Move ExecProcNode from dispatch to function pointer based model.

This allows us to add stack-depth checks the first time an executor
node is called, and skip that overhead on following
calls. Additionally it yields a nice speedup.

While it'd probably have been a good idea to have that check all
along, it has become more important after the new expression
evaluation framework in b8d7f053c5c2bf2a7e - there's no stack depth
check in common paths anymore now. We previously relied on
ExecEvalExpr() being executed somewhere.

We should move towards that model for further routines, but as this is
required for v10, it seems better to only do the necessary (which
already is quite large).

Author: Andres Freund, Tom Lane
Reported-By: Julien Rouhaud
Discussion:
    https://postgr.es/m/[email protected]
    https://postgr.es/m/[email protected]

1 files changed, 66 insertions, 186 deletions

diff --git a/src/backend/executor/execProcnode.c b/src/backend/executor/execProcnode.c
index 20fd9f822ed..396920c0a23 100644
--- a/src/backend/executor/execProcnode.c
+++ b/src/backend/executor/execProcnode.c
@@ -17,15 +17,10 @@
  *-------------------------------------------------------------------------
  */
 /*
- *	 INTERFACE ROUTINES
- *		ExecInitNode	-		initialize a plan node and its subplans
- *		ExecProcNode	-		get a tuple by executing the plan node
- *		ExecEndNode		-		shut down a plan node and its subplans
- *
  *	 NOTES
  *		This used to be three files.  It is now all combined into
- *		one file so that it is easier to keep ExecInitNode, ExecProcNode,
- *		and ExecEndNode in sync when new nodes are added.
+ *		one file so that it is easier to keep the dispatch routines
+ *		in sync when new nodes are added.
  *
  *	 EXAMPLE
  *		Suppose we want the age of the manager of the shoe department and
@@ -122,6 +117,10 @@
 #include "miscadmin.h"
 
 
+static TupleTableSlot *ExecProcNodeFirst(PlanState *node);
+static TupleTableSlot *ExecProcNodeInstr(PlanState *node);
+
+
 /* ------------------------------------------------------------------------
  *		ExecInitNode
  *
@@ -149,6 +148,13 @@ ExecInitNode(Plan *node, EState *estate, int eflags)
 	if (node == NULL)
 		return NULL;
 
+	/*
+	 * Make sure there's enough stack available. Need to check here, in
+	 * addition to ExecProcNode() (via ExecProcNodeFirst()), to ensure the
+	 * stack isn't overrun while initializing the node tree.
+	 */
+	check_stack_depth();
+
 	switch (nodeTag(node))
 	{
 			/*
@@ -365,6 +371,13 @@ ExecInitNode(Plan *node, EState *estate, int eflags)
 	}
 
 	/*
+	 * Add a wrapper around the ExecProcNode callback that checks stack depth
+	 * during the first execution.
+	 */
+	result->ExecProcNodeReal = result->ExecProcNode;
+	result->ExecProcNode = ExecProcNodeFirst;
+
+	/*
 	 * Initialize any initPlans present in this node.  The planner put them in
 	 * a separate list for us.
 	 */
@@ -388,195 +401,51 @@ ExecInitNode(Plan *node, EState *estate, int eflags)
 }
 
 
-/* ----------------------------------------------------------------
- *		ExecProcNode
- *
- *		Execute the given node to return a(nother) tuple.
- * ----------------------------------------------------------------
+/*
+ * ExecProcNode wrapper that performs some one-time checks, before calling
+ * the relevant node method (possibly via an instrumentation wrapper).
  */
-TupleTableSlot *
-ExecProcNode(PlanState *node)
+static TupleTableSlot *
+ExecProcNodeFirst(PlanState *node)
 {
-	TupleTableSlot *result;
-
-	if (node->chgParam != NULL) /* something changed */
-		ExecReScan(node);		/* let ReScan handle this */
+	/*
+	 * Perform stack depth check during the first execution of the node.  We
+	 * only do so the first time round because it turns out to not be cheap on
+	 * some common architectures (eg. x86).  This relies on the assumption that
+	 * ExecProcNode calls for a given plan node will always be made at roughly
+	 * the same stack depth.
+	 */
+	check_stack_depth();
 
