diff options
| author | Alvaro Herrera | 2018-01-19 14:49:22 +0000 |
|---|---|---|
| committer | Alvaro Herrera | 2018-01-19 14:49:22 +0000 |
| commit | 8b08f7d4820fd7a8ef6152a9dd8c6e3cb01e5f99 (patch) | |
| tree | 0ecb879a9568e7e14275afe72a77e316e72376c6 /src/backend | |
| parent | 1ef61ddce9086c30a18a6ecc48bc3ce0ef62cb39 (diff) | |
Local partitioned indexes
When CREATE INDEX is run on a partitioned table, create catalog entries
for an index on the partitioned table (which is just a placeholder since
the table proper has no data of its own), and recurse to create actual
indexes on the existing partitions; create them in future partitions
also.
As a convenience gadget, if the new index definition matches some
existing index in partitions, these are picked up and used instead of
creating new ones. Whichever way these indexes come about, they become
attached to the index on the parent table and are dropped alongside it,
and cannot be dropped on isolation unless they are detached first.
To support pg_dump'ing these indexes, add commands
CREATE INDEX ON ONLY <table>
(which creates the index on the parent partitioned table, without
recursing) and
ALTER INDEX ATTACH PARTITION
(which is used after the indexes have been created individually on each
partition, to attach them to the parent index). These reconstruct prior
database state exactly.
Reviewed-by: (in alphabetical order) Peter Eisentraut, Robert Haas, Amit
Langote, Jesper Pedersen, Simon Riggs, David Rowley
Discussion: https://postgr.es/m/[email protected]
Diffstat (limited to 'src/backend')
23 files changed, 1429 insertions, 145 deletions
diff --git a/src/backend/access/common/reloptions.c b/src/backend/access/common/reloptions.c index 425bc5d06ea..274f7aa8e97 100644 --- a/src/backend/access/common/reloptions.c +++ b/src/backend/access/common/reloptions.c @@ -993,6 +993,7 @@ extractRelOptions(HeapTuple tuple, TupleDesc tupdesc, options = view_reloptions(datum, false); break; case RELKIND_INDEX: + case RELKIND_PARTITIONED_INDEX: options = index_reloptions(amoptions, datum, false); break; case RELKIND_FOREIGN_TABLE: diff --git a/src/backend/access/heap/heapam.c b/src/backend/access/heap/heapam.c index dbc8f2d6c71..be263850cd3 100644 --- a/src/backend/access/heap/heapam.c +++ b/src/backend/access/heap/heapam.c @@ -1293,7 +1293,8 @@ heap_open(Oid relationId, LOCKMODE lockmode) r = relation_open(relationId, lockmode); - if (r->rd_rel->relkind == RELKIND_INDEX) + if (r->rd_rel->relkind == RELKIND_INDEX || + r->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is an index", @@ -1321,7 +1322,8 @@ heap_openrv(const RangeVar *relation, LOCKMODE lockmode) r = relation_openrv(relation, lockmode); - if (r->rd_rel->relkind == RELKIND_INDEX) + if (r->rd_rel->relkind == RELKIND_INDEX || + r->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is an index", @@ -1353,7 +1355,8 @@ heap_openrv_extended(const RangeVar *relation, LOCKMODE lockmode, if (r) { - if (r->rd_rel->relkind == RELKIND_INDEX) + if (r->rd_rel->relkind == RELKIND_INDEX || + r->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is an index", diff --git a/src/backend/access/index/indexam.c b/src/backend/access/index/indexam.c index 1b61cd95151..91247f0fa59 100644 --- a/src/backend/access/index/indexam.c +++ b/src/backend/access/index/indexam.c @@ -154,7 +154,8 @@ index_open(Oid relationId, LOCKMODE lockmode) r = relation_open(relationId, lockmode); - if (r->rd_rel->relkind != RELKIND_INDEX) + if (r->rd_rel->relkind != RELKIND_INDEX && + r->rd_rel->relkind != RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is not an index", diff --git a/src/backend/bootstrap/bootparse.y b/src/backend/bootstrap/bootparse.y index 8c52846a92e..dfd53fa054a 100644 --- a/src/backend/bootstrap/bootparse.y +++ b/src/backend/bootstrap/bootparse.y @@ -321,6 +321,7 @@ Boot_DeclareIndexStmt: DefineIndex(relationId, stmt, $4, + InvalidOid, false, false, false, @@ -365,6 +366,7 @@ Boot_DeclareUniqueIndexStmt: DefineIndex(relationId, stmt, $5, + InvalidOid, false, false, false, diff --git a/src/backend/catalog/aclchk.c b/src/backend/catalog/aclchk.c index fac80612b80..50a2e2681b6 100644 --- a/src/backend/catalog/aclchk.c +++ b/src/backend/catalog/aclchk.c @@ -1824,7 +1824,8 @@ ExecGrant_Relation(InternalGrant *istmt) pg_class_tuple = (Form_pg_class) GETSTRUCT(tuple); /* Not sensible to grant on an index */ - if (pg_class_tuple->relkind == RELKIND_INDEX) + if (pg_class_tuple->relkind == RELKIND_INDEX || + pg_class_tuple->relkind == RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is an index", @@ -5405,7 +5406,8 @@ recordExtObjInitPriv(Oid objoid, Oid classoid) pg_class_tuple = (Form_pg_class) GETSTRUCT(tuple); /* Indexes don't have permissions */ - if (pg_class_tuple->relkind == RELKIND_INDEX) + if (pg_class_tuple->relkind == RELKIND_INDEX || + pg_class_tuple->relkind == RELKIND_PARTITIONED_INDEX) return; /* Composite types don't have permissions either */ @@ -5690,7 +5692,8 @@ removeExtObjInitPriv(Oid objoid, Oid classoid) pg_class_tuple = (Form_pg_class) GETSTRUCT(tuple); /* Indexes don't have permissions */ - if (pg_class_tuple->relkind == RELKIND_INDEX) + if (pg_class_tuple->relkind == RELKIND_INDEX || + pg_class_tuple->relkind == RELKIND_PARTITIONED_INDEX) return; /* Composite types don't have permissions either */ diff --git a/src/backend/catalog/dependency.c b/src/backend/catalog/dependency.c index 269111b4c16..be60270ea5f 100644 --- a/src/backend/catalog/dependency.c +++ b/src/backend/catalog/dependency.c @@ -582,6 +582,7 @@ findDependentObjects(const ObjectAddress *object, /* FALL THRU */ case DEPENDENCY_INTERNAL: + case DEPENDENCY_INTERNAL_AUTO: /* * This object is part of the internal implementation of @@ -633,6 +634,14 @@ findDependentObjects(const ObjectAddress *object, * transform this deletion request into a delete of this * owning object. * + * For INTERNAL_AUTO dependencies, we don't enforce this; + * in other words, we don't follow the links back to the + * owning object. + */ + if (foundDep->deptype == DEPENDENCY_INTERNAL_AUTO) + break; + + /* * First, release caller's lock on this object and get * deletion lock on the owning object. (We must release * caller's lock to avoid deadlock against a concurrent @@ -675,6 +684,7 @@ findDependentObjects(const ObjectAddress *object, /* And we're done here. */ systable_endscan(scan); return; + case DEPENDENCY_PIN: /* @@ -762,6 +772,7 @@ findDependentObjects(const ObjectAddress *object, case DEPENDENCY_AUTO_EXTENSION: subflags = DEPFLAG_AUTO; break; + case DEPENDENCY_INTERNAL_AUTO: case DEPENDENCY_INTERNAL: subflags = DEPFLAG_INTERNAL; break; @@ -1109,7 +1120,8 @@ doDeletion(const ObjectAddress *object, int flags) { char relKind = get_rel_relkind(object->objectId); - if (relKind == RELKIND_INDEX) + if (relKind == RELKIND_INDEX || + relKind == RELKIND_PARTITIONED_INDEX) { bool concurrent = ((flags & PERFORM_DELETION_CONCURRENTLY) != 0); diff --git a/src/backend/catalog/heap.c b/src/backend/catalog/heap.c index 089b7965f2a..99f4d59863c 100644 --- a/src/backend/catalog/heap.c +++ b/src/backend/catalog/heap.c @@ -294,6 +294,7 @@ heap_create(const char *relname, case RELKIND_COMPOSITE_TYPE: case RELKIND_FOREIGN_TABLE: case RELKIND_PARTITIONED_TABLE: + case RELKIND_PARTITIONED_INDEX: create_storage = false; /* diff --git a/src/backend/catalog/index.c b/src/backend/catalog/index.c index 007b929a6fa..f0223416adf 100644 --- a/src/backend/catalog/index.c +++ b/src/backend/catalog/index.c @@ -41,6 +41,8 @@ #include "catalog/pg_collation.h" #include "catalog/pg_constraint.h" #include "catalog/pg_constraint_fn.h" +#include "catalog/pg_depend.h" +#include "catalog/pg_inherits_fn.h" #include "catalog/pg_operator.h" #include "catalog/pg_opclass.h" #include "catalog/pg_tablespace.h" @@ -55,6 +57,7 @@ #include "nodes/nodeFuncs.h" #include "optimizer/clauses.h" #include "parser/parser.h" +#include "rewrite/rewriteManip.h" #include "storage/bufmgr.h" #include "storage/lmgr.h" #include "storage/predicate.h" @@ -110,6 +113,7 @@ static void InitializeAttributeOids(Relation indexRelation, int numatts, Oid indexoid); static void AppendAttributeTuples(Relation indexRelation, int numatts); static void UpdateIndexRelation(Oid indexoid, Oid heapoid, + Oid parentIndexId, IndexInfo *indexInfo, Oid *collationOids, Oid *classOids, @@ -117,7 +121,8 @@ static void UpdateIndexRelation(Oid indexoid, Oid heapoid, bool primary, bool isexclusion, bool immediate, - bool isvalid); + bool isvalid, + bool isready); static void index_update_stats(Relation rel, bool hasindex, bool isprimary, double reltuples); @@ -563,6 +568,7 @@ AppendAttributeTuples(Relation indexRelation, int numatts) static void UpdateIndexRelation(Oid indexoid, Oid heapoid, + Oid parentIndexOid, IndexInfo *indexInfo, Oid *collationOids, Oid *classOids, @@ -570,7 +576,8 @@ UpdateIndexRelation(Oid indexoid, bool primary, bool isexclusion, bool immediate, - bool isvalid) + bool isvalid, + bool isready) { int2vector *indkey; oidvector *indcollation; @@ -644,8 +651,7 @@ UpdateIndexRelation(Oid indexoid, values[Anum_pg_index_indisclustered - 1] = BoolGetDatum(false); values[Anum_pg_index_indisvalid - 1] = BoolGetDatum(isvalid); values[Anum_pg_index_indcheckxmin - 1] = BoolGetDatum(false); - /* we set isvalid and isready the same way */ - values[Anum_pg_index_indisready - 1] = BoolGetDatum(isvalid); + values[Anum_pg_index_indisready - 1] = BoolGetDatum(isready); values[Anum_pg_index_indislive - 1] = BoolGetDatum(true); values[Anum_pg_index_indisreplident - 1] = BoolGetDatum(false); values[Anum_pg_index_indkey - 1] = PointerGetDatum(indkey); @@ -682,6 +688,8 @@ UpdateIndexRelation(Oid indexoid, * indexRelationId: normally, pass InvalidOid to let this routine * generate an OID for the index. During bootstrap this may be * nonzero to specify a preselected OID. + * parentIndexRelid: if creating an index partition, the OID of the + * parent index; otherwise InvalidOid. * relFileNode: normally, pass InvalidOid to get new storage. May be * nonzero to attach an existing valid build. * indexInfo: same info executor uses to insert into the index @@ -707,6 +715,8 @@ UpdateIndexRelation(Oid indexoid, * INDEX_CREATE_IF_NOT_EXISTS: * do not throw an error if a relation with the same name * already exists. + * INDEX_CREATE_PARTITIONED: + * create a partitioned index (table must be partitioned) * constr_flags: flags passed to index_constraint_create * (only if INDEX_CREATE_ADD_CONSTRAINT is set) * allow_system_table_mods: allow table to be a system catalog @@ -718,6 +728,7 @@ Oid index_create(Relation heapRelation, const char *indexRelationName, Oid indexRelationId, + Oid parentIndexRelid, Oid relFileNode, IndexInfo *indexInfo, List *indexColNames, @@ -743,12 +754,18 @@ index_create(Relation heapRelation, int i; char relpersistence; bool isprimary = (flags & INDEX_CREATE_IS_PRIMARY) != 0; + bool invalid = (flags & INDEX_CREATE_INVALID) != 0; bool concurrent = (flags & INDEX_CREATE_CONCURRENT) != 0; + bool partitioned = (flags & INDEX_CREATE_PARTITIONED) != 0; + char relkind; /* constraint flags can only be set when a constraint is requested */ Assert((constr_flags == 0) || ((flags & INDEX_CREATE_ADD_CONSTRAINT) != 0)); + /* partitioned indexes must never be "built" by themselves */ + Assert(!partitioned || (flags & INDEX_CREATE_SKIP_BUILD)); + relkind = partitioned ? RELKIND_PARTITIONED_INDEX : RELKIND_INDEX; is_exclusion = (indexInfo->ii_ExclusionOps != NULL); pg_class = heap_open(RelationRelationId, RowExclusiveLock); @@ -866,9 +883,9 @@ index_create(Relation heapRelation, } /* - * create the index relation's relcache entry and physical disk file. (If - * we fail further down, it's the smgr's responsibility to remove the disk - * file again.) + * create the index relation's relcache entry and, if necessary, the + * physical disk file. (If we fail further down, it's the smgr's + * responsibility to remove the disk file again, if any.) */ indexRelation = heap_create(indexRelationName, namespaceId, @@ -876,7 +893,7 @@ index_create(Relation heapRelation, indexRelationId, relFileNode, indexTupDesc, - RELKIND_INDEX, + relkind, relpersistence, shared_relation, mapped_relation, @@ -933,12 +950,18 @@ index_create(Relation heapRelation, * (Or, could define a rule to maintain the predicate) --Nels, Feb '92 * ---------------- */ - UpdateIndexRelation(indexRelationId, heapRelationId, indexInfo, + UpdateIndexRelation(indexRelationId, heapRelationId, parentIndexRelid, + indexInfo, collationObjectId, classObjectId, coloptions, isprimary, is_exclusion, (constr_flags & INDEX_CONSTR_CREATE_DEFERRABLE) == 0, + !concurrent && !invalid, !concurrent); + /* update pg_inherits, if needed */ + if (OidIsValid(parentIndexRelid)) + StoreSingleInheritance(indexRelationId, parentIndexRelid, 1); + /* * Register constraint and dependencies for the index. * @@ -990,6 +1013,9 @@ index_create(Relation heapRelation, else { bool have_simple_col = false; + DependencyType deptype; + + deptype = OidIsValid(parentIndexRelid) ? DEPENDENCY_INTERNAL_AUTO : DEPENDENCY_AUTO; /* Create auto dependencies on simply-referenced columns */ for (i = 0; i < indexInfo->ii_NumIndexAttrs; i++) @@ -1000,7 +1026,7 @@ index_create(Relation heapRelation, referenced.objectId = heapRelationId; referenced.objectSubId = indexInfo->ii_KeyAttrNumbers[i]; - recordDependencyOn(&myself, &referenced, DEPENDENCY_AUTO); + recordDependencyOn(&myself, &referenced, deptype); have_simple_col = true; } @@ -1018,10 +1044,20 @@ index_create(Relation heapRelation, referenced.objectId = heapRelationId; referenced.objectSubId = 0; - recordDependencyOn(&myself, &referenced, DEPENDENCY_AUTO); + recordDependencyOn(&myself, &referenced, deptype); } } + /* Store