDR/5D9CAxaGQzDPIS3ueWb6idy7X1Shg=
+github.com/vulcanize/ipfs-ethdb/v4 v4.0.7-alpha h1:UdOYaVmK/QaFyUDg44eEcEaNivP6059hmK6Qeh+3Mg8=
+github.com/vulcanize/ipfs-ethdb/v4 v4.0.7-alpha/go.mod h1:gK23M2S0428yKsJoiUzBMTMVRt6OdOuFEFPcnO/ksGA=
 github.com/wI2L/jsondiff v0.2.0 h1:dE00WemBa1uCjrzQUUTE/17I6m5qAaN0EMFOg2Ynr/k=
 github.com/wI2L/jsondiff v0.2.0/go.mod h1:axTcwtBkY4TsKuV+RgoMhHyHKKFRI6nnjRLi8LLYQnA=
 github.com/wangjia184/sortedset v0.0.0-20160527075905-f5d03557ba30/go.mod h1:YkocrP2K2tcw938x9gCOmT5G5eCD6jsTz0SZuyAqwIE=
@@ -1677,8 +1679,8 @@ golang.org/x/crypto v0.0.0-20210711020723-a769d52b0f97/go.mod h1:GvvjBRRGRdwPK5y
 golang.org/x/crypto v0.0.0-20210813211128-0a44fdfbc16e/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
 golang.org/x/crypto v0.0.0-20211108221036-ceb1ce70b4fa/go.mod h1:GvvjBRRGRdwPK5ydBHafDWAxML/pGHZbMvKqRZ5+Abc=
 golang.org/x/crypto v0.0.0-20220518034528-6f7dac969898/go.mod h1:IxCIyHEi3zRg3s0A5j5BB6A9Jmi73HwBIUl50j+osU4=
-golang.org/x/crypto v0.0.0-20220525230936-793ad666bf5e h1:T8NU3HyQ8ClP4SEE+KbFlg6n0NhuTsN4MyznaarGsZM=
-golang.org/x/crypto v0.0.0-20220525230936-793ad666bf5e/go.mod h1:IxCIyHEi3zRg3s0A5j5BB6A9Jmi73HwBIUl50j+osU4=
+golang.org/x/crypto v0.0.0-20220829220503-c86fa9a7ed90 h1:Y/gsMcFOcR+6S6f3YeMKl5g+dZMEWqcz5Czj/GWYbkM=
+golang.org/x/crypto v0.0.0-20220829220503-c86fa9a7ed90/go.mod h1:IxCIyHEi3zRg3s0A5j5BB6A9Jmi73HwBIUl50j+osU4=
 golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
 golang.org/x/exp v0.0.0-20190306152737-a1d7652674e8/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
 golang.org/x/exp v0.0.0-20190510132918-efd6b22b2522/go.mod h1:ZjyILWgesfNpC6sMxTJOJm9Kp84zZh5NQWvqDGG3Qr8=

pkg/validator.go

@@ -17,17 +17,23 @@
 package validator
 
 import (
+	"bytes"
+	"fmt"
+	"sync"
 	"time"
 
 	"github.com/ethereum/go-ethereum/common"
 	"github.com/ethereum/go-ethereum/core/state"
+	"github.com/ethereum/go-ethereum/core/types"
 	"github.com/ethereum/go-ethereum/crypto"
 	"github.com/ethereum/go-ethereum/ethdb"
+	"github.com/ethereum/go-ethereum/rlp"
 	"github.com/ethereum/go-ethereum/trie"
 	"github.com/ipfs/go-blockservice"
 	"github.com/jmoiron/sqlx"
 	"github.com/mailgun/groupcache/v2"
 
+	nodeiter "github.com/vulcanize/go-eth-state-node-iterator"
 	ipfsethdb "github.com/vulcanize/ipfs-ethdb/v4"
 	pgipfsethdb "github.com/vulcanize/ipfs-ethdb/v4/postgres"
 )
@@ -38,10 +44,14 @@ type Validator struct {
 	trieDB        *trie.Database
 	stateDatabase state.Database
 	db            *pgipfsethdb.Database
+
+	iterWorkers uint
 }
 
+var emptyCodeHash = crypto.Keccak256(nil)
+
 // NewPGIPFSValidator returns a new trie validator ontop of a connection pool for an IPFS backing Postgres database
-func NewPGIPFSValidator(db *sqlx.DB) *Validator {
+func NewPGIPFSValidator(db *sqlx.DB, workers uint) *Validator {
 	kvs := pgipfsethdb.NewKeyValueStore(db, pgipfsethdb.CacheConfig{
 		Name:           "kv",
 		Size:           16 * 1000 * 1000, // 16MB
@@ -54,11 +64,15 @@ func NewPGIPFSValidator(db *sqlx.DB) *Validator {
 		ExpiryDuration: time.Hour * 8,    // 8 hours
 	})
 
+	if workers == 0 {
+		workers = 1
+	}
 	return &Validator{
 		kvs:           kvs,
 		trieDB:        trie.NewDatabase(kvs),
 		stateDatabase: state.NewDatabase(database),
 		db:            database.(*pgipfsethdb.Database),
+		iterWorkers:   workers,
 	}
 }
 
