execReplication.c File Reference

#include "postgres.h"
#include "access/amapi.h"
#include "access/commit_ts.h"
#include "access/genam.h"
#include "access/gist.h"
#include "access/relscan.h"
#include "access/tableam.h"
#include "access/transam.h"
#include "access/xact.h"
#include "access/heapam.h"
#include "catalog/pg_am_d.h"
#include "commands/trigger.h"
#include "executor/executor.h"
#include "executor/nodeModifyTable.h"
#include "replication/conflict.h"
#include "replication/logicalrelation.h"
#include "storage/lmgr.h"
#include "utils/builtins.h"
#include "utils/lsyscache.h"
#include "utils/rel.h"
#include "utils/snapmgr.h"
#include "utils/syscache.h"
#include "utils/typcache.h"

Include dependency graph for execReplication.c:

Go to the source code of this file.

Functions
static bool	tuples_equal (TupleTableSlot *slot1, TupleTableSlot *slot2, TypeCacheEntry **eq, Bitmapset *columns)
static int	build_replindex_scan_key (ScanKey skey, Relation rel, Relation idxrel, TupleTableSlot *searchslot)
static bool	should_refetch_tuple (TM_Result res, TM_FailureData *tmfd)
bool	RelationFindReplTupleByIndex (Relation rel, Oid idxoid, LockTupleMode lockmode, TupleTableSlot *searchslot, TupleTableSlot *outslot)
bool	RelationFindReplTupleSeq (Relation rel, LockTupleMode lockmode, TupleTableSlot *searchslot, TupleTableSlot *outslot)
static void	BuildConflictIndexInfo (ResultRelInfo *resultRelInfo, Oid conflictindex)
static void	update_most_recent_deletion_info (TupleTableSlot *scanslot, TransactionId oldestxmin, TransactionId *delete_xid, TimestampTz *delete_time, ReplOriginId *delete_origin)
bool	RelationFindDeletedTupleInfoSeq (Relation rel, TupleTableSlot *searchslot, TransactionId oldestxmin, TransactionId *delete_xid, ReplOriginId *delete_origin, TimestampTz *delete_time)
bool	RelationFindDeletedTupleInfoByIndex (Relation rel, Oid idxoid, TupleTableSlot *searchslot, TransactionId oldestxmin, TransactionId *delete_xid, ReplOriginId *delete_origin, TimestampTz *delete_time)
static bool	FindConflictTuple (ResultRelInfo *resultRelInfo, EState *estate, Oid conflictindex, TupleTableSlot *slot, TupleTableSlot **conflictslot)
static void	CheckAndReportConflict (ResultRelInfo *resultRelInfo, EState *estate, ConflictType type, List *recheckIndexes, TupleTableSlot *searchslot, TupleTableSlot *remoteslot)
void	ExecSimpleRelationInsert (ResultRelInfo *resultRelInfo, EState *estate, TupleTableSlot *slot)
void	ExecSimpleRelationUpdate (ResultRelInfo *resultRelInfo, EState *estate, EPQState *epqstate, TupleTableSlot *searchslot, TupleTableSlot *slot)
void	ExecSimpleRelationDelete (ResultRelInfo *resultRelInfo, EState *estate, EPQState *epqstate, TupleTableSlot *searchslot)
void	CheckCmdReplicaIdentity (Relation rel, CmdType cmd)
void	CheckSubscriptionRelkind (char localrelkind, char remoterelkind, const char *nspname, const char *relname)

Function Documentation

build_replindex_scan_key()

static int build_replindex_scan_key ( ScanKey  skey, Relation  rel, Relation  idxrel, TupleTableSlot *  searchslot  )

static

Definition at line 58 of file execReplication.c.

{
    int         index_attoff;
    int         skey_attoff = 0;
    Datum       indclassDatum;
    oidvector  *opclass;
    int2vector *indkey = &idxrel->rd_index->indkey;
 
    indclassDatum = SysCacheGetAttrNotNull(INDEXRELID, idxrel->rd_indextuple,
                                           Anum_pg_index_indclass);
    opclass = (oidvector *) DatumGetPointer(indclassDatum);
 
    /* Build scankey for every non-expression attribute in the index. */
    for (index_attoff = 0; index_attoff < IndexRelationGetNumberOfKeyAttributes(idxrel);
         index_attoff++)
    {
        Oid         operator;
        Oid         optype;
        Oid         opfamily;
        RegProcedure regop;
        int         table_attno = indkey->values[index_attoff];
        StrategyNumber eq_strategy;
 
        if (!AttributeNumberIsValid(table_attno))
        {
            /*
             * XXX: Currently, we don't support expressions in the scan key,
             * see code below.
             */
            continue;
        }
 
        /*
         * Load the operator info.  We need this to get the equality operator
         * function for the scan key.
         */
        optype = get_opclass_input_type(opclass->values[index_attoff]);
        opfamily = get_opclass_family(opclass->values[index_attoff]);
        eq_strategy = IndexAmTranslateCompareType(COMPARE_EQ, idxrel->rd_rel->relam, opfamily, false);
        operator = get_opfamily_member(opfamily, optype,
                                       optype,
                                       eq_strategy);
 
        if (!OidIsValid(operator))
            elog(ERROR, "missing operator %d(%u,%u) in opfamily %u",
                 eq_strategy, optype, optype, opfamily);
 
        regop = get_opcode(operator);
 
        /* Initialize the scankey. */
        ScanKeyInit(&skey[skey_attoff],
                    index_attoff + 1,
                    eq_strategy,
                    regop,
                    searchslot->tts_values[table_attno - 1]);
 
        skey[skey_attoff].sk_collation = idxrel->rd_indcollation[index_attoff];
 
        /* Check for null value. */
        if (searchslot->tts_isnull[table_attno - 1])
            skey[skey_attoff].sk_flags |= (SK_ISNULL | SK_SEARCHNULL);
 
        skey_attoff++;
    }
 
    /* There must always be at least one attribute for the index scan. */
    Assert(skey_attoff > 0);
 
    return skey_attoff;
}

References Assert, AttributeNumberIsValid, COMPARE_EQ, DatumGetPointer(), elog, ERROR, fb(), get_opclass_family(), get_opclass_input_type(), get_opcode(), get_opfamily_member(), IndexAmTranslateCompareType(), IndexRelationGetNumberOfKeyAttributes, OidIsValid, ScanKeyInit(), SK_ISNULL, SK_SEARCHNULL, SysCacheGetAttrNotNull(), and oidvector::values.

