Fix rare assertion failure in parallel hash join.

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
1804284042e659e7d16904e7bbb0ad546394b6a3.

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
index 0a519fae313cc6ec532b522486adaff48a38fbf6..04eb3650aa3df37a9f74320bbdd6f7bbd1cc42b1 100644 (file)
--- 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.