core/filtermaps: allow log search while head indexing (ethereum#31429)

This PR changes the matcher syncing conditions so that it is possible to
run a search while head indexing is in progress. Previously it was a
requirement to have the head indexed in order to perform matcher sync
before and after a search. This was unnecessarily strict as the purpose
was just to avoid syncing the valid range with the temporary shortened
indexed range applied while updating existing head maps. Now the sync
condition explicitly checks whether the indexer has a temporary indexed
range with some head maps being partially updated.
It also fixes a deadlock that happened when matcher synchronization was
attempted in the event handler called from the `writeFinishedMaps`
periodical callback.

Author: zsfelfoldi

core/filtermaps/filtermaps.go

@@ -70,6 +70,7 @@ type FilterMaps struct {
 	indexLock    sync.RWMutex
 	indexedRange filterMapsRange
 	indexedView  *ChainView // always consistent with the log index
+	hasTempRange bool
 
 	// also accessed by indexer and matcher backend but no locking needed.
 	filterMapCache *lru.Cache[uint32, filterMap]
@@ -94,7 +95,7 @@ type FilterMaps struct {
 	ptrTailUnindexMap                                            uint32
 
 	targetView            *ChainView
-	matcherSyncRequest    *FilterMapsMatcherBackend
+	matcherSyncRequests   []*FilterMapsMatcherBackend
 	historyCutoff         uint64
 	finalBlock, lastFinal uint64
 	lastFinalEpoch        uint32
@@ -330,7 +331,7 @@ func (f *FilterMaps) init() error {
 		fmr.blocks = common.NewRange(cp.BlockNumber+1, 0)
 		fmr.maps = common.NewRange(uint32(bestLen)<<f.logMapsPerEpoch, 0)
 	}
-	f.setRange(batch, f.targetView, fmr)
+	f.setRange(batch, f.targetView, fmr, false)
 	return batch.Write()
 }
 
@@ -373,9 +374,10 @@ func (f *FilterMaps) removeDbWithPrefix(prefix []byte, action string) bool {
 // setRange updates the indexed chain view and covered range and also adds the
 // changes to the given batch.
 // Note that this function assumes that the index write lock is being held.
-func (f *FilterMaps) setRange(batch ethdb.KeyValueWriter, newView *ChainView, newRange filterMapsRange) {
+func (f *FilterMaps) setRange(batch ethdb.KeyValueWriter, newView *ChainView, newRange filterMapsRange, isTempRange bool) {
 	f.indexedView = newView
 	f.indexedRange = newRange
+	f.hasTempRange = isTempRange
 	f.updateMatchersValidRange()
 	if newRange.initialized {
 		rs := rawdb.FilterMapsRange{
@@ -666,7 +668,7 @@ func (f *FilterMaps) deleteTailEpoch(epoch uint32) error {
 	} else {
 		return errors.New("invalid tail epoch number")
 	}
-	f.setRange(f.db, f.indexedView, fmr)
+	f.setRange(f.db, f.indexedView, fmr, false)
 	first := f.mapRowIndex(firstMap, 0)
 	count := f.mapRowIndex(firstMap+f.mapsPerEpoch, 0) - first
 	rawdb.DeleteFilterMapRows(f.db, common.NewRange(first, count))

core/filtermaps/indexer.go

@@ -136,15 +136,18 @@ func (f *FilterMaps) processEvents() {
 // processSingleEvent processes a single event either in a blocking or
 // non-blocking manner.
 func (f *FilterMaps) processSingleEvent(blocking bool) bool {
-	if f.matcherSyncRequest != nil && f.targetHeadIndexed() {
-		f.matcherSyncRequest.synced()
-		f.matcherSyncRequest = nil
+	if !f.hasTempRange {
+		for _, mb := range f.matcherSyncRequests {
+			mb.synced()
+		}
+		f.matcherSyncRequests = nil
 	}
 	if blocking {
 		select {
 		case target := <-f.targetCh:
 			f.setTarget(target)
-		case f.matcherSyncRequest = <-f.matcherSyncCh:
+		case mb := <-f.matcherSyncCh:
+			f.matcherSyncRequests = append(f.matcherSyncRequests, mb)
 		case f.blockProcessing = <-f.blockProcessingCh:
 		case <-f.closeCh:
 			f.stop = true
@@ -160,7 +163,8 @@ func (f *FilterMaps) processSingleEvent(blocking bool) bool {
 		select {
 		case target := <-f.targetCh:
 			f.setTarget(target)
-		case f.matcherSyncRequest = <-f.matcherSyncCh:
+		case mb := <-f.matcherSyncCh:
+			f.matcherSyncRequests = append(f.matcherSyncRequests, mb)
 		case f.blockProcessing = <-f.blockProcessingCh:
 		case <-f.closeCh:
 			f.stop = true

core/filtermaps/map_renderer.go

@@ -392,12 +392,16 @@ func (r *mapRenderer) writeFinishedMaps(pauseCb func() bool) error {
 			}
 			// do not exit while in partially written state but do allow processing
 			// events and pausing while block processing is in progress
+			r.f.indexLock.Unlock()
 			pauseCb()
+			r.f.indexLock.Lock()
 			batch = r.f.db.NewBatch()
 		}
 	}
 
-	r.f.setRange(batch, r.f.indexedView, tempRange)
+	if tempRange != r.f.indexedRange {
+		r.f.setRange(batch, r.f.indexedView, tempRange, true)
+	}
 	// add or update filter rows
 	for rowIndex := uint32(0); rowIndex < r.f.mapHeight; rowIndex++ {
 		var (
@@ -469,7 +473,7 @@ func (r *mapRenderer) writeFinishedMaps(pauseCb func() bool) error {
 	}
 	r.finishedMaps = make(map[uint32]*renderedMap)
 	r.finished.SetFirst(r.finished.AfterLast())
-	r.f.setRange(batch, renderedView, newRange)
+	r.f.setRange(batch, renderedView, newRange, false)
 	if err := batch.Write(); err != nil {
 		log.Crit("Error writing log index update batch", "error", err)
 	}

core/filtermaps/matcher_backend.go

@@ -125,7 +125,7 @@ func (fm *FilterMapsMatcherBackend) synced() {
 		indexedBlocks.SetAfterLast(indexedBlocks.Last()) // remove partially indexed last block
 	}
 	fm.syncCh <- SyncRange{
-		HeadNumber:    fm.f.indexedView.headNumber,
+		HeadNumber:    fm.f.targetView.headNumber,
 		ValidBlocks:   fm.validBlocks,
 		IndexedBlocks: indexedBlocks,
 	}