Fix state reversal after partition tuple routing

We make some changes to ModifyTableState and the EState it uses whenever
we route tuples to partitions; but we weren't restoring properly in all
cases, possibly causing crashes when partitions with different tuple
descriptors are targeted by tuples inserted in the same command.
Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the
needed state changing logic, and have it invoked one level above its
current place (ie. put it in ExecModifyTable instead of ExecInsert);
this makes it all more readable.

Add a test case to exercise this.

We don't support having views as partitions; and since only views can
have INSTEAD OF triggers, there is no point in testing for INSTEAD OF
when processing insertions into a partitioned table.  Remove code that
appears to support this (but which is actually never relevant.)

In passing, fix location of some very confusing comments in
ModifyTableState.

Reported-by: Amit Langote
Author: Etsuro Fujita, Amit Langote
Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp

diff --git a/src/backend/commands/copy.c b/src/backend/commands/copy.c
index 4562a5121d4d287a48fe4c6d5b7de39892a3c144..a42861da0df7da4858b8047471602916944835a5 100644 (file)
--- a/src/backend/commands/copy.c
+++ b/src/backend/commands/copy.c
@@ -2633,13 +2633,12 @@ CopyFrom(CopyState cstate)
                        if (cstate->transition_capture != NULL)
                        {
                                if (resultRelInfo->ri_TrigDesc &&
-                                       (resultRelInfo->ri_TrigDesc->trig_insert_before_row ||
-                                        resultRelInfo->ri_TrigDesc->trig_insert_instead_row))
+                                       resultRelInfo->ri_TrigDesc->trig_insert_before_row)
                                {
                                        /*
-                                        * If there are any BEFORE or INSTEAD triggers on the
-                                        * partition, we'll have to be ready to convert their
-                                        * result back to tuplestore format.
+                                        * If there are any BEFORE triggers on the partition,
+                                        * we'll have to be ready to convert their result back to
+                                        * tuplestore format.
                                         */
                                        cstate->transition_capture->tcs_original_insert_tuple = NULL;
                                        cstate->transition_capture->tcs_map =
@@ -2768,12 +2767,13 @@ CopyFrom(CopyState cstate)
                         * tuples inserted by an INSERT command.
                         */
                        processed++;
+               }
 
-                       if (saved_resultRelInfo)
-                       {
-                               resultRelInfo = saved_resultRelInfo;
-                               estate->es_result_relation_info = resultRelInfo;
-                       }
+               /* Restore the saved ResultRelInfo */
+               if (saved_resultRelInfo)
+               {
+                       resultRelInfo = saved_resultRelInfo;
+                       estate->es_result_relation_info = resultRelInfo;
                }
        }
 