Referenced by RelationFindDeletedTupleInfoByIndex(), and RelationFindReplTupleByIndex().

BuildConflictIndexInfo()

static void BuildConflictIndexInfo ( ResultRelInfo *  resultRelInfo, Oid  conflictindex  )

static

Definition at line 454 of file execReplication.c.

{
    for (int i = 0; i < resultRelInfo->ri_NumIndices; i++)
    {
        Relation    indexRelation = resultRelInfo->ri_IndexRelationDescs[i];
        IndexInfo  *indexRelationInfo = resultRelInfo->ri_IndexRelationInfo[i];
 
        if (conflictindex != RelationGetRelid(indexRelation))
            continue;
 
        /*
         * This Assert will fail if BuildSpeculativeIndexInfo() is called
         * twice for the given index.
         */
        Assert(indexRelationInfo->ii_UniqueOps == NULL);
 
        BuildSpeculativeIndexInfo(indexRelation, indexRelationInfo);
    }
}

References Assert, BuildSpeculativeIndexInfo(), fb(), i, RelationGetRelid, ResultRelInfo::ri_IndexRelationDescs, ResultRelInfo::ri_IndexRelationInfo, and ResultRelInfo::ri_NumIndices.

Referenced by FindConflictTuple().

CheckAndReportConflict()

static void CheckAndReportConflict ( ResultRelInfo *  resultRelInfo, EState *  estate, ConflictType  type, List *  recheckIndexes, TupleTableSlot *  searchslot, TupleTableSlot *  remoteslot  )

static

Definition at line 770 of file execReplication.c.

{
    List       *conflicttuples = NIL;
    TupleTableSlot *conflictslot;
 
    /* Check all the unique indexes for conflicts */
    foreach_oid(uniqueidx, resultRelInfo->ri_onConflictArbiterIndexes)
    {
        if (list_member_oid(recheckIndexes, uniqueidx) &&
            FindConflictTuple(resultRelInfo, estate, uniqueidx, remoteslot,
                              &conflictslot))
        {
            ConflictTupleInfo *conflicttuple = palloc0_object(ConflictTupleInfo);
 
            conflicttuple->slot = conflictslot;
            conflicttuple->indexoid = uniqueidx;
 
            GetTupleTransactionInfo(conflictslot, &conflicttuple->xmin,
                                    &conflicttuple->origin, &conflicttuple->ts);
 
            conflicttuples = lappend(conflicttuples, conflicttuple);
        }
    }
 
    /* Report the conflict, if found */
    if (conflicttuples)
        ReportApplyConflict(estate, resultRelInfo, ERROR,
                            list_length(conflicttuples) > 1 ? CT_MULTIPLE_UNIQUE_CONFLICTS : type,
                            searchslot, remoteslot, conflicttuples);
}

References CT_MULTIPLE_UNIQUE_CONFLICTS, ERROR, fb(), FindConflictTuple(), foreach_oid, GetTupleTransactionInfo(), lappend(), list_length(), list_member_oid(), NIL, palloc0_object, ReportApplyConflict(), ResultRelInfo::ri_onConflictArbiterIndexes, and type.

Referenced by ExecSimpleRelationInsert(), and ExecSimpleRelationUpdate().

CheckCmdReplicaIdentity()

void CheckCmdReplicaIdentity	(	Relation	rel,
		CmdType	cmd
	)

Definition at line 1029 of file execReplication.c.

{
    PublicationDesc pubdesc;
 
    /*
     * Skip checking the replica identity for partitioned tables, because the
     * operations are actually performed on the leaf partitions.
     */
    if (rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE)
        return;
 
    /* We only need to do checks for UPDATE and DELETE. */
    if (cmd != CMD_UPDATE && cmd != CMD_DELETE)
        return;
 
    /*
     * It is only safe to execute UPDATE/DELETE if the relation does not
     * publish UPDATEs or DELETEs, or all the following conditions are
     * satisfied:
     *
     * 1. All columns, referenced in the row filters from publications which
     * the relation is in, are valid - i.e. when all referenced columns are
     * part of REPLICA IDENTITY.
     *
     * 2. All columns, referenced in the column lists are valid - i.e. when
     * all columns referenced in the REPLICA IDENTITY are covered by the
     * column list.
     *
     * 3. All generated columns in REPLICA IDENTITY of the relation, are valid
     * - i.e. when all these generated columns are published.
     *
     * XXX We could optimize it by first checking whether any of the
     * publications have a row filter or column list for this relation, or if
     * the relation contains a generated column. If none of these exist and
     * the relation has replica identity then we can avoid building the
     * descriptor but as this happens only one time it doesn't seem worth the
     * additional complexity.
     */
    RelationBuildPublicationDesc(rel, &pubdesc);
    if (cmd == CMD_UPDATE && !pubdesc.rf_valid_for_update)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot update table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Column used in the publication WHERE expression is not part of the replica identity.")));
    else if (cmd == CMD_UPDATE && !pubdesc.cols_valid_for_update)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot update table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Column list used by the publication does not cover the replica identity.")));
    else if (cmd == CMD_UPDATE && !pubdesc.gencols_valid_for_update)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot update table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Replica identity must not contain unpublished generated columns.")));
    else if (cmd == CMD_DELETE && !pubdesc.rf_valid_for_delete)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot delete from table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Column used in the publication WHERE expression is not part of the replica identity.")));
    else if (cmd == CMD_DELETE && !pubdesc.cols_valid_for_delete)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot delete from table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Column list used by the publication does not cover the replica identity.")));
    else if (cmd == CMD_DELETE && !pubdesc.gencols_valid_for_delete)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_COLUMN_REFERENCE),
                 errmsg("cannot delete from table \"%s\"",
                        RelationGetRelationName(rel)),
                 errdetail("Replica identity must not contain unpublished generated columns.")));
 
