Change executor to just Assert that table locks were already obtained.

tglsfdc · tglsfdc · commit 9a3cebeaa7fd · 2018-10-03T16:05:12.000-04:00
Instead of locking tables during executor startup, just Assert that suitable locks were obtained already during the parse/plan pipeline (or re-obtained by the plan cache). This must be so, else we have a hazard that concurrent DDL has invalidated the plan. This is pretty inefficient as well as undercommented, but it's all going to go away shortly, so I didn't try hard. This commit is just another attempt to use the buildfarm to see if we've missed anything in the plan to simplify the executor's table management. Note that the change needed here in relation_open() exposes that parallel workers now really are accessing tables without holding any lock of their own, whereas they were not doing that before this commit. This does not give me a warm fuzzy feeling about that aspect of parallel query; it does not seem like a good design, and we now know that it's had exactly no actual testing. I think that we should modify parallel query so that that change can be reverted. Discussion: https://postgr.es/m/468c85d9-540e-66a2-1dde-fec2b741e688@lab.ntt.co.jp
diff --git a/src/backend/access/heap/heapam.c b/src/backend/access/heap/heapam.c
@@ -1140,9 +1140,13 @@ relation_open(Oid relationId, LOCKMODE lockmode)
 	/*
 	 * If we didn't get the lock ourselves, assert that caller holds one,
 	 * except in bootstrap mode where no locks are used.
+	 *
+	 * Also, parallel workers currently assume that their parent holds locks
+	 * for tables used in the parallel query (a mighty shaky assumption).
 	 */
 	Assert(lockmode != NoLock ||
 		   IsBootstrapProcessingMode() ||
+		   IsParallelWorker() ||
 		   CheckRelationLockedByMe(r, AccessShareLock, true));
 
 	/* Make note that we've accessed a temporary relation */
diff --git a/src/backend/executor/execMain.c b/src/backend/executor/execMain.c
@@ -849,8 +849,8 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 			Relation	resultRelation;
 
 			resultRelationOid = getrelid(resultRelationIndex, rangeTable);
-			Assert(rt_fetch(resultRelationIndex, rangeTable)->rellockmode == RowExclusiveLock);
-			resultRelation = heap_open(resultRelationOid, RowExclusiveLock);
+			resultRelation = heap_open(resultRelationOid, NoLock);
+			Assert(CheckRelationLockedByMe(resultRelation, RowExclusiveLock, true));
 
 			InitResultRelInfo(resultRelInfo,
 							  resultRelation,
@@ -890,8 +890,8 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 				Relation	resultRelDesc;
 
 				resultRelOid = getrelid(resultRelIndex, rangeTable);
-				Assert(rt_fetch(resultRelIndex, rangeTable)->rellockmode == RowExclusiveLock);
-				resultRelDesc = heap_open(resultRelOid, RowExclusiveLock);
+				resultRelDesc = heap_open(resultRelOid, NoLock);
+				Assert(CheckRelationLockedByMe(resultRelDesc, RowExclusiveLock, true));
 				InitResultRelInfo(resultRelInfo,
 								  resultRelDesc,
 								  lfirst_int(l),
@@ -903,7 +903,8 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 			estate->es_root_result_relations = resultRelInfos;
 			estate->es_num_root_result_relations = num_roots;
 
-			/* Simply lock the rest of them. */
+			/* Simply check the rest of them are locked. */
+#ifdef USE_ASSERT_CHECKING
 			foreach(l, plannedstmt->nonleafResultRelations)
 			{
 				Index		resultRelIndex = lfirst_int(l);
@@ -912,11 +913,15 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 				if (!list_member_int(plannedstmt->rootResultRelations,
 									 resultRelIndex))
 				{
-					Assert(rt_fetch(resultRelIndex, rangeTable)->rellockmode == RowExclusiveLock);
-					LockRelationOid(getrelid(resultRelIndex, rangeTable),
-									RowExclusiveLock);
+					Relation	resultRelDesc;
+
+					resultRelDesc = heap_open(getrelid(resultRelIndex, rangeTable),
+											  NoLock);
+					Assert(CheckRelationLockedByMe(resultRelDesc, RowExclusiveLock, true));
+					heap_close(resultRelDesc, NoLock);
 				}
 			}
+#endif
 		}
 	}
 	else
@@ -945,7 +950,6 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 	{
 		PlanRowMark *rc = (PlanRowMark *) lfirst(l);
 		Oid			relid;
-		LOCKMODE	rellockmode;
 		Relation	relation;
 		ExecRowMark *erm;
 
@@ -963,8 +967,10 @@ InitPlan(QueryDesc *queryDesc, int eflags)
 			case ROW_MARK_SHARE:
 			case ROW_MARK_KEYSHARE:
 			case ROW_MARK_REFERENCE:
-				rellockmode = rt_fetch(rc->rti, rangeTable)->rellockmode;
-				relation = heap_open(relid, rellockmode);
+				relation = heap_open(relid, NoLock);
+				Assert(CheckRelationLockedByMe(relation,
+											   rt_fetch(rc->rti, rangeTable)->rellockmode,
+											   true));
 				break;
 			case ROW_MARK_COPY:
 				/* no physical table access is required */
diff --git a/src/backend/executor/execUtils.c b/src/backend/executor/execUtils.c
@@ -42,6 +42,7 @@
 
 #include "postgres.h"
 
+#include "access/parallel.h"
 #include "access/relscan.h"
 #include "access/transam.h"
 #include "executor/executor.h"
@@ -648,12 +649,14 @@ ExecOpenScanRelation(EState *estate, Index scanrelid, int eflags)
 {
 	Relation	rel;
 	Oid			reloid;
-	LOCKMODE	lockmode;
 
-	/* Open the relation and acquire lock as needed */
+	/* Open the relation and verify lock was obtained upstream */
 	reloid = getrelid(scanrelid, estate->es_range_table);
-	lockmode = rt_fetch(scanrelid, estate->es_range_table)->rellockmode;
-	rel = heap_open(reloid, lockmode);
+	rel = heap_open(reloid, NoLock);
+	Assert(IsParallelWorker() ||
+		   CheckRelationLockedByMe(rel,
+								   rt_fetch(scanrelid, estate->es_range_table)->rellockmode,
+								   true));
 
 	/*
 	 * Complain if we're attempting a scan of an unscannable relation, except