diff --git a/src/backend/executor/nodeModifyTable.c b/src/backend/executor/nodeModifyTable.c
index 3332ae4bf3d65217d08b1a21eb87ec5546a47793..9260236e4d93faec9eb508d2882f371527521f73 100644 (file)
--- a/src/backend/executor/nodeModifyTable.c
+++ b/src/backend/executor/nodeModifyTable.c
@@ -62,6 +62,11 @@ static bool ExecOnConflictUpdate(ModifyTableState *mtstate,
                                         EState *estate,
                                         bool canSetTag,
                                         TupleTableSlot **returning);
+static TupleTableSlot *ExecPrepareTupleRouting(ModifyTableState *mtstate,
+                                               EState *estate,
+                                               PartitionTupleRouting *proute,
+                                               ResultRelInfo *targetRelInfo,
+                                               TupleTableSlot *slot);
 static ResultRelInfo *getTargetResultRelInfo(ModifyTableState *node);
 static void ExecSetupChildParentMapForTcs(ModifyTableState *mtstate);
 static void ExecSetupChildParentMapForSubplan(ModifyTableState *mtstate);
@@ -265,7 +270,6 @@ ExecInsert(ModifyTableState *mtstate,
 {
        HeapTuple       tuple;
        ResultRelInfo *resultRelInfo;
-       ResultRelInfo *saved_resultRelInfo = NULL;
        Relation        resultRelationDesc;
        Oid                     newId;
        List       *recheckIndexes = NIL;
@@ -282,100 +286,6 @@ ExecInsert(ModifyTableState *mtstate,
         * get information on the (current) result relation
         */
        resultRelInfo = estate->es_result_relation_info;
-
-       /* Determine the partition to heap_insert the tuple into */
-       if (mtstate->mt_partition_tuple_routing)
-       {
-               int                     leaf_part_index;
-               PartitionTupleRouting *proute = mtstate->mt_partition_tuple_routing;
-
-               /*
-                * Away we go ... If we end up not finding a partition after all,
-                * ExecFindPartition() does not return and errors out instead.
-                * Otherwise, the returned value is to be used as an index into arrays
-                * proute->partitions[] and proute->partition_tupconv_maps[] that will
-                * get us the ResultRelInfo and TupleConversionMap for the partition,
-                * respectively.
-                */
-               leaf_part_index = ExecFindPartition(resultRelInfo,
-                                                                                       proute->partition_dispatch_info,
-                                                                                       slot,
-                                                                                       estate);
-               Assert(leaf_part_index >= 0 &&
-                          leaf_part_index < proute->num_partitions);
-
-               /*
-                * Save the old ResultRelInfo and switch to the one corresponding to
-                * the selected partition.  (We might need to initialize it first.)
-                */
-               saved_resultRelInfo = resultRelInfo;
-               resultRelInfo = proute->partitions[leaf_part_index];
-               if (resultRelInfo == NULL)
-               {
-                       resultRelInfo = ExecInitPartitionInfo(mtstate,
-                                                                                                 saved_resultRelInfo,
-                                                                                                 proute, estate,
-                                                                                                 leaf_part_index);
-                       Assert(resultRelInfo != NULL);
-               }
-
-               /* We do not yet have a way to insert into a foreign partition */
-               if (resultRelInfo->ri_FdwRoutine)
-                       ereport(ERROR,
-                                       (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
-                                        errmsg("cannot route inserted tuples to a foreign table")));
-
-               /* For ExecInsertIndexTuples() to work on the partition's indexes */
-               estate->es_result_relation_info = resultRelInfo;
-
-               /*
-                * If we're capturing transition tuples, we might need to convert from
-                * the partition rowtype to parent rowtype.
-                */
-               if (mtstate->mt_transition_capture != NULL)
-               {
-                       if (resultRelInfo->ri_TrigDesc &&
-                               (resultRelInfo->ri_TrigDesc->trig_insert_before_row ||
-                                resultRelInfo->ri_TrigDesc->trig_insert_instead_row))
-                       {
-                               /*
-                                * If there are any BEFORE or INSTEAD triggers on the
-                                * partition, we'll have to be ready to convert their result
-                                * back to tuplestore format.
-                                */
-                               mtstate->mt_transition_capture->tcs_original_insert_tuple = NULL;
-
-                               mtstate->mt_transition_capture->tcs_map =
-                                       TupConvMapForLeaf(proute, saved_resultRelInfo,
-                                                                         leaf_part_index);
-                       }
-                       else
-                       {
-                               /*
-                                * Otherwise, just remember the original unconverted tuple, to
-                                * avoid a needless round trip conversion.
-                                */
-                               mtstate->mt_transition_capture->tcs_original_insert_tuple = tuple;
-                               mtstate->mt_transition_capture->tcs_map = NULL;
-                       }
-               }
-               if (mtstate->mt_oc_transition_capture != NULL)
-               {
-                       mtstate->mt_oc_transition_capture->tcs_map =
-                               TupConvMapForLeaf(proute, saved_resultRelInfo,
-                                                                 leaf_part_index);
-               }
-
-               /*
-                * We might need to convert from the parent rowtype to the partition
-                * rowtype.
-                */
-               tuple = ConvertPartitionTupleSlot(proute->parent_child_tupconv_maps[leaf_part_index],
-                                                                                 tuple,
-                                                                                 proute->partition_tuple_slot,
-                                                                                 &slot);
-       }
-
        resultRelationDesc = resultRelInfo->ri_RelationDesc;
 