dependency on parent index, if any */ + if (OidIsValid(parentIndexRelid)) + { + referenced.classId = RelationRelationId; + referenced.objectId = parentIndexRelid; + referenced.objectSubId = 0; + + recordDependencyOn(&myself, &referenced, DEPENDENCY_INTERNAL_AUTO); + } + /* Store dependency on collations */ /* The default collation is pinned, so don't bother recording it */ for (i = 0; i < indexInfo->ii_NumIndexAttrs; i++) @@ -1567,9 +1603,10 @@ index_drop(Oid indexId, bool concurrent) } /* - * Schedule physical removal of the files + * Schedule physical removal of the files (if any) */ - RelationDropStorage(userIndexRelation); + if (userIndexRelation->rd_rel->relkind != RELKIND_PARTITIONED_INDEX) + RelationDropStorage(userIndexRelation); /* * Close and flush the index's relcache entry, to ensure relcache doesn't @@ -1614,6 +1651,11 @@ index_drop(Oid indexId, bool concurrent) DeleteRelationTuple(indexId); /* + * fix INHERITS relation + */ + DeleteInheritsTuple(indexId, InvalidOid); + + /* * We are presently too lazy to attempt to compute the new correct value * of relhasindex (the next VACUUM will fix it if necessary). So there is * no need to update the pg_class tuple for the owning relation. But we @@ -1706,12 +1748,120 @@ BuildIndexInfo(Relation index) ii->ii_BrokenHotChain = false; /* set up for possible use by index AM */ + ii->ii_Am = index->rd_rel->relam; ii->ii_AmCache = NULL; ii->ii_Context = CurrentMemoryContext; return ii; } +/* + * CompareIndexInfo + * Return whether the properties of two indexes (in different tables) + * indicate that they have the "same" definitions. + * + * Note: passing collations and opfamilies separately is a kludge. Adding + * them to IndexInfo may result in better coding here and elsewhere. + * + * Use convert_tuples_by_name_map(index2, index1) to build the attmap. + */ +bool +CompareIndexInfo(IndexInfo *info1, IndexInfo *info2, + Oid *collations1, Oid *collations2, + Oid *opfamilies1, Oid *opfamilies2, + AttrNumber *attmap, int maplen) +{ + int i; + + if (info1->ii_Unique != info2->ii_Unique) + return false; + + /* indexes are only equivalent if they have the same access method */ + if (info1->ii_Am != info2->ii_Am) + return false; + + /* and same number of attributes */ + if (info1->ii_NumIndexAttrs != info2->ii_NumIndexAttrs) + return false; + + /* + * and columns match through the attribute map (actual attribute numbers + * might differ!) Note that this implies that index columns that are + * expressions appear in the same positions. We will next compare the + * expressions themselves. + */ + for (i = 0; i < info1->ii_NumIndexAttrs; i++) + { + if (maplen < info2->ii_KeyAttrNumbers[i]) + elog(ERROR, "incorrect attribute map"); + + if (attmap[info2->ii_KeyAttrNumbers[i] - 1] != + info1->ii_KeyAttrNumbers[i]) + return false; + + if (collations1[i] != collations2[i]) + return false; + if (opfamilies1[i] != opfamilies2[i]) + return false; + } + + /* + * For expression indexes: either both are expression indexes, or neither + * is; if they are, make sure the expressions match. + */ + if ((info1->ii_Expressions != NIL) != (info2->ii_Expressions != NIL)) + return false; + if (info1->ii_Expressions != NIL) + { + bool found_whole_row; + Node *mapped; + + mapped = map_variable_attnos((Node *) info2->ii_Expressions, + 1, 0, attmap, maplen, + InvalidOid, &found_whole_row); + if (found_whole_row) + { + /* + * we could throw an error here, but seems out of scope for this + * routine. + */ + return false; + } + + if (!equal(info1->ii_Expressions, mapped)) + return false; + } + + /* Partial index predicates must be identical, if they exist */ + if ((info1->ii_Predicate == NULL) != (info2->ii_Predicate == NULL)) + return false; + if (info1->ii_Predicate != NULL) + { + bool found_whole_row; + Node *mapped; + + mapped = map_variable_attnos((Node *) info2->ii_Predicate, + 1, 0, attmap, maplen, + InvalidOid, &found_whole_row); + if (found_whole_row) + { + /* + * we could throw an error here, but seems out of scope for this + * routine. + */ + return false; + } + if (!equal(info1->ii_Predicate, mapped)) + return false; + } + + /* No support currently for comparing exclusion indexes. */ + if (info1->ii_ExclusionOps != NULL || info2->ii_ExclusionOps != NULL) + return false; + + return true; +} + /* ---------------- * BuildSpeculativeIndexInfo * Add extra state to IndexInfo record @@ -1934,6 +2084,9 @@ index_update_stats(Relation rel, elog(ERROR, "could not find tuple for relation %u", relid); rd_rel = (Form_pg_class) GETSTRUCT(tuple); + /* Should this be a more comprehensive test? */ + Assert(rd_rel->relkind != RELKIND_PARTITIONED_INDEX); + /* Apply required updates, if any, to copied tuple */ dirty = false; @@ -3344,6 +3497,14 @@ reindex_index(Oid indexId, bool skip_constraint_checks, char persistence, iRel = index_open(indexId, AccessExclusiveLock); /* + * The case of reindexing partitioned tables and indexes is handled + * differently by upper layers, so this case shouldn't arise. + */ + if (iRel->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) + elog(ERROR, "unsupported relation kind for index \"%s\"", + RelationGetRelationName(iRel)); + + /* * Don't allow reindex on temp tables of other backends ... their local * buffer manager is not going to cope. */ @@ -3542,6 +3703,22 @@ reindex_relation(Oid relid, int flags, int options) */ rel = heap_open(relid, ShareLock); + /* + * This may be useful when implemented someday; but that day is not today. + * For now, avoid erroring out when called in a multi-table context + * (REINDEX SCHEMA) and happen to come across a partitioned table. The + * partitions may be reindexed on their own anyway. + */ + if (rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE) + { + ereport(WARNING, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("REINDEX of partitioned tables is not yet implemented, skipping \"%s\"", + RelationGetRelationName(rel)))); + heap_close(rel, ShareLock); + return false; + } + toast_relid = rel->rd_rel->reltoastrelid; /* diff --git a/src/backend/catalog/objectaddress.c b/src/backend/catalog/objectaddress.c index bc999ca3c43..7576606c1b2 100644 --- a/src/backend/catalog/objectaddress.c +++ b/src/backend/catalog/objectaddress.c @@ -1217,7 +1217,8 @@ get_relation_by_qualified_name(ObjectType objtype, List *object, switch (objtype) { case OBJECT_INDEX: - if (relation->rd_rel->relkind != RELKIND_INDEX) + if (relation->rd_rel->relkind != RELKIND_INDEX && + relation->rd_rel->relkind != RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is not an index", @@ -3483,6 +3484,7 @@ getRelationDescription(StringInfo buffer, Oid relid) relname); break; case RELKIND_INDEX: + case RELKIND_PARTITIONED_INDEX: appendStringInfo(buffer, _("index %s"), relname); break; @@ -3957,6 +3959,7 @@ getRelationTypeDescription(StringInfo buffer, Oid relid, int32 objectSubId) appendStringInfoString(buffer, "table"); break; case RELKIND_INDEX: + case RELKIND_PARTITIONED_INDEX: appendStringInfoString(buffer, "index"); break; case RELKIND_SEQUENCE: diff --git a/src/backend/catalog/pg_depend.c b/src/backend/catalog/pg_depend.c index 9dfbe123b52..2ea05f350b3 100644 --- a/src/backend/catalog/pg_depend.c +++ b/src/backend/catalog/pg_depend.c @@ -656,14 +656,19 @@ get_constraint_index(Oid constraintId) /* * We assume any internal dependency of an index on the constraint - * must be what we are looking for. (The relkind test is just - * paranoia; there shouldn't be any such dependencies otherwise.) + * must be what we are looking for. */ if (deprec->classid == RelationRelationId && deprec->objsubid == 0 && - deprec->deptype == DEPENDENCY_INTERNAL && - get_rel_relkind(deprec->objid) == RELKIND_INDEX) + deprec->deptype == DEPENDENCY_INTERNAL) { + char relkind = get_rel_relkind(deprec->objid); + + /* This is pure paranoia; there shouldn't be any such */ + if (relkind != RELKIND_INDEX && + relkind != RELKIND_PARTITIONED_INDEX) + break; + indexId = deprec->objid; break; } diff --git a/src/backend/catalog/pg_inherits.c b/src/backend/catalog/pg_inherits.c index b32d677347f..5a5beb9273e 100644 --- a/src/backend/catalog/pg_inherits.c +++ b/src/backend/catalog/pg_inherits.c @@ -405,3 +405,83 @@ typeInheritsFrom(Oid subclassTypeId, Oid superclassTypeId) return result; } + +/* + * Create a single pg_inherits row with the given data + */ +void +StoreSingleInheritance(Oid relationId, Oid parentOid, int32 seqNumber) +{ + Datum values[Natts_pg_inherits]; + bool nulls[Natts_pg_inherits]; + HeapTuple tuple; + Relation inhRelation; + + inhRelation = heap_open(InheritsRelationId, RowExclusiveLock); + + /* + * Make the pg_inherits entry + */ + values[Anum_pg_inherits_inhrelid - 1] = ObjectIdGetDatum(relationId); + values[Anum_pg_inherits_inhparent - 1] = ObjectIdGetDatum(parentOid); + values[Anum_pg_inherits_inhseqno - 1] = Int32GetDatum(seqNumber); + + memset(nulls, 0, sizeof(nulls)); + + tuple = heap_form_tuple(RelationGetDescr(inhRelation), values, nulls); + + CatalogTupleInsert(inhRelation, tuple); + + heap_freetuple(tuple); + + heap_close(inhRelation, RowExclusiveLock); +} + +/* + * DeleteInheritsTuple + * + * Delete pg_inherits tuples with the given inhrelid. inhparent may be given + * as InvalidOid, in which case all tuples matching inhrelid are deleted; + * otherwise only delete tuples with the specified inhparent. + * + * Returns whether at least one row was deleted. + */ +bool +DeleteInheritsTuple(Oid inhrelid, Oid inhparent) +{ + bool found = false; + Relation catalogRelation; + ScanKeyData key; + SysScanDesc scan; + HeapTuple inheritsTuple; + + /* + * Find pg_inherits entries by inhrelid. + */ + catalogRelation = heap_open(InheritsRelationId, RowExclusiveLock); + ScanKeyInit(&key, + Anum_pg_inherits_inhrelid, + BTEqualStrategyNumber, F_OIDEQ, + ObjectIdGetDatum(inhrelid)); + scan = systable_beginscan(catalogRelation, InheritsRelidSeqnoIndexId, + true, NULL, 1, &key); + + while (HeapTupleIsValid(inheritsTuple = systable_getnext(scan))) + { + Oid parent; + + /* Compare inhparent if it was given, and do the actual deletion. */ + parent = ((Form_pg_inherits) GETSTRUCT(inheritsTuple))->inhparent; + if (!OidIsValid(inhparent) || parent == inhparent) + { + CatalogTupleDelete(catalogRelation, &inheritsTuple->t_self); + found = true; + } + } + + /* Done */ + systable_endscan(scan); + heap_close(catalogRelation, RowExclusiveLock); + + return found; +} diff --git a/src/backend/catalog/toasting.c b/src/backend/catalog/toasting.c index 0b4b5631a1b..cf37011b73f 100644 --- a/src/backend/catalog/toasting.c +++ b/src/backend/catalog/toasting.c @@ -315,6 +315,7 @@ create_toast_table(Relation rel, Oid toastOid, Oid toastIndexOid, indexInfo->ii_ReadyForInserts = true; indexInfo->ii_Concurrent = false; indexInfo->ii_BrokenHotChain = false; + indexInfo->ii_Am = BTREE_AM_OID; indexInfo->ii_AmCache = NULL; indexInfo->ii_Context = CurrentMemoryContext; @@ -328,6 +329,7 @@ create_toast_table(Relation rel, Oid toastOid, Oid toastIndexOid, coloptions[1] = 0; index_create(toast_rel, toast_idxname, toastIndexOid, InvalidOid, + InvalidOid, indexInfo, list_make2("chunk_id", "chunk_seq"), BTREE_AM_OID, diff --git a/src/backend/commands/indexcmds.c b/src/backend/commands/indexcmds.c index 9e6ba920086..8118a39a7b1 100644 --- a/src/backend/commands/indexcmds.c +++ b/src/backend/commands/indexcmds.c @@ -23,7 +23,10 @@ #include "catalog/catalog.h" #include "catalog/index.h" #include "catalog/indexing.h" +#include "catalog/partition.h" #include "catalog/pg_am.h" +#include "catalog/pg_inherits.h" +#include "catalog/pg_inherits_fn.h" #include "catalog/pg_opclass.h" #include "catalog/pg_opfamily.h" #include "catalog/pg_tablespace.h" @@ -35,6 +38,7 @@ #include "commands/tablespace.h" #include "mb/pg_wchar.h" #include "miscadmin.h" +#include "nodes/makefuncs.h" #include "nodes/nodeFuncs.h" #include "optimizer/clauses.h" #include "optimizer/planner.h" @@ -42,6 +46,7 @@ #include "parser/parse_coerce.h" #include "parser/parse_func.h" #include "parser/parse_oper.h" +#include "rewrite/rewriteManip.h" #include "storage/lmgr.h" #include "storage/proc.h" #include "storage/procarray.h" @@ -77,6 +82,7 @@ static char *ChooseIndexNameAddition(List *colnames); static List *ChooseIndexColumnNames(List *indexElems); static void RangeVarCallbackForReindexIndex(const RangeVar *relation, Oid relId, Oid oldRelId, void *arg); +static void ReindexPartitionedIndex(Relation parentIdx); /* * CheckIndexCompatible @@ -183,6 +189,7 @@ CheckIndexCompatible(Oid oldId, indexInfo->ii_ExclusionOps = NULL; indexInfo->ii_ExclusionProcs = NULL; indexInfo->ii_ExclusionStrats = NULL; + indexInfo->ii_Am = accessMethodId; indexInfo->ii_AmCache = NULL; indexInfo->ii_Context = CurrentMemoryContext; typeObjectId = (Oid *) palloc(numberOfAttributes * sizeof(Oid)); @@ -292,14 +299,15 @@ CheckIndexCompatible(Oid oldId, * 'stmt': IndexStmt describing the properties of the new index. * 'indexRelationId': normally InvalidOid, but during bootstrap can be * nonzero to specify a preselected OID for the index. + * 'parentIndexId': the OID of the parent index; InvalidOid if not the child + * of a partitioned index. * 'is_alter_table': this is due to an ALTER rather than a CREATE operation. * 'check_rights': check for CREATE rights in namespace and tablespace. (This * should be true except when ALTER is deleting/recreating an index.) * 'check_not_in_use': check for table not already in use in current session. * This should be true unless caller is holding the table open, in which * case the caller had better have checked it earlier. - * 'skip_build': make the catalog entries but leave the index file