From bdee4af8e07648008fe522fc5a562db453be5ad7 Mon Sep 17 00:00:00 2001
From: David Rowley <drowley@postgresql.org>
Date: Fri, 19 Jun 2020 17:25:07 +1200
Subject: Fix EXPLAIN ANALYZE for parallel HashAgg plans

Since 1f39bce02, HashAgg nodes have had the ability to spill to disk when
memory consumption exceeds work_mem. That commit added new properties to
EXPLAIN ANALYZE to show the maximum memory usage and disk usage, however,
it didn't quite go as far as showing that information for parallel
workers.  Since workers may have experienced something very different from
the main process, we should show this information per worker, as is done
in Sort.

Reviewed-by: Justin Pryzby
Reviewed-by: Jeff Davis
Discussion: https://postgr.es/m/CAApHDvpEKbfZa18mM1TD7qV6PG+w97pwCWq5tVD0dX7e11gRJw@mail.gmail.com
Backpatch-through: 13, where the hashagg spilling code was added.
---
 src/backend/executor/nodeAgg.c | 103 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 103 insertions(+)

(limited to 'src/backend/executor/nodeAgg.c')

diff --git a/src/backend/executor/nodeAgg.c b/src/backend/executor/nodeAgg.c
index 331acee2814..a20554ae65a 100644
--- a/src/backend/executor/nodeAgg.c
+++ b/src/backend/executor/nodeAgg.c
@@ -240,6 +240,7 @@
 #include "postgres.h"
 
 #include "access/htup_details.h"
+#include "access/parallel.h"
 #include "catalog/objectaccess.h"
 #include "catalog/pg_aggregate.h"
 #include "catalog/pg_proc.h"
@@ -4483,6 +4484,22 @@ ExecEndAgg(AggState *node)
 	int			numGroupingSets = Max(node->maxsets, 1);
 	int			setno;
 
+	/*
+	 * When ending a parallel worker, copy the statistics gathered by the
+	 * worker back into shared memory so that it can be picked up by the main
+	 * process to report in EXPLAIN ANALYZE.
+	 */
+	if (node->shared_info && IsParallelWorker())
+	{
+		AggregateInstrumentation *si;
+
+		Assert(ParallelWorkerNumber <= node->shared_info->num_workers);
+		si = &node->shared_info->sinstrument[ParallelWorkerNumber];
+		si->hash_batches_used = node->hash_batches_used;
+		si->hash_disk_used = node->hash_disk_used;
+		si->hash_mem_peak = node->hash_mem_peak;
+	}
+
 	/* Make sure we have closed any open tuplesorts */
 
 	if (node->sort_in)
@@ -4854,3 +4871,89 @@ aggregate_dummy(PG_FUNCTION_ARGS)
 		 fcinfo->flinfo->fn_oid);
 	return (Datum) 0;			/* keep compiler quiet */
 }
+
+/* ----------------------------------------------------------------
+ *						Parallel Query Support
+ * ----------------------------------------------------------------
+ */
+
+ /* ----------------------------------------------------------------
+  *		ExecAggEstimate
+  *
+  *		Estimate space required to propagate aggregate statistics.
+  * ----------------------------------------------------------------
+  */
+void
+ExecAggEstimate(AggState *node, ParallelContext *pcxt)
+{
+	Size		size;
+
+	/* don't need this if not instrumenting or no workers */
+	if (!node->ss.ps.instrument || pcxt->nworkers == 0)
+		return;
+
+	size = mul_size(pcxt->nworkers, sizeof(AggregateInstrumentation));
+	size = add_size(size, offsetof(SharedAggInfo, sinstrument));
+	shm_toc_estimate_chunk(&pcxt->estimator, size);
+	shm_toc_estimate_keys(&pcxt->estimator, 1);
+}
+
+/* ----------------------------------------------------------------
+ *		ExecAggInitializeDSM
+ *
+ *		Initialize DSM space for aggregate statistics.
+ * ----------------------------------------------------------------
+ */
+void
+ExecAggInitializeDSM(AggState *node, ParallelContext *pcxt)
+{
+	Size		size;
+
+	/* don't need this if not instrumenting or no workers */
+	if (!node->ss.ps.instrument || pcxt->nworkers == 0)
+		return;
+
+	size = offsetof(SharedAggInfo, sinstrument)
+		+ pcxt->nworkers * sizeof(AggregateInstrumentation);
+	node->shared_info = shm_toc_allocate(pcxt->toc, size);
+	/* ensure any unfilled slots will contain zeroes */
+	memset(node->shared_info, 0, size);
+	node->shared_info->num_workers = pcxt->nworkers;
+	shm_toc_insert(pcxt->toc, node->ss.ps.plan->plan_node_id,
+				   node->shared_info);
+}
+
+/* ----------------------------------------------------------------
+ *		ExecAggInitializeWorker
+ *
+ *		Attach worker to DSM space for aggregate statistics.
+ * ----------------------------------------------------------------
+ */
+void
+ExecAggInitializeWorker(AggState *node, ParallelWorkerContext *pwcxt)
+{
+	node->shared_info =
+		shm_toc_lookup(pwcxt->toc, node->ss.ps.plan->plan_node_id, true);
+}
+
+/* ----------------------------------------------------------------
+ *		ExecAggRetrieveInstrumentation
+ *
+ *		Transfer aggregate statistics from DSM to private memory.
+ * ----------------------------------------------------------------
+ */
+void
+ExecAggRetrieveInstrumentation(AggState *node)
+{
+	Size		size;
+	SharedAggInfo *si;
+
+	if (node->shared_info == NULL)
+		return;
+
+	size = offsetof(SharedAggInfo, sinstrument)
+		+ node->shared_info->num_workers * sizeof(AggregateInstrumentation);
+	si = palloc(size);
+	memcpy(si, node->shared_info, size);
+	node->shared_info = si;
+}
-- 
cgit v1.2.3