    /* If relation has replica identity we are always good. */
    if (OidIsValid(RelationGetReplicaIndex(rel)))
        return;
 
    /* REPLICA IDENTITY FULL is also good for UPDATE/DELETE. */
    if (rel->rd_rel->relreplident == REPLICA_IDENTITY_FULL)
        return;
 
    /*
     * This is UPDATE/DELETE and there is no replica identity.
     *
     * Check if the table publishes UPDATES or DELETES.
     */
    if (cmd == CMD_UPDATE && pubdesc.pubactions.pubupdate)
        ereport(ERROR,
                (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
                 errmsg("cannot update table \"%s\" because it does not have a replica identity and publishes updates",
                        RelationGetRelationName(rel)),
                 errhint("To enable updating the table, set REPLICA IDENTITY using ALTER TABLE.")));
    else if (cmd == CMD_DELETE && pubdesc.pubactions.pubdelete)
        ereport(ERROR,
                (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
                 errmsg("cannot delete from table \"%s\" because it does not have a replica identity and publishes deletes",
                        RelationGetRelationName(rel)),
                 errhint("To enable deleting from the table, set REPLICA IDENTITY using ALTER TABLE.")));
}

References CMD_DELETE, CMD_UPDATE, ereport, errcode(), errdetail(), errhint(), errmsg, ERROR, fb(), OidIsValid, RelationData::rd_rel, RelationBuildPublicationDesc(), RelationGetRelationName, and RelationGetReplicaIndex().

Referenced by CheckValidResultRel(), ExecSimpleRelationDelete(), ExecSimpleRelationInsert(), and ExecSimpleRelationUpdate().

CheckSubscriptionRelkind()

void CheckSubscriptionRelkind	(	char	localrelkind,
		char	remoterelkind,
		const char *	nspname,
		const char *	relname
	)

Definition at line 1140 of file execReplication.c.

{
    if (localrelkind != RELKIND_RELATION &&
        localrelkind != RELKIND_PARTITIONED_TABLE &&
        localrelkind != RELKIND_SEQUENCE)
        ereport(ERROR,
                (errcode(ERRCODE_WRONG_OBJECT_TYPE),
                 errmsg("cannot use relation \"%s.%s\" as logical replication target",
                        nspname, relname),
                 errdetail_relkind_not_supported(localrelkind)));
 
    /*
     * Allow RELKIND_RELATION and RELKIND_PARTITIONED_TABLE to be treated
     * interchangeably, but ensure that sequences (RELKIND_SEQUENCE) match
     * exactly on both publisher and subscriber.
     */
    if ((localrelkind == RELKIND_SEQUENCE && remoterelkind != RELKIND_SEQUENCE) ||
        (localrelkind != RELKIND_SEQUENCE && remoterelkind == RELKIND_SEQUENCE))
        ereport(ERROR,
                errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
        /* translator: 3rd and 4th %s are "sequence" or "table" */
                errmsg("relation \"%s.%s\" type mismatch: source \"%s\", target \"%s\"",
                       nspname, relname,
                       remoterelkind == RELKIND_SEQUENCE ? "sequence" : "table",
                       localrelkind == RELKIND_SEQUENCE ? "sequence" : "table"));
}

References ereport, errcode(), errdetail_relkind_not_supported(), errmsg, ERROR, fb(), and relname.

Referenced by AlterSubscription_refresh(), apply_handle_tuple_routing(), CreateSubscription(), and logicalrep_rel_open().

ExecSimpleRelationDelete()

void ExecSimpleRelationDelete	(	ResultRelInfo *	resultRelInfo,
		EState *	estate,
		EPQState *	epqstate,
		TupleTableSlot *	searchslot
	)

Definition at line 996 of file execReplication.c.

{
    bool        skip_tuple = false;
    Relation    rel = resultRelInfo->ri_RelationDesc;
    ItemPointer tid = &searchslot->tts_tid;
 
    CheckCmdReplicaIdentity(rel, CMD_DELETE);
 
    /* BEFORE ROW DELETE Triggers */
    if (resultRelInfo->ri_TrigDesc &&
        resultRelInfo->ri_TrigDesc->trig_delete_before_row)
    {
        skip_tuple = !ExecBRDeleteTriggers(estate, epqstate, resultRelInfo,
                                           tid, NULL, NULL, NULL, NULL, false);
    }
 
    if (!skip_tuple)
    {
        /* OK, delete the tuple */
        simple_table_tuple_delete(rel, tid, estate->es_snapshot);
 
        /* AFTER ROW DELETE Triggers */
        ExecARDeleteTriggers(estate, resultRelInfo,
                             tid, NULL, NULL, false);
    }
}

References CheckCmdReplicaIdentity(), CMD_DELETE, EState::es_snapshot, ExecARDeleteTriggers(), ExecBRDeleteTriggers(), fb(), ResultRelInfo::ri_RelationDesc, ResultRelInfo::ri_TrigDesc, simple_table_tuple_delete(), and TriggerDesc::trig_delete_before_row.

Referenced by apply_handle_delete_internal(), and apply_handle_tuple_routing().

ExecSimpleRelationInsert()

void ExecSimpleRelationInsert	(	ResultRelInfo *	resultRelInfo,
		EState *	estate,
		TupleTableSlot *	slot
	)

Definition at line 810 of file execReplication.c.