empty; - * it will be filled later. + * 'skip_build': make the catalog entries but don't create the index files * 'quiet': suppress the NOTICE chatter ordinarily provided for constraints. * * Returns the object address of the created index. @@ -308,6 +316,7 @@ ObjectAddress DefineIndex(Oid relationId, IndexStmt *stmt, Oid indexRelationId, + Oid parentIndexId, bool is_alter_table, bool check_rights, bool check_not_in_use, @@ -330,6 +339,7 @@ DefineIndex(Oid relationId, IndexAmRoutine *amRoutine; bool amcanorder; amoptions_function amoptions; + bool partitioned; Datum reloptions; int16 *coloptions; IndexInfo *indexInfo; @@ -382,23 +392,56 @@ DefineIndex(Oid relationId, { case RELKIND_RELATION: case RELKIND_MATVIEW: + case RELKIND_PARTITIONED_TABLE: /* OK */ break; case RELKIND_FOREIGN_TABLE: + /* + * Custom error message for FOREIGN TABLE since the term is close + * to a regular table and can confuse the user. + */ ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("cannot create index on foreign table \"%s\"", RelationGetRelationName(rel)))); - case RELKIND_PARTITIONED_TABLE: - ereport(ERROR, - (errcode(ERRCODE_WRONG_OBJECT_TYPE), - errmsg("cannot create index on partitioned table \"%s\"", - RelationGetRelationName(rel)))); default: ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is not a table or materialized view", RelationGetRelationName(rel)))); + break; + } + + /* + * Establish behavior for partitioned tables, and verify sanity of + * parameters. + * + * We do not build an actual index in this case; we only create a few + * catalog entries. The actual indexes are built by recursing for each + * partition. + */ + partitioned = rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE; + if (partitioned) + { + if (stmt->concurrent) + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("cannot create index on partitioned table \"%s\" concurrently", + RelationGetRelationName(rel)))); + if (stmt->unique) + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("cannot create unique index on partitioned table \"%s\"", + RelationGetRelationName(rel)))); + if (stmt->excludeOpNames) + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("cannot create exclusion constraints on partitioned table \"%s\"", + RelationGetRelationName(rel)))); + if (stmt->primary || stmt->isconstraint) + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("cannot create constraints on partitioned tables"))); } /* @@ -574,6 +617,7 @@ DefineIndex(Oid relationId, indexInfo->ii_ReadyForInserts = !stmt->concurrent; indexInfo->ii_Concurrent = stmt->concurrent; indexInfo->ii_BrokenHotChain = false; + indexInfo->ii_Am = accessMethodId; indexInfo->ii_AmCache = NULL; indexInfo->ii_Context = CurrentMemoryContext; @@ -665,19 +709,24 @@ DefineIndex(Oid relationId, /* * Make the catalog entries for the index, including constraints. This * step also actually builds the index, except if caller requested not to - * or in concurrent mode, in which case it'll be done later. + * or in concurrent mode, in which case it'll be done later, or + * doing a partitioned index (because those don't have storage). */ flags = constr_flags = 0; if (stmt->isconstraint) flags |= INDEX_CREATE_ADD_CONSTRAINT; - if (skip_build || stmt->concurrent) + if (skip_build || stmt->concurrent || partitioned) flags |= INDEX_CREATE_SKIP_BUILD; if (stmt->if_not_exists) flags |= INDEX_CREATE_IF_NOT_EXISTS; if (stmt->concurrent) flags |= INDEX_CREATE_CONCURRENT; + if (partitioned) + flags |= INDEX_CREATE_PARTITIONED; if (stmt->primary) flags |= INDEX_CREATE_IS_PRIMARY; + if (partitioned && stmt->relation && !stmt->relation->inh) + flags |= INDEX_CREATE_INVALID; if (stmt->deferrable) constr_flags |= INDEX_CONSTR_CREATE_DEFERRABLE; @@ -685,8 +734,8 @@ DefineIndex(Oid relationId, constr_flags |= INDEX_CONSTR_CREATE_INIT_DEFERRED; indexRelationId = - index_create(rel, indexRelationName, indexRelationId, stmt->oldNode, - indexInfo, indexColNames, + index_create(rel, indexRelationName, indexRelationId, parentIndexId, + stmt->oldNode, indexInfo, indexColNames, accessMethodId, tablespaceId, collationObjectId, classObjectId, coloptions, reloptions, @@ -706,6 +755,160 @@ DefineIndex(Oid relationId, CreateComments(indexRelationId, RelationRelationId, 0, stmt->idxcomment); + if (partitioned) + { + /* + * Unless caller specified to skip this step (via ONLY), process + * each partition to make sure they all contain a corresponding index. + * + * If we're called internally (no stmt->relation), recurse always. + */ + if (!stmt->relation || stmt->relation->inh) + { + PartitionDesc partdesc = RelationGetPartitionDesc(rel); + int nparts = partdesc->nparts; + Oid *part_oids = palloc(sizeof(Oid) * nparts); + bool invalidate_parent = false; + TupleDesc parentDesc; + Oid *opfamOids; + + memcpy(part_oids, partdesc->oids, sizeof(Oid) * nparts); + + parentDesc = CreateTupleDescCopy(RelationGetDescr(rel)); + opfamOids = palloc(sizeof(Oid) * numberOfAttributes); + for (i = 0; i < numberOfAttributes; i++) + opfamOids[i] = get_opclass_family(classObjectId[i]); + + heap_close(rel, NoLock); + + /* + * For each partition, scan all existing indexes; if one matches + * our index definition and is not already attached to some other + * parent index, attach it to the one we just created. + * + * If none matches, build a new index by calling ourselves + * recursively with the same options (except for the index name). + */ + for (i = 0; i < nparts; i++) + { + Oid childRelid = part_oids[i]; + Relation childrel; + List *childidxs; + ListCell *cell; + AttrNumber *attmap; + bool found = false; + int maplen; + + childrel = heap_open(childRelid, lockmode); + childidxs = RelationGetIndexList(childrel); + attmap = + convert_tuples_by_name_map(RelationGetDescr(childrel), + parentDesc, + gettext_noop("could not convert row type")); + maplen = parentDesc->natts; + + + foreach(cell, childidxs) + { + Oid cldidxid = lfirst_oid(cell); + Relation cldidx; + IndexInfo *cldIdxInfo; + + /* this index is already partition of another one */ + if (has_superclass(cldidxid)) + continue; + + cldidx = index_open(cldidxid, lockmode); + cldIdxInfo = BuildIndexInfo(cldidx); + if (CompareIndexInfo(cldIdxInfo, indexInfo, + cldidx->rd_indcollation, + collationObjectId, + cldidx->rd_opfamily, + opfamOids, + attmap, maplen)) + { + /* + * Found a match. Attach index to parent and we're + * done, but keep lock till commit. + */ + IndexSetParentIndex(cldidx, indexRelationId); + + if (!IndexIsValid(cldidx->rd_index)) + invalidate_parent = true; + + found = true; + index_close(cldidx, NoLock); + break; + } + + index_close(cldidx, lockmode); + } + + list_free(childidxs); + heap_close(childrel, NoLock); + + /* + * If no matching index was found, create our own. + */ + if (!found) + { + IndexStmt *childStmt = copyObject(stmt); + bool