        /*
@@ -495,7 +405,7 @@ ExecInsert(ModifyTableState *mtstate,
                 * No need though if the tuple has been routed, and a BR trigger
                 * doesn't exist.
                 */
-               if (saved_resultRelInfo != NULL &&
+               if (resultRelInfo->ri_PartitionRoot != NULL &&
                        !(resultRelInfo->ri_TrigDesc &&
                          resultRelInfo->ri_TrigDesc->trig_insert_before_row))
                        check_partition_constr = false;
@@ -686,9 +596,6 @@ ExecInsert(ModifyTableState *mtstate,
        if (resultRelInfo->ri_projectReturning)
                result = ExecProcessReturning(resultRelInfo, slot, planSlot);
 
-       if (saved_resultRelInfo)
-               estate->es_result_relation_info = saved_resultRelInfo;
-
        return result;
 }
 
@@ -1209,27 +1116,22 @@ lreplace:;
                                                                                          proute->root_tuple_slot,
                                                                                          &slot);
 
-
-                       /*
-                        * For ExecInsert(), make it look like we are inserting into the
-                        * root.
-                        */
+                       /* Prepare for tuple routing */
                        Assert(mtstate->rootResultRelInfo != NULL);
-                       estate->es_result_relation_info = mtstate->rootResultRelInfo;
+                       slot = ExecPrepareTupleRouting(mtstate, estate, proute,
+                                                                                  mtstate->rootResultRelInfo, slot);
 
                        ret_slot = ExecInsert(mtstate, slot, planSlot, NULL,
                                                                  ONCONFLICT_NONE, estate, canSetTag);
 
-                       /*
-                        * Revert back the active result relation and the active
-                        * transition capture map that we changed above.
-                        */
+                       /* Revert ExecPrepareTupleRouting's node change. */
                        estate->es_result_relation_info = resultRelInfo;
                        if (mtstate->mt_transition_capture)
                        {
                                mtstate->mt_transition_capture->tcs_original_insert_tuple = NULL;
                                mtstate->mt_transition_capture->tcs_map = saved_tcs_map;
                        }
+
                        return ret_slot;
                }
 
@@ -1710,6 +1612,108 @@ ExecSetupTransitionCaptureState(ModifyTableState *mtstate, EState *estate)
        }
 }
 