{
    bool        skip_tuple = false;
    Relation    rel = resultRelInfo->ri_RelationDesc;
 
    /* For now we support only tables. */
    Assert(rel->rd_rel->relkind == RELKIND_RELATION);
 
    CheckCmdReplicaIdentity(rel, CMD_INSERT);
 
    /* BEFORE ROW INSERT Triggers */
    if (resultRelInfo->ri_TrigDesc &&
        resultRelInfo->ri_TrigDesc->trig_insert_before_row)
    {
        if (!ExecBRInsertTriggers(estate, resultRelInfo, slot))
            skip_tuple = true;  /* "do nothing" */
    }
 
    if (!skip_tuple)
    {
        List       *recheckIndexes = NIL;
        List       *conflictindexes;
        bool        conflict = false;
 
        /* Compute stored generated columns */
        if (rel->rd_att->constr &&
            rel->rd_att->constr->has_generated_stored)
            ExecComputeStoredGenerated(resultRelInfo, estate, slot,
                                       CMD_INSERT);
 
        /* Check the constraints of the tuple */
        if (rel->rd_att->constr)
            ExecConstraints(resultRelInfo, slot, estate);
        if (rel->rd_rel->relispartition)
            ExecPartitionCheck(resultRelInfo, slot, estate, true);
 
        /* OK, store the tuple and create index entries for it */
        simple_table_tuple_insert(resultRelInfo->ri_RelationDesc, slot);
 
        conflictindexes = resultRelInfo->ri_onConflictArbiterIndexes;
 
        if (resultRelInfo->ri_NumIndices > 0)
        {
            uint32      flags;
 
            if (conflictindexes != NIL)
                flags = EIIT_NO_DUPE_ERROR;
            else
                flags = 0;
            recheckIndexes = ExecInsertIndexTuples(resultRelInfo,
                                                   estate, flags,
                                                   slot, conflictindexes,
                                                   &conflict);
        }
 
        /*
         * Checks the conflict indexes to fetch the conflicting local row and
         * reports the conflict. We perform this check here, instead of
         * performing an additional index scan before the actual insertion and
         * reporting the conflict if any conflicting rows are found. This is
         * to avoid the overhead of executing the extra scan for each INSERT
         * operation, even when no conflict arises, which could introduce
         * significant overhead to replication, particularly in cases where
         * conflicts are rare.
         *
         * XXX OTOH, this could lead to clean-up effort for dead tuples added
         * in heap and index in case of conflicts. But as conflicts shouldn't
         * be a frequent thing so we preferred to save the performance
         * overhead of extra scan before each insertion.
         */
        if (conflict)
            CheckAndReportConflict(resultRelInfo, estate, CT_INSERT_EXISTS,
                                   recheckIndexes, NULL, slot);
 
        /* AFTER ROW INSERT Triggers */
        ExecARInsertTriggers(estate, resultRelInfo, slot,
                             recheckIndexes, NULL);
 
        /*
         * XXX we should in theory pass a TransitionCaptureState object to the
         * above to capture transition tuples, but after statement triggers
         * don't actually get fired by replication yet anyway
         */
 
        list_free(recheckIndexes);
    }
}

References Assert, CheckAndReportConflict(), CheckCmdReplicaIdentity(), CMD_INSERT, TupleDescData::constr, CT_INSERT_EXISTS, EIIT_NO_DUPE_ERROR, ExecARInsertTriggers(), ExecBRInsertTriggers(), ExecComputeStoredGenerated(), ExecConstraints(), ExecInsertIndexTuples(), ExecPartitionCheck(), fb(), TupleConstr::has_generated_stored, list_free(), NIL, RelationData::rd_att, RelationData::rd_rel, ResultRelInfo::ri_NumIndices, ResultRelInfo::ri_onConflictArbiterIndexes, ResultRelInfo::ri_RelationDesc, ResultRelInfo::ri_TrigDesc, simple_table_tuple_insert(), and TriggerDesc::trig_insert_before_row.

Referenced by apply_handle_insert_internal().

ExecSimpleRelationUpdate()

void ExecSimpleRelationUpdate	(	ResultRelInfo *	resultRelInfo,
		EState *	estate,
		EPQState *	epqstate,
		TupleTableSlot *	searchslot,
		TupleTableSlot *	slot
	)

Definition at line 906 of file execReplication.c.

{
    bool        skip_tuple = false;
    Relation    rel = resultRelInfo->ri_RelationDesc;
    ItemPointer tid = &(searchslot->tts_tid);
 
    /*
     * We support only non-system tables, with
     * check_publication_add_relation() accountable.
     */
    Assert(rel->rd_rel->relkind == RELKIND_RELATION);
    Assert(!IsCatalogRelation(rel));
 
    CheckCmdReplicaIdentity(rel, CMD_UPDATE);
 
    /* BEFORE ROW UPDATE Triggers */
    if (resultRelInfo->ri_TrigDesc &&
        resultRelInfo->ri_TrigDesc->trig_update_before_row)
    {
        if (!ExecBRUpdateTriggers(estate, epqstate, resultRelInfo,
                                  tid, NULL, slot, NULL, NULL, false))
            skip_tuple = true;  /* "do nothing" */
    }
 
    if (!skip_tuple)
    {
        List       *recheckIndexes = NIL;
        TU_UpdateIndexes update_indexes;
        List       *conflictindexes;
        bool        conflict = false;
 
        /* Compute stored generated columns */
        if (rel->rd_att->constr &&
            rel->rd_att->constr->has_generated_stored)
            ExecComputeStoredGenerated(resultRelInfo, estate, slot,
                                       CMD_UPDATE);
 
        /* Check the constraints of the tuple */
        if (rel->rd_att->constr)
            ExecConstraints(resultRelInfo, slot, estate);
        if (rel->rd_rel->relispartition)
            ExecPartitionCheck(resultRelInfo, slot, estate, true);
 
        simple_table_tuple_update(rel, tid, slot, estate->es_snapshot,
                                  &update_indexes);
 
        conflictindexes = resultRelInfo->ri_onConflictArbiterIndexes;
 
        if (resultRelInfo->ri_NumIndices > 0 && (update_indexes != TU_None))
        {
            uint32      flags = EIIT_IS_UPDATE;
 
            if (conflictindexes != NIL)
                flags |= EIIT_NO_DUPE_ERROR;
            if (update_indexes == TU_Summarizing)
                flags |= EIIT_ONLY_SUMMARIZING;
            recheckIndexes = ExecInsertIndexTuples(resultRelInfo,
                                                   estate, flags,
                                                   slot, conflictindexes,
                                                   &conflict);
        }
 