+	/*
+	 * If instrumentation is required, change the wrapper to one that just
+	 * does instrumentation.  Otherwise we can dispense with all wrappers and
+	 * have ExecProcNode() directly call the relevant function from now on.
+	 */
 	if (node->instrument)
-		InstrStartNode(node->instrument);
-
-	switch (nodeTag(node))
-	{
-			/*
-			 * control nodes
-			 */
-		case T_ResultState:
-			result = ExecResult((ResultState *) node);
-			break;
-
-		case T_ProjectSetState:
-			result = ExecProjectSet((ProjectSetState *) node);
-			break;
-
-		case T_ModifyTableState:
-			result = ExecModifyTable((ModifyTableState *) node);
-			break;
-
-		case T_AppendState:
-			result = ExecAppend((AppendState *) node);
-			break;
-
-		case T_MergeAppendState:
-			result = ExecMergeAppend((MergeAppendState *) node);
-			break;
-
-		case T_RecursiveUnionState:
-			result = ExecRecursiveUnion((RecursiveUnionState *) node);
-			break;
-
-			/* BitmapAndState does not yield tuples */
-
-			/* BitmapOrState does not yield tuples */
-
-			/*
-			 * scan nodes
-			 */
-		case T_SeqScanState:
-			result = ExecSeqScan((SeqScanState *) node);
-			break;
-
-		case T_SampleScanState:
-			result = ExecSampleScan((SampleScanState *) node);
-			break;
-
-		case T_IndexScanState:
-			result = ExecIndexScan((IndexScanState *) node);
-			break;
-
-		case T_IndexOnlyScanState:
-			result = ExecIndexOnlyScan((IndexOnlyScanState *) node);
-			break;
-
-			/* BitmapIndexScanState does not yield tuples */
-
-		case T_BitmapHeapScanState:
-			result = ExecBitmapHeapScan((BitmapHeapScanState *) node);
-			break;
-
-		case T_TidScanState:
-			result = ExecTidScan((TidScanState *) node);
-			break;
-
-		case T_SubqueryScanState:
-			result = ExecSubqueryScan((SubqueryScanState *) node);
-			break;
-
-		case T_FunctionScanState:
-			result = ExecFunctionScan((FunctionScanState *) node);
-			break;
-
-		case T_TableFuncScanState:
-			result = ExecTableFuncScan((TableFuncScanState *) node);
-			break;
-
-		case T_ValuesScanState:
-			result = ExecValuesScan((ValuesScanState *) node);
-			break;
-
-		case T_CteScanState:
-			result = ExecCteScan((CteScanState *) node);
-			break;
-
-		case T_NamedTuplestoreScanState:
-			result = ExecNamedTuplestoreScan((NamedTuplestoreScanState *) node);
-			break;
-
-		case T_WorkTableScanState:
-			result = ExecWorkTableScan((WorkTableScanState *) node);
-			break;
-
-		case T_ForeignScanState:
-			result = ExecForeignScan((ForeignScanState *) node);
-			break;
-
-		case T_CustomScanState:
-			result = ExecCustomScan((CustomScanState *) node);
-			break;
-
-			/*
-			 * join nodes
-			 */
-		case T_NestLoopState:
-			result = ExecNestLoop((NestLoopState *) node);
-			break;
-
-		case T_MergeJoinState:
-			result = ExecMergeJoin((MergeJoinState *) node);
-			break;
-
-		case T_HashJoinState:
-			result = ExecHashJoin((HashJoinState *) node);
-			break;
-
-			/*
-			 * materialization nodes
-			 */
-		case T_MaterialState:
-			result = ExecMaterial((MaterialState *) node);
-			break;
-
-		case T_SortState:
-			result = ExecSort((SortState *) node);
-			break;
-
-		case T_GroupState:
-			result = ExecGroup((GroupState *) node);
-			break;
+		node->ExecProcNode = ExecProcNodeInstr;
+	else
+		node->ExecProcNode = node->ExecProcNodeReal;
 
-		case T_AggState:
-			result = ExecAgg((AggState *) node);
-			break;
-
-		case T_WindowAggState:
-			result = ExecWindowAgg((WindowAggState *) node);
-			break;
-
-		case T_UniqueState:
-			result = ExecUnique((UniqueState *) node);
-			break;
-
-		case T_GatherState:
-			result = ExecGather((GatherState *) node);
-			break;
-
-		case T_GatherMergeState:
-			result = ExecGatherMerge((GatherMergeState *) node);
-			break;
-
-		case T_HashState:
-			result = ExecHash((HashState *) node);
-			break;
+	return node->ExecProcNode(node);
+}
 
-		case T_SetOpState:
-			result = ExecSetOp((SetOpState *) node);
-			break;
 
-		case T_LockRowsState:
-			result = ExecLockRows((LockRowsState *) node);
-			break;
+/*
+ * ExecProcNode wrapper that performs instrumentation calls.  By keeping
+ * this a separate function, we avoid overhead in the normal case where
+ * no instrumentation is wanted.
+ */
+static TupleTableSlot *
+ExecProcNodeInstr(PlanState *node)
+{
+	TupleTableSlot *result;
 
-		case T_LimitState:
-			result = ExecLimit((LimitState *) node);
-			break;
+	InstrStartNode(node->instrument);
 
-		default:
-			elog(ERROR, "unrecognized node type: %d", (int) nodeTag(node));
-			result = NULL;
-			break;
-	}
+	result = node->ExecProcNodeReal(node);
 
-	if (node->instrument)
-		InstrStopNode(node->instrument, TupIsNull(result) ? 0.0 : 1.0);
+	InstrStopNode(node->instrument, TupIsNull(result) ? 0.0 : 1.0);
 
 	return result;
 }
@@ -600,6 +469,8 @@ MultiExecProcNode(PlanState *node)
 {
 	Node	   *result;
 
+	check_stack_depth();
+
 	CHECK_FOR_INTERRUPTS();
 
 	if (node->chgParam != NULL) /* something changed */
@@ -657,6 +528,13 @@ ExecEndNode(PlanState *node)
 	if (node == NULL)
 		return;
 
+	/*
+	 * Make sure there's enough stack available. Need to check here, in
+	 * addition to ExecProcNode() (via ExecProcNodeFirst()), because it's not
+	 * guaranteed that ExecProcNode() is reached for all nodes.
+	 */
+	check_stack_depth();
+
 	if (node->chgParam != NULL)
 	{
 		bms_free(node->chgParam);
@@ -855,6 +733,8 @@ ExecShutdownNode(PlanState *node)
 	if (node == NULL)
 		return false;
 
+	check_stack_depth();
+
 	planstate_tree_walker(node, ExecShutdownNode, NULL);
 
 	switch (nodeTag(node))