@@ -67,13 +81,17 @@ func (v *Validator) GetCacheStats() groupcache.Stats {
 }
 
 // NewIPFSValidator returns a new trie validator ontop of an IPFS blockservice
-func NewIPFSValidator(bs blockservice.BlockService) *Validator {
+func NewIPFSValidator(bs blockservice.BlockService, workers uint) *Validator {
 	kvs := ipfsethdb.NewKeyValueStore(bs)
 	database := ipfsethdb.NewDatabase(bs)
+	if workers == 0 {
+		workers = 1
+	}
 	return &Validator{
 		kvs:           kvs,
 		trieDB:        trie.NewDatabase(kvs),
 		stateDatabase: state.NewDatabase(database),
+		iterWorkers:   workers,
 	}
 }
 
@@ -88,21 +106,82 @@ func NewValidator(kvs ethdb.KeyValueStore, database ethdb.Database) *Validator {
 	}
 }
 
+// Traverses each iterator in a separate goroutine.
+// If storage = true, also traverse storage tries for each leaf.
+func (v *Validator) iterateAsync(iters []trie.NodeIterator, storage bool) error {
+	var wg sync.WaitGroup
+	errors := make(chan error)
+	for _, it := range iters {
+		wg.Add(1)
+		go func(it trie.NodeIterator) {
+			defer wg.Done()
+			for it.Next(true) {
+				// Iterate through entire state trie. it.Next() will return false when we have
+				// either completed iteration of the entire trie or run into an error (e.g. a
+				// missing node). If we are able to iterate through the entire trie without error
+				// then the trie is complete.
+
+				// If storage is not requested, or the state trie node is an internal entry, leave as is
+				if !storage || !it.Leaf() {
+					continue
+				}
+				// Otherwise we've reached an account node, initiate data iteration
+				var account types.StateAccount
+				if err := rlp.Decode(bytes.NewReader(it.LeafBlob()), &account); err != nil {
+					errors <- err
+					break
+				}
+				dataTrie, err := v.stateDatabase.OpenStorageTrie(common.BytesToHash(it.LeafKey()), account.Root)
+				if err != nil {
+					errors <- err
+					break
+				}
+				dataIt := dataTrie.NodeIterator(nil)
+				if !bytes.Equal(account.CodeHash, emptyCodeHash) {
+					addrHash := common.BytesToHash(it.LeafKey())
+					_, err := v.stateDatabase.ContractCode(addrHash, common.BytesToHash(account.CodeHash))
+					if err != nil {
+						errors <- fmt.Errorf("code %x: %v", account.CodeHash, err)
+						break
+					}
+				}
+				for dataIt.Next(true) {
+				}
+				if err = dataIt.Error(); err != nil {
+					errors <- err
+					break
+				}
+
+			}
+			if it.Error() != nil {
+				errors <- it.Error()
+			}
+		}(it)
+	}
+
+	done := make(chan struct{})
+	go func() {
+		wg.Wait()
+		done <- struct{}{}
+	}()
+	var err error
+	select {
+	case err = <-errors:
+	case <-done:
+		close(errors)
+	}
+	return err
+}
+
 // ValidateTrie returns an error if the state and storage tries for the provided state root cannot be confirmed as complete
 // This does consider child storage tries
 func (v *Validator) ValidateTrie(stateRoot common.Hash) error {
-	// Generate the state.NodeIterator for this root
-	stateDB, err := state.New(stateRoot, v.stateDatabase, nil)
+	t, err := v.stateDatabase.OpenTrie(stateRoot)
 	if err != nil {
 		return err
 	}
-	it := state.NewNodeIterator(stateDB)
-	for it.Next() {
-		// iterate through entire state trie and descendent storage tries
-		// it.Next() will return false when we have either completed iteration of the entire trie or have ran into an error (e.g. a missing node)
-		// if we are able to iterate through the entire trie without error then the trie is complete
-	}
-	return it.Error
+	iters := nodeiter.SubtrieIterators(t, v.iterWorkers)
+	return v.iterateAsync(iters, true)
 }
 
 // ValidateStateTrie returns an error if the state trie for the provided state root cannot be confirmed as complete
@@ -113,13 +192,8 @@ func (v *Validator) ValidateStateTrie(stateRoot common.Hash) error {
 	if err != nil {
 		return err
 	}
-	it := t.NodeIterator(nil)
-	for it.Next(true) {
-		// iterate through entire state trie
-		// it.Next() will return false when we have either completed iteration of the entire trie or have ran into an error (e.g. a missing node)
-		// if we are able to iterate through the entire trie without error then the trie is complete
-	}
-	return it.Error()
+	iters := nodeiter.SubtrieIterators(t, v.iterWorkers)
+	return v.iterateAsync(iters, false)
 }
 
 // ValidateStorageTrie returns an error if the storage trie for the provided storage root and contract address cannot be confirmed as complete
@@ -130,13 +204,8 @@ func (v *Validator) ValidateStorageTrie(address common.Address, storageRoot comm
 	if err != nil {
 		return err
 	}
-	it := t.NodeIterator(nil)
-	for it.Next(true) {
-		// iterate through entire storage trie
-		// it.Next() will return false when we have either completed iteration of the entire trie or have ran into an error (e.g. a missing node)
-		// if we are able to iterate through the entire trie without error then the trie is complete
-	}
-	return it.Error()
+	iters := nodeiter.SubtrieIterators(t, v.iterWorkers)
+	return v.iterateAsync(iters, false)
 }
 
 // Close implements io.Closer

pkg/validator_test.go

@@ -200,7 +200,7 @@ var _ = Describe("PG-IPFS Validator", func() {
 	BeforeEach(func() {
 		db, err = pgipfsethdb.TestDB()
 		Expect(err).ToNot(HaveOccurred())
-		v = validator.NewPGIPFSValidator(db)
+		v = validator.NewPGIPFSValidator(db, 4)
 	})
 	AfterEach(func() {
 		v.Close()