        /*
         * Refer to the comments above the call to CheckAndReportConflict() in
         * ExecSimpleRelationInsert to understand why this check is done at
         * this point.
         */
        if (conflict)
            CheckAndReportConflict(resultRelInfo, estate, CT_UPDATE_EXISTS,
                                   recheckIndexes, searchslot, slot);
 
        /* AFTER ROW UPDATE Triggers */
        ExecARUpdateTriggers(estate, resultRelInfo,
                             NULL, NULL,
                             tid, NULL, slot,
                             recheckIndexes, NULL, false);
 
        list_free(recheckIndexes);
    }
}

References Assert, CheckAndReportConflict(), CheckCmdReplicaIdentity(), CMD_UPDATE, TupleDescData::constr, CT_UPDATE_EXISTS, EIIT_IS_UPDATE, EIIT_NO_DUPE_ERROR, EIIT_ONLY_SUMMARIZING, EState::es_snapshot, ExecARUpdateTriggers(), ExecBRUpdateTriggers(), ExecComputeStoredGenerated(), ExecConstraints(), ExecInsertIndexTuples(), ExecPartitionCheck(), fb(), TupleConstr::has_generated_stored, IsCatalogRelation(), list_free(), NIL, RelationData::rd_att, RelationData::rd_rel, ResultRelInfo::ri_NumIndices, ResultRelInfo::ri_onConflictArbiterIndexes, ResultRelInfo::ri_RelationDesc, ResultRelInfo::ri_TrigDesc, simple_table_tuple_update(), TriggerDesc::trig_update_before_row, TU_None, and TU_Summarizing.

Referenced by apply_handle_tuple_routing(), and apply_handle_update_internal().

FindConflictTuple()

static bool FindConflictTuple ( ResultRelInfo *  resultRelInfo, EState *  estate, Oid  conflictindex, TupleTableSlot *  slot, TupleTableSlot **  conflictslot  )

static

Definition at line 716 of file execReplication.c.

{
    Relation    rel = resultRelInfo->ri_RelationDesc;
    ItemPointerData conflictTid;
    TM_FailureData tmfd;
    TM_Result   res;
 
    *conflictslot = NULL;
 
    /*
     * Build additional information required to check constraints violations.
     * See check_exclusion_or_unique_constraint().
     */
    BuildConflictIndexInfo(resultRelInfo, conflictindex);
 
retry:
    if (ExecCheckIndexConstraints(resultRelInfo, slot, estate,
                                  &conflictTid, &slot->tts_tid,
                                  list_make1_oid(conflictindex)))
    {
        if (*conflictslot)
            ExecDropSingleTupleTableSlot(*conflictslot);
 
        *conflictslot = NULL;
        return false;
    }
 
    *conflictslot = table_slot_create(rel, NULL);
 
    PushActiveSnapshot(GetLatestSnapshot());
 
    res = table_tuple_lock(rel, &conflictTid, GetActiveSnapshot(),
                           *conflictslot,
                           GetCurrentCommandId(false),
                           LockTupleShare,
                           LockWaitBlock,
                           0 /* don't follow updates */ ,
                           &tmfd);
 
    PopActiveSnapshot();
 
    if (should_refetch_tuple(res, &tmfd))
        goto retry;
 
    return true;
}

References BuildConflictIndexInfo(), ExecCheckIndexConstraints(), ExecDropSingleTupleTableSlot(), fb(), GetActiveSnapshot(), GetCurrentCommandId(), GetLatestSnapshot(), list_make1_oid, LockTupleShare, LockWaitBlock, PopActiveSnapshot(), PushActiveSnapshot(), ResultRelInfo::ri_RelationDesc, should_refetch_tuple(), table_slot_create(), table_tuple_lock(), and TupleTableSlot::tts_tid.

Referenced by CheckAndReportConflict().

RelationFindDeletedTupleInfoByIndex()

bool RelationFindDeletedTupleInfoByIndex	(	Relation	rel,
		Oid	idxoid,
		TupleTableSlot *	searchslot,
		TransactionId	oldestxmin,
		TransactionId *	delete_xid,
		ReplOriginId *	delete_origin,
		TimestampTz *	delete_time
	)

Definition at line 634 of file execReplication.c.

{
    Relation    idxrel;
    ScanKeyData skey[INDEX_MAX_KEYS];
    int         skey_attoff;
    IndexScanDesc scan;
    TupleTableSlot *scanslot;
    TypeCacheEntry **eq = NULL;
    bool        isIdxSafeToSkipDuplicates;
    TupleDesc   desc PG_USED_FOR_ASSERTS_ONLY = RelationGetDescr(rel);
 
    Assert(equalTupleDescs(desc, searchslot->tts_tupleDescriptor));
    Assert(OidIsValid(idxoid));
 
    *delete_xid = InvalidTransactionId;
    *delete_time = 0;
    *delete_origin = InvalidReplOriginId;
 
    isIdxSafeToSkipDuplicates = (GetRelationIdentityOrPK(rel) == idxoid);
 
    scanslot = table_slot_create(rel, NULL);
 
    idxrel = index_open(idxoid, RowExclusiveLock);
 
    /* Build scan key. */
    skey_attoff = build_replindex_scan_key(skey, rel, idxrel, searchslot);
 
    /*
     * Start an index scan using SnapshotAny to identify dead tuples that are
     * not visible under a standard MVCC snapshot. Tuples from transactions
     * not yet committed or those just committed prior to the scan are
     * excluded in update_most_recent_deletion_info().
     */
    scan = index_beginscan(rel, idxrel,
                           SnapshotAny, NULL, skey_attoff, 0, SO_NONE);
 
    index_rescan(scan, skey, skey_attoff, NULL, 0);
 
    /* Try to find the tuple */
    while (index_getnext_slot(scan, ForwardScanDirection, scanslot))
    {
        /*
         * Avoid expensive equality check if the index is primary key or
         * replica identity index.
         */
        if (!isIdxSafeToSkipDuplicates)
        {
            if (eq == NULL)
                eq = palloc0_array(TypeCacheEntry *, scanslot->tts_tupleDescriptor->natts);
 
            if (!tuples_equal(scanslot, searchslot, eq, NULL))
                continue;
        }
 
        update_most_recent_deletion_info(scanslot, oldestxmin, delete_xid,
                                         delete_time, delete_origin);
    }
 
    index_endscan(scan);
 
    index_close(idxrel, NoLock);
 
    ExecDropSingleTupleTableSlot(scanslot);
 
    return *delete_time != 0;
}

References Assert, build_replindex_scan_key(), equalTupleDescs(), ExecDropSingleTupleTableSlot(), fb(), ForwardScanDirection, GetRelationIdentityOrPK(), index_beginscan(), index_close(), index_endscan(), index_getnext_slot(), INDEX_MAX_KEYS, index_open(), index_rescan(), InvalidReplOriginId, InvalidTransactionId, NoLock, OidIsValid, palloc0_array, PG_USED_FOR_ASSERTS_ONLY, RelationGetDescr, RowExclusiveLock, SnapshotAny, SO_NONE, table_slot_create(), tuples_equal(), and update_most_recent_deletion_info().

