trie, les, tests, core: implement trie tracer (#24403)

Trie tracer is an auxiliary tool to capture all deleted nodes
which can't be captured by trie.Committer. The deleted nodes
can be removed from the disk later.

trie/trie.go

@@ -24,6 +24,7 @@ import (
 	"sync"
 
 	"github.com/ethereum/go-ethereum/common"
+	"github.com/ethereum/go-ethereum/core/rawdb"
 	"github.com/ethereum/go-ethereum/core/types"
 	"github.com/ethereum/go-ethereum/crypto"
 	"github.com/ethereum/go-ethereum/log"
@@ -62,10 +63,15 @@ type LeafCallback func(paths [][]byte, hexpath []byte, leaf []byte, parent commo
 type Trie struct {
 	db   *Database
 	root node
-	// Keep track of the number leafs which have been inserted since the last
+
+	// Keep track of the number leaves which have been inserted since the last
 	// hashing operation. This number will not directly map to the number of
 	// actually unhashed nodes
 	unhashed int
+
+	// tracer is the state diff tracer can be used to track newly added/deleted
+	// trie node. It will be reset after each commit operation.
+	tracer *tracer
 }
 
 // newFlag returns the cache flag value for a newly created node.
@@ -73,6 +79,16 @@ func (t *Trie) newFlag() nodeFlag {
 	return nodeFlag{dirty: true}
 }
 
+// Copy returns a copy of Trie.
+func (t *Trie) Copy() *Trie {
+	return &Trie{
+		db:       t.db,
+		root:     t.root,
+		unhashed: t.unhashed,
+		tracer:   t.tracer.copy(),
+	}
+}
+
 // New creates a trie with an existing root node from db.
 //
 // If root is the zero hash or the sha3 hash of an empty string, the
@@ -85,6 +101,7 @@ func New(root common.Hash, db *Database) (*Trie, error) {
 	}
 	trie := &Trie{
 		db: db,
+		//tracer: newTracer(),
 	}
 	if root != (common.Hash{}) && root != emptyRoot {
 		rootnode, err := trie.resolveHash(root[:], nil)
@@ -96,6 +113,16 @@ func New(root common.Hash, db *Database) (*Trie, error) {
 	return trie, nil
 }
 
+// newWithRootNode initializes the trie with the given root node.
+// It's only used by range prover.
+func newWithRootNode(root node) *Trie {
+	return &Trie{
+		root: root,
+		//tracer: newTracer(),
+		db: NewDatabase(rawdb.NewMemoryDatabase()),
+	}
+}
+
 // NodeIterator returns an iterator that returns nodes of the trie. Iteration starts at
 // the key after the given start key.
 func (t *Trie) NodeIterator(start []byte) NodeIterator {
@@ -317,7 +344,12 @@ func (t *Trie) insert(n node, prefix, key []byte, value node) (bool, node, error
 		if matchlen == 0 {
 			return true, branch, nil
 		}
-		// Otherwise, replace it with a short node leading up to the branch.
+		// New branch node is created as a child of the original short node.
+		// Track the newly inserted node in the tracer. The node identifier
+		// passed is the path from the root node.
+		t.tracer.onInsert(append(prefix, key[:matchlen]...))
+
+		// Replace it with a short node leading up to the branch.
 		return true, &shortNode{key[:matchlen], branch, t.newFlag()}, nil
 
 	case *fullNode:
@@ -331,6 +363,11 @@ func (t *Trie) insert(n node, prefix, key []byte, value node) (bool, node, error
 		return true, n, nil
 
 	case nil:
+		// New short node is created and track it in the tracer. The node identifier
+		// passed is the path from the root node. Note the valueNode won't be tracked
+		// since it's always embedded in its parent.
+		t.tracer.onInsert(prefix)
+
 		return true, &shortNode{key, value, t.newFlag()}, nil
 
 	case hashNode:
@@ -383,6 +420,11 @@ func (t *Trie) delete(n node, prefix, key []byte) (bool, node, error) {
 			return false, n, nil // don't replace n on mismatch
 		}
 		if matchlen == len(key) {
+			// The matched short node is deleted entirely and track
+			// it in the deletion set. The same the valueNode doesn't
+			// need to be tracked at all since it's always embedded.
+			t.tracer.onDelete(prefix)
+
 			return true, nil, nil // remove n entirely for whole matches
 		}
 		// The key is longer than n.Key. Remove the remaining suffix
@@ -395,6 +437,10 @@ func (t *Trie) delete(n node, prefix, key []byte) (bool, node, error) {
 		}
 		switch child := child.(type) {
 		case *shortNode:
+			// The child shortNode is merged into its parent, track
+			// is deleted as well.
+			t.tracer.onDelete(append(prefix, n.Key...))
+
 			// Deleting from the subtrie reduced it to another
 			// short node. Merge the nodes to avoid creating a
 			// shortNode{..., shortNode{...}}. Use concat (which
@@ -456,6 +502,11 @@ func (t *Trie) delete(n node, prefix, key []byte) (bool, node, error) {
 					return false, nil, err
 				}
 				if cnode, ok := cnode.(*shortNode); ok {
+					// Replace the entire full node with the short node.
+					// Mark the original short node as deleted since the
+					// value is embedded into the parent now.
+					t.tracer.onDelete(append(prefix, byte(pos)))
+
 					k := append([]byte{byte(pos)}, cnode.Key...)
 					return true, &shortNode{k, cnode.Val, t.newFlag()}, nil
 				}
@@ -537,6 +588,8 @@ func (t *Trie) Commit(onleaf LeafCallback) (common.Hash, int, error) {
 	if t.db == nil {
 		panic("commit called on trie with nil database")
 	}
+	defer t.tracer.reset()
+
 	if t.root == nil {
 		return emptyRoot, 0, nil
 	}
@@ -595,4 +648,5 @@ func (t *Trie) hashRoot() (node, node, error) {
 func (t *Trie) Reset() {
 	t.root = nil
 	t.unhashed = 0
+	t.tracer.reset()
 }