+/*
+ * ExecPrepareTupleRouting --- prepare for routing one tuple
+ *
+ * Determine the partition in which the tuple in slot is to be inserted,
+ * and modify mtstate and estate to prepare for it.
+ *
+ * Caller must revert the estate changes after executing the insertion!
+ * In mtstate, transition capture changes may also need to be reverted.
+ *
+ * Returns a slot holding the tuple of the partition rowtype.
+ */
+static TupleTableSlot *
+ExecPrepareTupleRouting(ModifyTableState *mtstate,
+                                               EState *estate,
+                                               PartitionTupleRouting *proute,
+                                               ResultRelInfo *targetRelInfo,
+                                               TupleTableSlot *slot)
+{
+       int                     partidx;
+       ResultRelInfo *partrel;
+       HeapTuple       tuple;
+
+       /*
+        * Determine the target partition.  If ExecFindPartition does not find
+        * a partition after all, it doesn't return here; otherwise, the returned
+        * value is to be used as an index into the arrays for the ResultRelInfo
+        * and TupleConversionMap for the partition.
+        */
+       partidx = ExecFindPartition(targetRelInfo,
+                                                               proute->partition_dispatch_info,
+                                                               slot,
+                                                               estate);
+       Assert(partidx >= 0 && partidx < proute->num_partitions);
+
+       /*
+        * Get the ResultRelInfo corresponding to the selected partition; if not
+        * yet there, initialize it.
+        */
+       partrel = proute->partitions[partidx];
+       if (partrel == NULL)
+               partrel = ExecInitPartitionInfo(mtstate, targetRelInfo,
+                                                                               proute, estate,
+                                                                               partidx);
+
+       /* We do not yet have a way to insert into a foreign partition */
+       if (partrel->ri_FdwRoutine)
+               ereport(ERROR,
+                               (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+                                errmsg("cannot route inserted tuples to a foreign table")));
+
+       /*
+        * Make it look like we are inserting into the partition.
+        */
+       estate->es_result_relation_info = partrel;
+
+       /* Get the heap tuple out of the given slot. */
+       tuple = ExecMaterializeSlot(slot);
+
+       /*
+        * If we're capturing transition tuples, we might need to convert from the
+        * partition rowtype to parent rowtype.
+        */
+       if (mtstate->mt_transition_capture != NULL)
+       {
+               if (partrel->ri_TrigDesc &&
+                       partrel->ri_TrigDesc->trig_insert_before_row)
+               {
+                       /*
+                        * If there are any BEFORE triggers on the partition, we'll have
+                        * to be ready to convert their result back to tuplestore format.
+                        */
+                       mtstate->mt_transition_capture->tcs_original_insert_tuple = NULL;
+                       mtstate->mt_transition_capture->tcs_map =
+                               TupConvMapForLeaf(proute, targetRelInfo, partidx);
+               }
+               else
+               {
+                       /*
+                        * Otherwise, just remember the original unconverted tuple, to
+                        * avoid a needless round trip conversion.
+                        */
+                       mtstate->mt_transition_capture->tcs_original_insert_tuple = tuple;
+                       mtstate->mt_transition_capture->tcs_map = NULL;
+               }
+       }
+       if (mtstate->mt_oc_transition_capture != NULL)
+       {
+               mtstate->mt_oc_transition_capture->tcs_map =
+                       TupConvMapForLeaf(proute, targetRelInfo, partidx);
+       }
+
+       /*
+        * Convert the tuple, if necessary.
+        */
+       ConvertPartitionTupleSlot(proute->parent_child_tupconv_maps[partidx],
+                                                         tuple,
+                                                         proute->partition_tuple_slot,
+                                                         &slot);
+
+       return slot;
+}
+
 /*
  * Initialize the child-to-root tuple conversion map array for UPDATE subplans.
  *
@@ -1846,6 +1850,7 @@ static TupleTableSlot *
 ExecModifyTable(PlanState *pstate)
 {
        ModifyTableState *node = castNode(ModifyTableState, pstate);
+       PartitionTupleRouting *proute = node->mt_partition_tuple_routing;
        EState     *estate = node->ps.state;
        CmdType         operation = node->operation;
        ResultRelInfo *saved_resultRelInfo;
@@ -2051,9 +2056,16 @@ ExecModifyTable(PlanState *pstate)
                switch (operation)
                {
                        case CMD_INSERT:
+                               /* Prepare for tuple routing if needed. */
+                               if (proute)
+                                       slot = ExecPrepareTupleRouting(node, estate, proute,
+                                                                                                  resultRelInfo, slot);
                                slot = ExecInsert(node, slot, planSlot,
                                                                  node->mt_arbiterindexes, node->mt_onconflict,
                                                                  estate, node->canSetTag);
+                               /* Revert ExecPrepareTupleRouting's state change. */
+                               if (proute)
+                                       estate->es_result_relation_info = resultRelInfo;
                                break;
                        case CMD_UPDATE:
                                slot = ExecUpdate(node, tupleid, oldtuple, slot, planSlot,

diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index a953820f43a5746db06e4f8cea10a8142fc3ef6c..b119ef861b9902d1baf7e9ad5d7ece8345616ed1 100644 (file)
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -995,14 +995,18 @@ typedef struct ModifyTableState
        TupleTableSlot *mt_existing;    /* slot to store existing target tuple in */
        List       *mt_excludedtlist;   /* the excluded pseudo relation's tlist  */
        TupleTableSlot *mt_conflproj;   /* CONFLICT ... SET ... projection target */
-       struct PartitionTupleRouting *mt_partition_tuple_routing;
+
        /* Tuple-routing support info */
-       struct TransitionCaptureState *mt_transition_capture;
+       struct PartitionTupleRouting *mt_partition_tuple_routing;
+
        /* controls transition table population for specified operation */
-       struct TransitionCaptureState *mt_oc_transition_capture;
+       struct TransitionCaptureState *mt_transition_capture;
+
        /* controls transition table population for INSERT...ON CONFLICT UPDATE */
-       TupleConversionMap **mt_per_subplan_tupconv_maps;
+       struct TransitionCaptureState *mt_oc_transition_capture;
+
        /* Per plan map for tuple conversion from child to root */
+       TupleConversionMap **mt_per_subplan_tupconv_maps;
 } ModifyTableState;
 
