Improved debugging for HNSW locks [skip ci]

CHANGELOG.md

@@ -1,7 +1,3 @@
-## 0.6.2 (unreleased)
-
-- Improved performance of parallel HNSW index builds
-
 ## 0.6.1 (2024-03-04)
 
 - Fixed error with `ANALYZE` and vectors with different dimensions

src/hnsw.c

@@ -17,7 +17,10 @@
 #endif
 
 int			hnsw_ef_search;
-int			hnsw_lock_tranche_id;
+int			hnsw_entry_lock_tranche_id;
+int			hnsw_allocator_lock_tranche_id;
+int			hnsw_flush_lock_tranche_id;
+int			hnsw_element_lock_tranche_id;
 static relopt_kind hnsw_relopt_kind;
 
 /*
@@ -36,16 +39,27 @@ HnswInitLockTranche(void)
 	bool		found;
 
 	LWLockAcquire(AddinShmemInitLock, LW_EXCLUSIVE);
-	tranche_ids = ShmemInitStruct("hnsw LWLock ids",
-								  sizeof(int) * 1,
+	tranche_ids = ShmemInitStruct("hnsw LWLock ids v2",
+								  sizeof(int) * 4,
 								  &found);
 	if (!found)
+	{
 		tranche_ids[0] = LWLockNewTrancheId();
-	hnsw_lock_tranche_id = tranche_ids[0];
+		tranche_ids[1] = LWLockNewTrancheId();
+		tranche_ids[2] = LWLockNewTrancheId();
+		tranche_ids[3] = LWLockNewTrancheId();
+	}
+	hnsw_entry_lock_tranche_id = tranche_ids[0];
+	hnsw_allocator_lock_tranche_id = tranche_ids[1];
+	hnsw_flush_lock_tranche_id = tranche_ids[2];
+	hnsw_element_lock_tranche_id = tranche_ids[3];
 	LWLockRelease(AddinShmemInitLock);
 
 	/* Per-backend registration of the tranche ID */
-	LWLockRegisterTranche(hnsw_lock_tranche_id, "HnswBuild");
+	LWLockRegisterTranche(hnsw_entry_lock_tranche_id, "HnswEntry");
+	LWLockRegisterTranche(hnsw_allocator_lock_tranche_id, "HnswAllocator");
+	LWLockRegisterTranche(hnsw_flush_lock_tranche_id, "HnswFlush");
+	LWLockRegisterTranche(hnsw_element_lock_tranche_id, "HnswElement");
 }
 
 /*

src/hnsw.h

@@ -113,7 +113,10 @@
 
 /* Variables */
 extern int	hnsw_ef_search;
-extern int	hnsw_lock_tranche_id;
+extern int	hnsw_entry_lock_tranche_id;
+extern int	hnsw_allocator_lock_tranche_id;
+extern int	hnsw_flush_lock_tranche_id;
+extern int	hnsw_element_lock_tranche_id;
 
 typedef struct HnswElementData HnswElementData;
 typedef struct HnswNeighborArray HnswNeighborArray;

src/hnswbuild.c

@@ -400,7 +400,7 @@ UpdateNeighborsInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswEleme
  * Update graph in memory
  */
 static void
-UpdateGraphInMemory(FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, bool updateEntryPoint, HnswBuildState * buildstate)
+UpdateGraphInMemory(FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, HnswBuildState * buildstate)
 {
 	HnswGraph  *graph = buildstate->graph;
 	char	   *base = buildstate->hnswarea;
@@ -416,7 +416,7 @@ UpdateGraphInMemory(FmgrInfo *procinfo, Oid collation, HnswElement element, int
 	UpdateNeighborsInMemory(base, procinfo, collation, element, m);
 
 	/* Update entry point if needed (already have lock) */
-	if (updateEntryPoint)
+	if (entryPoint == NULL || element->level > entryPoint->level)
 		HnswPtrStore(base, graph->entryPoint, element);
 }
 
@@ -434,29 +434,32 @@ InsertTupleInMemory(HnswBuildState * buildstate, HnswElement element)
 	int			efConstruction = buildstate->efConstruction;
 	int			m = buildstate->m;
 	char	   *base = buildstate->hnswarea;
-	bool		updateEntryPoint;
 
 	/* Get entry point */
-	LWLockAcquire(entryLock, LW_EXCLUSIVE);
+	LWLockAcquire(entryLock, LW_SHARED);
 	entryPoint = HnswPtrAccess(base, graph->entryPoint);
 
-	/* Pause new inserts when updating entry point */
-	/* May still be in-flight inserts */
-	updateEntryPoint = entryPoint == NULL || element->level > entryPoint->level;
-
-	/* Release entry lock */
-	if (!updateEntryPoint)
+	/* Prevent concurrent inserts when likely updating entry point */
+	if (entryPoint == NULL || element->level > entryPoint->level)
+	{
+		/* Release shared lock */
 		LWLockRelease(entryLock);
 
+		/* Get exclusive lock */
+		LWLockAcquire(entryLock, LW_EXCLUSIVE);
+
+		/* Get latest entry point after lock is acquired */
+		entryPoint = HnswPtrAccess(base, graph->entryPoint);
+	}
+
 	/* Find neighbors for element */
 	HnswFindElementNeighbors(base, element, entryPoint, NULL, procinfo, collation, m, efConstruction, false);
 
 	/* Update graph in memory */
-	UpdateGraphInMemory(procinfo, collation, element, m, efConstruction, updateEntryPoint, buildstate);
+	UpdateGraphInMemory(procinfo, collation, element, m, efConstruction, entryPoint, buildstate);
 
 	/* Release entry lock */
-	if (updateEntryPoint)
-		LWLockRelease(entryLock);
+	LWLockRelease(entryLock);
 }
 
 /*
@@ -545,7 +548,7 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
 	HnswPtrStore(base, element->value, valuePtr);
 
 	/* Create a lock for the element */
-	LWLockInitialize(&element->lock, hnsw_lock_tranche_id);
+	LWLockInitialize(&element->lock, hnsw_element_lock_tranche_id);
 
 	/* Insert tuple */
 	InsertTupleInMemory(buildstate, element);
@@ -608,9 +611,9 @@ InitGraph(HnswGraph * graph, char *base, long memoryTotal)
 	graph->flushed = false;
 	graph->indtuples = 0;
 	SpinLockInit(&graph->lock);
-	LWLockInitialize(&graph->entryLock, hnsw_lock_tranche_id);
-	LWLockInitialize(&graph->allocatorLock, hnsw_lock_tranche_id);
-	LWLockInitialize(&graph->flushLock, hnsw_lock_tranche_id);
+	LWLockInitialize(&graph->entryLock, hnsw_entry_lock_tranche_id);
+	LWLockInitialize(&graph->allocatorLock, hnsw_allocator_lock_tranche_id);
+	LWLockInitialize(&graph->flushLock, hnsw_flush_lock_tranche_id);
 }
 
 /*