Fix rare assertion failure in parallel hash join.

anarazel · anarazel · commit f83040c62a78 · 2017-12-28T02:41:53.000-08:00
When a backend runs out of inner tuples to hash, it should detach from grow_batch_barrier only after it has flushed all batches to disk and merged counters, not before. Otherwise a concurrent backend in ExecParallelHashIncreaseNumBatches() could stop waiting for this backend and try to read tuples before they have been written. This commit reorders those operations and should fix the assertion failures seen occasionally on the build farm since commit 1804284. Author: Thomas Munro Discussion: https://postgr.es/m/E1eRwXy-0004IK-TO%40gemulon.postgresql.org
diff --git a/src/backend/executor/nodeHash.c b/src/backend/executor/nodeHash.c
@@ -288,8 +288,6 @@ MultiExecParallelHash(HashState *node)
 					ExecParallelHashTableInsert(hashtable, slot, hashvalue);
 				hashtable->partialTuples++;
 			}
-			BarrierDetach(&pstate->grow_buckets_barrier);
-			BarrierDetach(&pstate->grow_batches_barrier);
 
 			/*
 			 * Make sure that any tuples we wrote to disk are visible to
@@ -304,6 +302,9 @@ MultiExecParallelHash(HashState *node)
 			 */
 			ExecParallelHashMergeCounters(hashtable);
 
+			BarrierDetach(&pstate->grow_buckets_barrier);
+			BarrierDetach(&pstate->grow_batches_barrier);
+
 			/*
 			 * Wait for everyone to finish building and flushing files and
 			 * counters.
