fix: flush base buffer at startup

trie/triedb/pathdb/nodebufferlist.go

@@ -128,11 +128,12 @@ func newNodeBufferList(
 		waitForceKeepCh: make(chan struct{}),
 		keepFunc:        keepFunc,
 	}
+
 	go nf.loop()
 
 	log.Info("new node buffer list", "proposed block interval", nf.wpBlocks,
 		"reserve multi difflayers", nf.rsevMdNum, "difflayers in multidifflayer", nf.dlInMd,
-		"limit", common.StorageSize(limit), "layers", layers, "persist id", nf.persistID)
+		"limit", common.StorageSize(limit), "layers", layers, "persist id", nf.persistID, "base_size", size)
 	return nf
 }
 
@@ -517,6 +518,7 @@ func (nf *nodebufferlist) diffToBase() {
 
 // backgroundFlush flush base node buffer to disk.
 func (nf *nodebufferlist) backgroundFlush() {
+	log.Info("Start to background flush")
 	nf.flushMux.Lock()
 	defer nf.flushMux.Unlock()
 	nf.baseMux.RLock()
@@ -590,7 +592,7 @@ func (nf *nodebufferlist) loop() {
 				continue
 			}
 			nf.diffToBase()
-			if nf.base.size > nf.base.limit {
+			if nf.base.size >= nf.base.limit {
 				nf.backgroundFlush()
 			}
 			nf.isFlushing.Swap(false)
@@ -859,6 +861,7 @@ func (mf *multiDifflayer) empty() bool {
 // flush persists the in-memory dirty trie node into the disk if the configured
 // memory threshold is reached. Note, all data must be written atomically.
 func (mf *multiDifflayer) flush(db ethdb.KeyValueStore, clean *fastcache.Cache, id uint64) error {
+	log.Info("Start to base flush")
 	// Ensure the target state id is aligned with the internal counter.
 	head := rawdb.ReadPersistentStateID(db)
 	if head+mf.layers != id {