Referenced by FindDeletedTupleInLocalRel().

RelationFindDeletedTupleInfoSeq()

bool RelationFindDeletedTupleInfoSeq	(	Relation	rel,
		TupleTableSlot *	searchslot,
		TransactionId	oldestxmin,
		TransactionId *	delete_xid,
		ReplOriginId *	delete_origin,
		TimestampTz *	delete_time
	)

Definition at line 564 of file execReplication.c.

{
    TupleTableSlot *scanslot;
    TableScanDesc scan;
    TypeCacheEntry **eq;
    Bitmapset  *indexbitmap;
    TupleDesc   desc PG_USED_FOR_ASSERTS_ONLY = RelationGetDescr(rel);
 
    Assert(equalTupleDescs(desc, searchslot->tts_tupleDescriptor));
 
    *delete_xid = InvalidTransactionId;
    *delete_origin = InvalidReplOriginId;
    *delete_time = 0;
 
    /*
     * If the relation has a replica identity key or a primary key that is
     * unusable for locating deleted tuples (see
     * IsIndexUsableForFindingDeletedTuple), a full table scan becomes
     * necessary. In such cases, comparing the entire tuple is not required,
     * since the remote tuple might not include all column values. Instead,
     * the indexed columns alone are sufficient to identify the target tuple
     * (see logicalrep_rel_mark_updatable).
     */
    indexbitmap = RelationGetIndexAttrBitmap(rel,
                                             INDEX_ATTR_BITMAP_IDENTITY_KEY);
 
    /* fallback to PK if no replica identity */
    if (!indexbitmap)
        indexbitmap = RelationGetIndexAttrBitmap(rel,
                                                 INDEX_ATTR_BITMAP_PRIMARY_KEY);
 
    eq = palloc0_array(TypeCacheEntry *, searchslot->tts_tupleDescriptor->natts);
 
    /*
     * Start a heap scan using SnapshotAny to identify dead tuples that are
     * not visible under a standard MVCC snapshot. Tuples from transactions
     * not yet committed or those just committed prior to the scan are
     * excluded in update_most_recent_deletion_info().
     */
    scan = table_beginscan(rel, SnapshotAny, 0, NULL,
                           SO_NONE);
    scanslot = table_slot_create(rel, NULL);
 
    table_rescan(scan, NULL);
 
    /* Try to find the tuple */
    while (table_scan_getnextslot(scan, ForwardScanDirection, scanslot))
    {
        if (!tuples_equal(scanslot, searchslot, eq, indexbitmap))
            continue;
 
        update_most_recent_deletion_info(scanslot, oldestxmin, delete_xid,
                                         delete_time, delete_origin);
    }
 
    table_endscan(scan);
    ExecDropSingleTupleTableSlot(scanslot);
 
    return *delete_time != 0;
}

References Assert, equalTupleDescs(), ExecDropSingleTupleTableSlot(), fb(), ForwardScanDirection, INDEX_ATTR_BITMAP_IDENTITY_KEY, INDEX_ATTR_BITMAP_PRIMARY_KEY, InvalidReplOriginId, InvalidTransactionId, palloc0_array, PG_USED_FOR_ASSERTS_ONLY, RelationGetDescr, RelationGetIndexAttrBitmap(), SnapshotAny, SO_NONE, table_beginscan(), table_endscan(), table_rescan(), table_scan_getnextslot(), table_slot_create(), tuples_equal(), and update_most_recent_deletion_info().

Referenced by FindDeletedTupleInLocalRel().

RelationFindReplTupleByIndex()

bool RelationFindReplTupleByIndex	(	Relation	rel,
		Oid	idxoid,
		LockTupleMode	lockmode,
		TupleTableSlot *	searchslot,
		TupleTableSlot *	outslot
	)

Definition at line 182 of file execReplication.c.

{
    ScanKeyData skey[INDEX_MAX_KEYS];
    int         skey_attoff;
    IndexScanDesc scan;
    SnapshotData snap;
    TransactionId xwait;
    Relation    idxrel;
    bool        found;
    TypeCacheEntry **eq = NULL;
    bool        isIdxSafeToSkipDuplicates;
 
    /* Open the index. */
    idxrel = index_open(idxoid, RowExclusiveLock);
 
    isIdxSafeToSkipDuplicates = (GetRelationIdentityOrPK(rel) == idxoid);
 
    InitDirtySnapshot(snap);
 
    /* Build scan key. */
    skey_attoff = build_replindex_scan_key(skey, rel, idxrel, searchslot);
 
    /* Start an index scan. */
    scan = index_beginscan(rel, idxrel,
                           &snap, NULL, skey_attoff, 0, SO_NONE);
 
retry:
    found = false;
 
    index_rescan(scan, skey, skey_attoff, NULL, 0);
 
    /* Try to find the tuple */
    while (index_getnext_slot(scan, ForwardScanDirection, outslot))
    {
        /*
         * Avoid expensive equality check if the index is primary key or
         * replica identity index.
         */
        if (!isIdxSafeToSkipDuplicates)
        {
            if (eq == NULL)
                eq = palloc0_array(TypeCacheEntry *, outslot->tts_tupleDescriptor->natts);
 
            if (!tuples_equal(outslot, searchslot, eq, NULL))
                continue;
        }
 