 /* ----------------

diff --git a/src/test/regress/expected/insert.out b/src/test/regress/expected/insert.out
index 2ca085e1ba31e09d1a4893cb8345d2eba24a5ea7..97419a744ff4efe922eec5bfd9ddda4a097ce578 100644 (file)
--- a/src/test/regress/expected/insert.out
+++ b/src/test/regress/expected/insert.out
@@ -748,6 +748,32 @@ drop role regress_coldesc_role;
 drop table inserttest3;
 drop table brtrigpartcon;
 drop function brtrigpartcon1trigf();
+-- check that "do nothing" BR triggers work with tuple-routing (this checks
+-- that estate->es_result_relation_info is appropriately set/reset for each
+-- routed tuple)
+create table donothingbrtrig_test (a int, b text) partition by list (a);
+create table donothingbrtrig_test1 (b text, a int);
+create table donothingbrtrig_test2 (c text, b text, a int);
+alter table donothingbrtrig_test2 drop column c;
+create or replace function donothingbrtrig_func() returns trigger as $$begin raise notice 'b: %', new.b; return NULL; end$$ language plpgsql;
+create trigger donothingbrtrig1 before insert on donothingbrtrig_test1 for each row execute procedure donothingbrtrig_func();
+create trigger donothingbrtrig2 before insert on donothingbrtrig_test2 for each row execute procedure donothingbrtrig_func();
+alter table donothingbrtrig_test attach partition donothingbrtrig_test1 for values in (1);
+alter table donothingbrtrig_test attach partition donothingbrtrig_test2 for values in (2);
+insert into donothingbrtrig_test values (1, 'foo'), (2, 'bar');
+NOTICE:  b: foo
+NOTICE:  b: bar
+copy donothingbrtrig_test from stdout;
+NOTICE:  b: baz
+NOTICE:  b: qux
+select tableoid::regclass, * from donothingbrtrig_test;
+ tableoid | a | b 
+----------+---+---
+(0 rows)
+
+-- cleanup
+drop table donothingbrtrig_test;
+drop function donothingbrtrig_func();
 -- check multi-column range partitioning with minvalue/maxvalue constraints
 create table mcrparted (a text, b int) partition by range(a, b);
 create table mcrparted1_lt_b partition of mcrparted for values from (minvalue, minvalue) to ('b', minvalue);

diff --git a/src/test/regress/sql/insert.sql b/src/test/regress/sql/insert.sql
index 3e3fd96c75758c99e76bb050fe8ecb99ee344028..a16f2a7f8901f2d402c1924f1ded52e5a7abc67b 100644 (file)
--- a/src/test/regress/sql/insert.sql
+++ b/src/test/regress/sql/insert.sql
@@ -489,6 +489,29 @@ drop table inserttest3;
 drop table brtrigpartcon;
 drop function brtrigpartcon1trigf();
 
+-- check that "do nothing" BR triggers work with tuple-routing (this checks
+-- that estate->es_result_relation_info is appropriately set/reset for each
+-- routed tuple)
+create table donothingbrtrig_test (a int, b text) partition by list (a);
+create table donothingbrtrig_test1 (b text, a int);
+create table donothingbrtrig_test2 (c text, b text, a int);
+alter table donothingbrtrig_test2 drop column c;
+create or replace function donothingbrtrig_func() returns trigger as $$begin raise notice 'b: %', new.b; return NULL; end$$ language plpgsql;
+create trigger donothingbrtrig1 before insert on donothingbrtrig_test1 for each row execute procedure donothingbrtrig_func();
+create trigger donothingbrtrig2 before insert on donothingbrtrig_test2 for each row execute procedure donothingbrtrig_func();
+alter table donothingbrtrig_test attach partition donothingbrtrig_test1 for values in (1);
+alter table donothingbrtrig_test attach partition donothingbrtrig_test2 for values in (2);
+insert into donothingbrtrig_test values (1, 'foo'), (2, 'bar');
+copy donothingbrtrig_test from stdout;
+1      baz
+2      qux
+\.
+select tableoid::regclass, * from donothingbrtrig_test;
+
+-- cleanup
+drop table donothingbrtrig_test;
+drop function donothingbrtrig_func();
+
 -- check multi-column range partitioning with minvalue/maxvalue constraints
 create table mcrparted (a text, b int) partition by range(a, b);
 create table mcrparted1_lt_b partition of mcrparted for values from (minvalue, minvalue) to ('b', minvalue);