core/state,triedb/pathdb: doc accuracy sweep — remove stale temporal markers

Addresses review suggestion S2/S7/S22.

Remove "NOT wired in yet", "in a later commit", and "Commit N"
references that were accurate at the time of their original commit
but became stale after subsequent commits landed on the same branch.
Update cross-references to name the actual functions and files
rather than commit numbers.

Specific fixes:
  * flat_codec_bintrie.go:50-57: "NOT wired" → "wired when isVerkle"
  * flat_codec_bintrie.go:360: "in a later commit" → "generate_bintrie.go"
  * database_hasher_binary.go:132: "in a later commit" → "StateDB.commit()"
  * journal.go:53-56: v4 comment updated from "reserved for" → actual description

core/state/database_hasher_binary.go

@@ -129,10 +129,11 @@ func (tr *warpBinTrie) copy() *warpBinTrie {
 //
 // binaryHasher also implements LeafProducer: alongside every trie mutation
 // it records the corresponding (stem, offset, value) write into an
-// internal buffer. The caller (StateDB.Commit in a later commit) drains
-// this buffer once per block and hands the writes to the pathdb flat-state
-// layer via the stateUpdate, keeping the bintrie trie and its flat-state
-// mirror consistent without recomputing the bintrie key derivation twice.
+// internal buffer. StateDB.commit() drains this buffer once per block
+// via LeafProducer.DrainStemWrites and hands the writes to the pathdb
+// flat-state layer via stateUpdate.encodeBinary, keeping the bintrie
+// trie and its flat-state mirror consistent without recomputing the
+// bintrie key derivation twice.
 type binaryHasher struct {
 	db   *triedb.Database
 	root common.Hash

triedb/pathdb/flat_codec_bintrie.go

@@ -47,15 +47,13 @@ import (
 // reads the stem from the store (not from the in-flight batch), so a
 // second write at the same stem would re-read the pre-flush state and
 // clobber the first write. The codec's public surface area is designed
-// around this assumption; Commit 8 of the bintrie flat-state plan
-// restructures writeStates to pre-aggregate per-stem writes so callers
-// do not have to handle this manually.
+// around this assumption; the Flush method pre-aggregates per-stem
+// writes so callers do not have to handle this manually.
 //
-// This codec is NOT wired into pathdb.Database.New yet — that happens in a
-// later commit once the leaf-production hook in binaryHasher and the
-// stateUpdate wiring are in place. Until then, all call sites still
-// dispatch through merkleFlatCodec and bintrie mode continues to use the
-// (soon to be replaced) keccak-shaped flat-state layout.
+// This codec is wired into pathdb.Database.New when isVerkle is true
+// (see database.go). The leaf-production hook in binaryHasher emits
+// per-offset writes via DrainStemWrites, which encodeBinary routes
+// into the per-offset accountData map consumed by Flush.
 type bintrieFlatCodec struct {
 	// db is the underlying key-value store used by applyWrites to read
 	// the current stem blob before merging in new (offset, value) pairs.
@@ -357,9 +355,10 @@ func (c *bintrieFlatCodec) AccountPrefix() []byte {
 
 // StoragePrefix returns the same prefix as AccountPrefix because bintrie
 // flat-state entries are stored in a single namespace (stems contain
-// both account and storage data). The generator in a later commit uses
-// a single iterator over this prefix rather than the two-tier
-// account-then-storage walk used by the merkle generator.
+// both account and storage data). The bintrie generator
+// (generate_bintrie.go) uses a single iterator over this prefix
+// rather than the two-tier account-then-storage walk used by the
+// merkle generator.
 func (c *bintrieFlatCodec) StoragePrefix() []byte {
 	return rawdb.BinTrieStemPrefix
 }

triedb/pathdb/journal.go

@@ -50,10 +50,10 @@ var (
 // - Version 1: storage.Incomplete field is removed
 // - Version 2: add post-modification state values
 // - Version 3: a flag has been added to indicate whether the storage slot key is the raw key or a hash
-// - Version 4: reserved for bintrie flat-state (per-stem layout). Bumping now
-//              discards any v3 journals belonging to a bintrie database so
-//              that the new layout can be introduced cleanly in follow-up
-//              commits without a migration path.
+// - Version 4: bintrie flat-state per-stem layout. The journalGenerator
+//              struct gains an IsBintrie flag (rlp:"optional", defaults to
+//              false) so the loader can discard journals from a mismatched
+//              scheme and trigger a full flat-state regeneration.
 const journalVersion uint64 = 4
 
 // loadJournal tries to parse the layer journal from the disk.