        ExecMaterializeSlot(outslot);
 
        xwait = TransactionIdIsValid(snap.xmin) ?
            snap.xmin : snap.xmax;
 
        /*
         * If the tuple is locked, wait for locking transaction to finish and
         * retry.
         */
        if (TransactionIdIsValid(xwait))
        {
            XactLockTableWait(xwait, NULL, NULL, XLTW_None);
            goto retry;
        }
 
        /* Found our tuple and it's not locked */
        found = true;
        break;
    }
 
    /* Found tuple, try to lock it in the lockmode. */
    if (found)
    {
        TM_FailureData tmfd;
        TM_Result   res;
 
        PushActiveSnapshot(GetLatestSnapshot());
 
        res = table_tuple_lock(rel, &(outslot->tts_tid), GetActiveSnapshot(),
                               outslot,
                               GetCurrentCommandId(false),
                               lockmode,
                               LockWaitBlock,
                               0 /* don't follow updates */ ,
                               &tmfd);
 
        PopActiveSnapshot();
 
        if (should_refetch_tuple(res, &tmfd))
            goto retry;
    }
 
    index_endscan(scan);
 
    /* Don't release lock until commit. */
    index_close(idxrel, NoLock);
 
    return found;
}

References build_replindex_scan_key(), ExecMaterializeSlot(), fb(), ForwardScanDirection, GetActiveSnapshot(), GetCurrentCommandId(), GetLatestSnapshot(), GetRelationIdentityOrPK(), index_beginscan(), index_close(), index_endscan(), index_getnext_slot(), INDEX_MAX_KEYS, index_open(), index_rescan(), InitDirtySnapshot, LockWaitBlock, NoLock, palloc0_array, PopActiveSnapshot(), PushActiveSnapshot(), RowExclusiveLock, should_refetch_tuple(), SO_NONE, table_tuple_lock(), TransactionIdIsValid, tuples_equal(), XactLockTableWait(), and XLTW_None.

Referenced by FindReplTupleInLocalRel().

RelationFindReplTupleSeq()

bool RelationFindReplTupleSeq	(	Relation	rel,
		LockTupleMode	lockmode,
		TupleTableSlot *	searchslot,
		TupleTableSlot *	outslot
	)

Definition at line 370 of file execReplication.c.

{
    TupleTableSlot *scanslot;
    TableScanDesc scan;
    SnapshotData snap;
    TypeCacheEntry **eq;
    TransactionId xwait;
    bool        found;
    TupleDesc   desc PG_USED_FOR_ASSERTS_ONLY = RelationGetDescr(rel);
 
    Assert(equalTupleDescs(desc, outslot->tts_tupleDescriptor));
 
    eq = palloc0_array(TypeCacheEntry *, outslot->tts_tupleDescriptor->natts);
 
    /* Start a heap scan. */
    InitDirtySnapshot(snap);
    scan = table_beginscan(rel, &snap, 0, NULL,
                           SO_NONE);
    scanslot = table_slot_create(rel, NULL);
 
retry:
    found = false;
 
    table_rescan(scan, NULL);
 
    /* Try to find the tuple */
    while (table_scan_getnextslot(scan, ForwardScanDirection, scanslot))
    {
        if (!tuples_equal(scanslot, searchslot, eq, NULL))
            continue;
 
        found = true;
        ExecCopySlot(outslot, scanslot);
 
        xwait = TransactionIdIsValid(snap.xmin) ?
            snap.xmin : snap.xmax;
 
        /*
         * If the tuple is locked, wait for locking transaction to finish and
         * retry.
         */
        if (TransactionIdIsValid(xwait))
        {
            XactLockTableWait(xwait, NULL, NULL, XLTW_None);
            goto retry;
        }
 
        /* Found our tuple and it's not locked */
        break;
    }
 
    /* Found tuple, try to lock it in the lockmode. */
    if (found)
    {
        TM_FailureData tmfd;
        TM_Result   res;
 
        PushActiveSnapshot(GetLatestSnapshot());
 
        res = table_tuple_lock(rel, &(outslot->tts_tid), GetActiveSnapshot(),
                               outslot,
                               GetCurrentCommandId(false),
                               lockmode,
                               LockWaitBlock,
                               0 /* don't follow updates */ ,
                               &tmfd);
 
        PopActiveSnapshot();
 
        if (should_refetch_tuple(res, &tmfd))
            goto retry;
    }
 
    table_endscan(scan);
    ExecDropSingleTupleTableSlot(scanslot);
 
    return found;
}

References Assert, equalTupleDescs(), ExecCopySlot(), ExecDropSingleTupleTableSlot(), fb(), ForwardScanDirection, GetActiveSnapshot(), GetCurrentCommandId(), GetLatestSnapshot(), InitDirtySnapshot, LockWaitBlock, palloc0_array, PG_USED_FOR_ASSERTS_ONLY, PopActiveSnapshot(), PushActiveSnapshot(), RelationGetDescr, should_refetch_tuple(), SO_NONE, table_beginscan(), table_endscan(), table_rescan(), table_scan_getnextslot(), table_slot_create(), table_tuple_lock(), TransactionIdIsValid, tuples_equal(), XactLockTableWait(), and XLTW_None.

Referenced by FindReplTupleInLocalRel().

should_refetch_tuple()

static bool should_refetch_tuple ( TM_Result  res, TM_FailureData *  tmfd  )

static

Definition at line 137 of file execReplication.c.

{
    bool        refetch = false;
 
    switch (res)
    {
        case TM_Ok:
            break;
        case TM_Updated:
            /* XXX: Improve handling here */
            if (ItemPointerIndicatesMovedPartitions(&tmfd->ctid))
                ereport(LOG,
                        (errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
                         errmsg("tuple to be locked was already moved to another partition due to concurrent update, retrying")));
            else
                ereport(LOG,
                        (errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
                         errmsg("concurrent update, retrying")));
            refetch = true;
            break;
        case TM_Deleted:
            /* XXX: Improve handling here */
            ereport(LOG,
                    (errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
                     errmsg("concurrent delete, retrying")));
            refetch = true;
            break;
        case TM_Invisible:
            elog(ERROR, "attempted to lock invisible tuple");
            break;
        default:
            elog(ERROR, "unexpected table_tuple_lock status: %u", res);
            break;
    }
 
    return refetch;
}

References TM_FailureData::ctid, elog, ereport, errcode(), ERRCODE_T_R_SERIALIZATION_FAILURE, errmsg, ERROR, fb(), ItemPointerIndicatesMovedPartitions(), LOG, TM_Deleted, TM_Invisible, TM_Ok, and TM_Updated.