found_whole_row; + + childStmt->whereClause = + map_variable_attnos(stmt->whereClause, 1, 0, + attmap, maplen, + InvalidOid, &found_whole_row); + if (found_whole_row) + elog(ERROR, "cannot convert whole-row table reference"); + + childStmt->idxname = NULL; + childStmt->relationId = childRelid; + DefineIndex(childRelid, childStmt, + InvalidOid, /* no predefined OID */ + indexRelationId, /* this is our child */ + false, check_rights, check_not_in_use, + false, quiet); + } + + pfree(attmap); + } + + /* + * The pg_index row we inserted for this index was marked + * indisvalid=true. But if we attached an existing index that + * is invalid, this is incorrect, so update our row to + * invalid too. + */ + if (invalidate_parent) + { + Relation pg_index = heap_open(IndexRelationId, RowExclusiveLock); + HeapTuple tup, + newtup; + + tup = SearchSysCache1(INDEXRELID, + ObjectIdGetDatum(indexRelationId)); + if (!tup) + elog(ERROR, "cache lookup failed for index %u", + indexRelationId); + newtup = heap_copytuple(tup); + ((Form_pg_index) GETSTRUCT(newtup))->indisvalid = false; + CatalogTupleUpdate(pg_index, &tup->t_self, newtup); + ReleaseSysCache(tup); + heap_close(pg_index, RowExclusiveLock); + heap_freetuple(newtup); + } + } + else + heap_close(rel, NoLock); + + /* + * Indexes on partitioned tables are not themselves built, so we're + * done here. + */ + return address; + } + if (!stmt->concurrent) { /* Close the heap and we're done, in the non-concurrent case */ @@ -1765,7 +1968,7 @@ ChooseIndexColumnNames(List *indexElems) * ReindexIndex * Recreate a specific index. */ -Oid +void ReindexIndex(RangeVar *indexRelation, int options) { Oid indOid; @@ -1788,12 +1991,17 @@ ReindexIndex(RangeVar *indexRelation, int options) * lock on the index. */ irel = index_open(indOid, NoLock); + + if (irel->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) + { + ReindexPartitionedIndex(irel); + return; + } + persistence = irel->rd_rel->relpersistence; index_close(irel, NoLock); reindex_index(indOid, false, persistence, options); - - return indOid; } /* @@ -1832,7 +2040,8 @@ RangeVarCallbackForReindexIndex(const RangeVar *relation, relkind = get_rel_relkind(relId); if (!relkind) return; - if (relkind != RELKIND_INDEX) + if (relkind != RELKIND_INDEX && + relkind != RELKIND_PARTITIONED_INDEX) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is not an index", relation->relname))); @@ -1976,6 +2185,12 @@ ReindexMultipleTables(const char *objectName, ReindexObjectType objectKind, /* * Only regular tables and matviews can have indexes, so ignore any * other kind of relation. + * + * It is tempting to also consider partitioned tables here, but that + * has the problem that if the children are in the same schema, they + * would be processed twice. Maybe we could have a separate list of + * partitioned tables, and expand that afterwards into relids, + * ignoring any duplicates. */ if (classtuple->relkind != RELKIND_RELATION && classtuple->relkind != RELKIND_MATVIEW) @@ -2038,3 +2253,155 @@ ReindexMultipleTables(const char *objectName, ReindexObjectType objectKind, MemoryContextDelete(private_context); } + +/* + * ReindexPartitionedIndex + * Reindex each child of the given partitioned index. + * + * Not yet implemented. + */ +static void +ReindexPartitionedIndex(Relation parentIdx) +{ + ereport(ERROR, + (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), + errmsg("REINDEX is not yet implemented for partitioned indexes"))); +} + +/* + * Insert or delete an appropriate pg_inherits tuple to make the given index + * be a partition of the indicated parent index. + * + * This also corrects the pg_depend information for the affected index. + */ +void +IndexSetParentIndex(Relation partitionIdx, Oid parentOid) +{ + Relation pg_inherits; + ScanKeyData key[2]; + SysScanDesc scan; + Oid partRelid = RelationGetRelid(partitionIdx); + HeapTuple tuple; + bool fix_dependencies; + + /* Make sure this is an index */ + Assert(partitionIdx->rd_rel->relkind == RELKIND_INDEX || + partitionIdx->rd_rel->relkind == RELKIND_PARTITIONED_INDEX); + + /* + * Scan pg_inherits for rows linking our index to some parent. + */ + pg_inherits = relation_open(InheritsRelationId, RowExclusiveLock); + ScanKeyInit(&key[0], + Anum_pg_inherits_inhrelid, + BTEqualStrategyNumber, F_OIDEQ, + ObjectIdGetDatum(partRelid)); + ScanKeyInit(&key[1], + Anum_pg_inherits_inhseqno, + BTEqualStrategyNumber, F_INT4EQ, + Int32GetDatum(1)); + scan = systable_beginscan(pg_inherits, InheritsRelidSeqnoIndexId, true, + NULL, 2, key); + tuple = systable_getnext(scan); + + if (!HeapTupleIsValid(tuple)) + { + if (parentOid == InvalidOid) + { + /* + * No pg_inherits row, and no parent wanted: nothing to do in + * this case. + */ + fix_dependencies = false; + } + else + { + Datum values[Natts_pg_inherits]; + bool isnull[Natts_pg_inherits]; + + /* + * No pg_inherits row exists, and we want a parent for this index, + * so insert it. + */ + values[Anum_pg_inherits_inhrelid - 1] = ObjectIdGetDatum(partRelid); + values[Anum_pg_inherits_inhparent - 1] = + ObjectIdGetDatum(parentOid); + values[Anum_pg_inherits_inhseqno - 1] = Int32GetDatum(1); + memset(isnull, false, sizeof(isnull)); + + tuple = heap_form_tuple(RelationGetDescr(pg_inherits), + values, isnull); + CatalogTupleInsert(pg_inherits, tuple); + + fix_dependencies = true; + } + } + else + { + Form_pg_inherits inhForm = (Form_pg_inherits) GETSTRUCT(tuple); + + if (parentOid == InvalidOid) + { + /* + * There exists a pg_inherits row, which we want to clear; do so. + */ + CatalogTupleDelete(pg_inherits, &tuple->t_self); + fix_dependencies = true; + } + else + { + /* + * A pg_inherits row exists. If it's the same we want, then we're + * good; if it differs, that amounts to a corrupt catalog and + * should not happen. + */ + if (inhForm->inhparent != parentOid) + { + /* unexpected: we should not get called in this case */ + elog(ERROR, "bogus pg_inherit row: inhrelid %u inhparent %u", + inhForm->inhrelid, inhForm->inhparent); + } + + /* already in the right state */ + fix_dependencies = false; + } + } + + /* done with pg_inherits */ + systable_endscan(scan); + relation_close(pg_inherits, RowExclusiveLock); + + if (fix_dependencies) + { + ObjectAddress partIdx; + + /* + * Insert/delete pg_depend rows. If setting a parent, add an + * INTERNAL_AUTO dependency to the parent index; if making standalone, + * remove all existing rows and put back the regular dependency on the + * table. + */ + ObjectAddressSet(partIdx, RelationRelationId, partRelid); + + if (OidIsValid(parentOid)) + { + ObjectAddress parentIdx; + + ObjectAddressSet(parentIdx, RelationRelationId, parentOid); + recordDependencyOn(&partIdx, &parentIdx, DEPENDENCY_INTERNAL_AUTO); + } + else + { + ObjectAddress partitionTbl; + + ObjectAddressSet(partitionTbl, RelationRelationId, + partitionIdx->rd_index->indrelid); + + deleteDependencyRecordsForClass(RelationRelationId, partRelid, + RelationRelationId, + DEPENDENCY_INTERNAL_AUTO); + + recordDependencyOn(&partIdx, &partitionTbl, DEPENDENCY_AUTO); + } + } +} diff --git a/src/backend/commands/tablecmds.c b/src/backend/commands/tablecmds.c index 59806349cc7..57ee112653f 100644 --- a/src/backend/commands/tablecmds.c +++ b/src/backend/commands/tablecmds.c @@ -266,6 +266,12 @@ static const struct dropmsgstrings dropmsgstringarray[] = { gettext_noop("table \"%s\" does not exist, skipping"), gettext_noop("\"%s\" is not a table"), gettext_noop("Use DROP TABLE to remove a table.")