Referenced by FindConflictTuple(), RelationFindReplTupleByIndex(), and RelationFindReplTupleSeq().

tuples_equal()

static bool tuples_equal ( TupleTableSlot *  slot1, TupleTableSlot *  slot2, TypeCacheEntry **  eq, Bitmapset *  columns  )

static

Definition at line 289 of file execReplication.c.

{
    int         attrnum;
 
    Assert(slot1->tts_tupleDescriptor->natts ==
           slot2->tts_tupleDescriptor->natts);
 
    slot_getallattrs(slot1);
    slot_getallattrs(slot2);
 
    /* Check equality of the attributes. */
    for (attrnum = 0; attrnum < slot1->tts_tupleDescriptor->natts; attrnum++)
    {
        Form_pg_attribute att;
        TypeCacheEntry *typentry;
 
        att = TupleDescAttr(slot1->tts_tupleDescriptor, attrnum);
 
        /*
         * Ignore dropped and generated columns as the publisher doesn't send
         * those
         */
        if (att->attisdropped || att->attgenerated)
            continue;
 
        /*
         * Ignore columns that are not listed for checking.
         */
        if (columns &&
            !bms_is_member(att->attnum - FirstLowInvalidHeapAttributeNumber,
                           columns))
            continue;
 
        /*
         * If one value is NULL and other is not, then they are certainly not
         * equal
         */
        if (slot1->tts_isnull[attrnum] != slot2->tts_isnull[attrnum])
            return false;
 
        /*
         * If both are NULL, they can be considered equal.
         */
        if (slot1->tts_isnull[attrnum] || slot2->tts_isnull[attrnum])
            continue;
 
        typentry = eq[attrnum];
        if (typentry == NULL)
        {
            typentry = lookup_type_cache(att->atttypid,
                                         TYPECACHE_EQ_OPR_FINFO);
            if (!OidIsValid(typentry->eq_opr_finfo.fn_oid))
                ereport(ERROR,
                        (errcode(ERRCODE_UNDEFINED_FUNCTION),
                         errmsg("could not identify an equality operator for type %s",
                                format_type_be(att->atttypid))));
            eq[attrnum] = typentry;
        }
 
        if (!DatumGetBool(FunctionCall2Coll(&typentry->eq_opr_finfo,
                                            att->attcollation,
                                            slot1->tts_values[attrnum],
                                            slot2->tts_values[attrnum])))
            return false;
    }
 
    return true;
}

References Assert, bms_is_member(), DatumGetBool(), TypeCacheEntry::eq_opr_finfo, ereport, errcode(), errmsg, ERROR, fb(), FirstLowInvalidHeapAttributeNumber, FmgrInfo::fn_oid, format_type_be(), FunctionCall2Coll(), lookup_type_cache(), OidIsValid, slot_getallattrs(), TupleDescAttr(), and TYPECACHE_EQ_OPR_FINFO.

Referenced by RelationFindDeletedTupleInfoByIndex(), RelationFindDeletedTupleInfoSeq(), RelationFindReplTupleByIndex(), and RelationFindReplTupleSeq().

update_most_recent_deletion_info()

static void update_most_recent_deletion_info ( TupleTableSlot *  scanslot, TransactionId  oldestxmin, TransactionId *  delete_xid, TimestampTz *  delete_time, ReplOriginId *  delete_origin  )

static

Definition at line 481 of file execReplication.c.

{
    BufferHeapTupleTableSlot *hslot;
    HeapTuple   tuple;
    Buffer      buf;
    bool        recently_dead = false;
    TransactionId xmax;
    TimestampTz localts;
    ReplOriginId localorigin;
 
    hslot = (BufferHeapTupleTableSlot *) scanslot;
 
    tuple = ExecFetchSlotHeapTuple(scanslot, false, NULL);
    buf = hslot->buffer;
 
    LockBuffer(buf, BUFFER_LOCK_SHARE);
 
    /*
     * We do not consider HEAPTUPLE_DEAD status because it indicates either
     * tuples whose inserting transaction was aborted (meaning there is no
     * commit timestamp or origin), or tuples deleted by a transaction older
     * than oldestxmin, making it safe to ignore them during conflict
     * detection (See comments atop worker.c for details).
     */
    if (HeapTupleSatisfiesVacuum(tuple, oldestxmin, buf) == HEAPTUPLE_RECENTLY_DEAD)
        recently_dead = true;
 
    LockBuffer(buf, BUFFER_LOCK_UNLOCK);
 
    if (!recently_dead)
        return;
 
    xmax = HeapTupleHeaderGetUpdateXid(tuple->t_data);
    if (!TransactionIdIsValid(xmax))
        return;
 
    /* Select the dead tuple with the most recent commit timestamp */
    if (TransactionIdGetCommitTsData(xmax, &localts, &localorigin) &&
        TimestampDifferenceExceeds(*delete_time, localts, 0))
    {
        *delete_xid = xmax;
        *delete_time = localts;
        *delete_origin = localorigin;
    }
}

References buf, BUFFER_LOCK_SHARE, BUFFER_LOCK_UNLOCK, ExecFetchSlotHeapTuple(), fb(), HEAPTUPLE_RECENTLY_DEAD, HeapTupleHeaderGetUpdateXid(), HeapTupleSatisfiesVacuum(), LockBuffer(), HeapTupleData::t_data, TimestampDifferenceExceeds(), TransactionIdGetCommitTsData(), and TransactionIdIsValid.

Referenced by RelationFindDeletedTupleInfoByIndex(), and RelationFindDeletedTupleInfoSeq().