}, + {RELKIND_PARTITIONED_INDEX, + ERRCODE_UNDEFINED_OBJECT, + gettext_noop("index \"%s\" does not exist"), + gettext_noop("index \"%s\" does not exist, skipping"), + gettext_noop("\"%s\" is not an index"), + gettext_noop("Use DROP INDEX to remove an index.")}, {'\0', 0, NULL, NULL, NULL, NULL} }; @@ -284,6 +290,7 @@ struct DropRelationCallbackState #define ATT_INDEX 0x0008 #define ATT_COMPOSITE_TYPE 0x0010 #define ATT_FOREIGN_TABLE 0x0020 +#define ATT_PARTITIONED_INDEX 0x0040 /* * Partition tables are expected to be dropped when the parent partitioned @@ -475,11 +482,17 @@ static void CreateInheritance(Relation child_rel, Relation parent_rel); static void RemoveInheritance(Relation child_rel, Relation parent_rel); static ObjectAddress ATExecAttachPartition(List **wqueue, Relation rel, PartitionCmd *cmd); +static void AttachPartitionEnsureIndexes(Relation rel, Relation attachrel); static void ValidatePartitionConstraints(List **wqueue, Relation scanrel, List *scanrel_children, List *partConstraint, bool validate_default); static ObjectAddress ATExecDetachPartition(Relation rel, RangeVar *name); +static ObjectAddress ATExecAttachPartitionIdx(List **wqueue, Relation rel, + RangeVar *name); +static void validatePartitionedIndex(Relation partedIdx, Relation partedTbl); +static void refuseDupeIndexAttach(Relation parentIdx, Relation partIdx, + Relation partitionTbl); /* ---------------------------------------------------------------- @@ -897,6 +910,53 @@ DefineRelation(CreateStmt *stmt, char relkind, Oid ownerId, StorePartitionKey(rel, strategy, partnatts, partattrs, partexprs, partopclass, partcollation); + + /* make it all visible */ + CommandCounterIncrement(); + } + + /* + * If we're creating a partition, create now all the indexes defined in + * the parent. We can't do it earlier, because DefineIndex wants to know + * the partition key which we just stored. + */ + if (stmt->partbound) + { + Oid parentId = linitial_oid(inheritOids); + Relation parent; + List *idxlist; + ListCell *cell; + + /* Already have strong enough lock on the parent */ + parent = heap_open(parentId, NoLock); + idxlist = RelationGetIndexList(parent); + + /* + * For each index in the parent table, create one in the partition + */ + foreach(cell, idxlist) + { + Relation idxRel = index_open(lfirst_oid(cell), AccessShareLock); + AttrNumber *attmap; + IndexStmt *idxstmt; + + attmap = convert_tuples_by_name_map(RelationGetDescr(rel), + RelationGetDescr(parent), + gettext_noop("could not convert row type")); + idxstmt = + generateClonedIndexStmt(NULL, RelationGetRelid(rel), idxRel, + attmap, RelationGetDescr(rel)->natts); + DefineIndex(RelationGetRelid(rel), + idxstmt, + InvalidOid, + RelationGetRelid(idxRel), + false, false, false, false, false); + + index_close(idxRel, AccessShareLock); + } + + list_free(idxlist); + heap_close(parent, NoLock); } /* @@ -1179,10 +1239,13 @@ RangeVarCallbackForDropRelation(const RangeVar *rel, Oid relOid, Oid oldRelOid, * but RemoveRelations() can only pass one relkind for a given relation. * It chooses RELKIND_RELATION for both regular and partitioned tables. * That means we must be careful before giving the wrong type error when - * the relation is RELKIND_PARTITIONED_TABLE. + * the relation is RELKIND_PARTITIONED_TABLE. An equivalent problem + * exists with indexes. */ if (classform->relkind == RELKIND_PARTITIONED_TABLE) expected_relkind = RELKIND_RELATION; + else if (classform->relkind == RELKIND_PARTITIONED_INDEX) + expected_relkind = RELKIND_INDEX; else expected_relkind = classform->relkind; @@ -1210,7 +1273,8 @@ RangeVarCallbackForDropRelation(const RangeVar *rel, Oid relOid, Oid oldRelOid, * we do it the other way around. No error if we don't find a pg_index * entry, though --- the relation may have been dropped. */ - if (relkind == RELKIND_INDEX && relOid != oldRelOid) + if ((relkind == RELKIND_INDEX || relkind == RELKIND_PARTITIONED_INDEX) && + relOid != oldRelOid) { state->heapOid = IndexGetRelation(relOid, true); if (OidIsValid(state->heapOid)) @@ -2396,27 +2460,11 @@ StoreCatalogInheritance1(Oid relationId, Oid parentOid, int32 seqNumber, Relation inhRelation, bool child_is_partition) { - TupleDesc desc = RelationGetDescr(inhRelation); - Datum values[Natts_pg_inherits]; - bool nulls[Natts_pg_inherits]; ObjectAddress childobject, parentobject; - HeapTuple tuple; - - /* - * Make the pg_inherits entry - */ - values[Anum_pg_inherits_inhrelid - 1] = ObjectIdGetDatum(relationId); - values[Anum_pg_inherits_inhparent - 1] = ObjectIdGetDatum(parentOid); - values[Anum_pg_inherits_inhseqno - 1] = Int32GetDatum(seqNumber); - memset(nulls, 0, sizeof(nulls)); - - tuple = heap_form_tuple(desc, values, nulls); - - CatalogTupleInsert(inhRelation, tuple); - - heap_freetuple(tuple); + /* store the pg_inherits row */ + StoreSingleInheritance(relationId, parentOid, seqNumber); /* * Store a dependency too @@ -2540,6 +2588,7 @@ renameatt_check(Oid myrelid, Form_pg_class classform, bool recursing) relkind != RELKIND_MATVIEW && relkind != RELKIND_COMPOSITE_TYPE && relkind != RELKIND_INDEX && + relkind != RELKIND_PARTITIONED_INDEX && relkind != RELKIND_FOREIGN_TABLE && relkind != RELKIND_PARTITIONED_TABLE) ereport(ERROR, @@ -3019,7 +3068,8 @@ RenameRelationInternal(Oid myrelid, const char *newrelname, bool is_internal) /* * Also rename the associated constraint, if any. */ - if (targetrelation->rd_rel->relkind == RELKIND_INDEX) + if (targetrelation->rd_rel->relkind == RELKIND_INDEX || + targetrelation->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) { Oid constraintId = get_index_constraint(myrelid); @@ -3073,6 +3123,7 @@ CheckTableNotInUse(Relation rel, const char *stmt) stmt, RelationGetRelationName(rel)))); if (rel->rd_rel->relkind != RELKIND_INDEX && + rel->rd_rel->relkind != RELKIND_PARTITIONED_INDEX && AfterTriggerPendingOnRel(RelationGetRelid(rel))) ereport(ERROR, (errcode(ERRCODE_OBJECT_IN_USE), @@ -3764,6 +3815,10 @@ ATPrepCmd(List **wqueue, Relation rel, AlterTableCmd *cmd, pass = AT_PASS_MISC; break; case AT_AttachPartition: + ATSimplePermissions(rel, ATT_TABLE | ATT_PARTITIONED_INDEX); + /* No command-specific prep needed */ + pass = AT_PASS_MISC; + break; case AT_DetachPartition: ATSimplePermissions(rel, ATT_TABLE); /* No command-specific prep needed */ @@ -4112,9 +4167,15 @@ ATExecCmd(List **wqueue, AlteredTableInfo *tab, Relation rel, ATExecGenericOptions(rel, (List *) cmd->def); break; case AT_AttachPartition: - ATExecAttachPartition(wqueue, rel, (PartitionCmd *) cmd->def); + if (rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE) + ATExecAttachPartition(wqueue, rel, (PartitionCmd *) cmd->def); + else + ATExecAttachPartitionIdx(wqueue, rel, + ((PartitionCmd *) cmd->def)->name); break; case AT_DetachPartition: + /* ATPrepCmd ensures it must be a table */ + Assert(rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE); ATExecDetachPartition(rel, ((PartitionCmd *) cmd->def)->name); break; default: /* oops */ @@ -4148,9 +4209,13 @@ ATRewriteTables(AlterTableStmt *parsetree, List **wqueue, LOCKMODE lockmode) { AlteredTableInfo *tab = (AlteredTableInfo *) lfirst(ltab); - /* Foreign tables have no storage, nor do partitioned tables. */ + /* + * Foreign tables have no storage, nor do partitioned tables and + * indexes. + */ if (tab->relkind == RELKIND_FOREIGN_TABLE || - tab->relkind == RELKIND_PARTITIONED_TABLE) + tab->relkind == RELKIND_PARTITIONED_TABLE || + tab->relkind == RELKIND_PARTITIONED_INDEX) continue; /* @@ -4752,6 +4817,9 @@ ATSimplePermissions(Relation rel, int allowed_targets) case RELKIND_INDEX: actual_target = ATT_INDEX; break; + case RELKIND_PARTITIONED_INDEX: + actual_target = ATT_PARTITIONED_INDEX; + break; case RELKIND_COMPOSITE_TYPE: actual_target = ATT_COMPOSITE_TYPE; break; @@ -6194,6 +6262,7 @@ ATPrepSetStatistics(Relation rel, const char *colName, int16 colNum, Node *newVa if (rel->rd_rel->relkind != RELKIND_RELATION && rel->rd_rel->relkind != RELKIND_MATVIEW && rel->rd_rel->relkind != RELKIND_INDEX && + rel->rd_rel->relkind != RELKIND_PARTITIONED_INDEX && rel->rd_rel->relkind != RELKIND_FOREIGN_TABLE && rel->rd_rel->relkind != RELKIND_PARTITIONED_TABLE) ereport(ERROR, @@ -6205,7 +6274,9 @@ ATPrepSetStatistics(Relation rel, const char *colName, int16 colNum, Node *newVa * We allow referencing columns by numbers only for indexes, since table * column numbers could contain gaps if columns are later dropped. */ - if (rel->rd_rel->relkind != RELKIND_INDEX && !colName) + if (rel->rd_rel->relkind != RELKIND_INDEX && + rel->rd_rel->relkind != RELKIND_PARTITIONED_INDEX && + !colName) ereport(ERROR, (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), errmsg("cannot refer to non-index column by number"))); @@ -6283,7 +6354,8 @@ ATExecSetStatistics(Relation rel, const char *colName, int16 colNum, Node *newVa errmsg("cannot alter system column \"%s\"", colName))); - if (rel->rd_rel->relkind == RELKIND_INDEX && + if ((rel->rd_rel->relkind == RELKIND_INDEX || + rel->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) && rel->rd_index->indkey.values[attnum - 1] != 0) ereport(ERROR, (errcode(ERRCODE_FEATURE_NOT_SUPPORTED), @@ -6736,6 +6808,7 @@ ATExecAddIndex(AlteredTableInfo *tab, Relation rel, address = DefineIndex(RelationGetRelid(rel), stmt, InvalidOid, /* no predefined OID */ + InvalidOid, /* no parent index */ true, /* is_alter_table */ check_rights, false, /* check_not_in_use - we did it already */ @@ -9139,7 +9212,8 @@ ATExecAlterColumnType(AlteredTableInfo *tab, Relation rel, { char relKind = get_rel_relkind(foundObject.objectId); - if (relKind == RELKIND_INDEX) + if (relKind == RELKIND_INDEX || + relKind == RELKIND_PARTITIONED_INDEX) { Assert(foundObject.objectSubId == 0); if (!list_member_oid(tab->changedIndexOids, foundObject.objectId)) @@ -9982,6 +10056,15 @@ ATExecChangeOwner(Oid relationOid, Oid newOwnerId, bool recursing, LOCKMODE lock newOwnerId = tuple_class->relowner; } break; + case RELKIND_PARTITIONED_INDEX: + if (recursing) + break; + ereport(ERROR, + (errcode(ERRCODE_WRONG_OBJECT_TYPE), + errmsg("cannot change owner of index \"%s\"", + NameStr(tuple_class->relname)), + errhint("Change the ownership of the index's table, instead."))); + break; case RELKIND_SEQUENCE: if (!recursing && tuple_class->relowner != newOwnerId) @@ -10103,6 +10186,7 @@ ATExecChangeOwner(Oid relationOid, Oid newOwnerId, bool recursing, LOCKMODE lock */ if (tuple_class->relkind != RELKIND_COMPOSITE_TYPE && tuple_class->relkind != RELKIND_INDEX && + tuple_class->relkind != RELKIND_PARTITIONED_INDEX && tuple_class->relkind != RELKIND_TOASTVALUE) changeDependencyOnOwner(RelationRelationId, relationOid, newOwnerId); @@ -10110,7 +10194,8 @@ ATExecChangeOwner(Oid relationOid, Oid newOwnerId, bool recursing, LOCKMODE lock /* * Also change the ownership of the table's row type, if it has one */ - if (tuple_class->relkind != RELKIND_INDEX) + if (tuple_class->relkind != RELKIND_INDEX && + tuple_class->relkind != RELKIND_PARTITIONED_INDEX) AlterTypeOwnerInternal(tuple_class->reltype, newOwnerId); /* @@ -10119,6 +10204,7 @@ ATExecChangeOwner(Oid relationOid, Oid newOwnerId, bool recursing, LOCKMODE lock * relation, as well as its toast table (if it has one). */ if (tuple_class->relkind == RELKIND_RELATION || + tuple_class->relkind == RELKIND_PARTITIONED_TABLE || tuple_class->relkind == RELKIND_MATVIEW || tuple_class->relkind == RELKIND_TOASTVALUE) { @@ -10427,6 +10513,7 @@ ATExecSetRelOptions(Relation rel, List *defList, AlterTableType operation, (void) view_reloptions(newOptions, true); break; case RELKIND_INDEX: + case RELKIND_PARTITIONED_INDEX: (void) index_reloptions(rel->rd_amroutine->amoptions, newOptions, true); break; default: @@ -10839,7 +10926,8 @@ AlterTableMoveAll(AlterTableMoveAllStmt *stmt) relForm->relkind != RELKIND_RELATION && relForm->relkind != RELKIND_PARTITIONED_TABLE) || (stmt->objtype == OBJECT_INDEX && - relForm->relkind != RELKIND_INDEX) || + relForm->relkind != RELKIND_INDEX && + relForm->relkind != RELKIND_PARTITIONED_INDEX) || (stmt->objtype == OBJECT_MATVIEW && relForm->relkind != RELKIND_MATVIEW)) continue; @@ -11633,45 +11721,18 @@ RemoveInheritance(Relation child_rel, Relation parent_rel) Relation catalogRelation; SysScanDesc scan; ScanKeyData key[3]; - HeapTuple inheritsTuple, - attributeTuple, + HeapTuple attributeTuple, constraintTuple; List *connames; - bool found = false; + bool found; bool child_is_partition = false; /* If parent_rel is a partitioned table, child_rel must be a partition */ if (parent_rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE) child_is_partition = true; - /* - * Find and destroy the pg_inherits entry linking the two, or error out if - * there is none. - */ - catalogRelation = heap_open(InheritsRelationId, RowExclusiveLock); - ScanKeyInit(&key[0], - Anum_pg_inherits_inhrelid, - BTEqualStrategyNumber, F_OIDEQ, - ObjectIdGetDatum(RelationGetRelid(child_rel))); - scan = systable_beginscan(catalogRelation, InheritsRelidSeqnoIndexId, - true, NULL, 1, key); - - while (HeapTupleIsValid(inheritsTuple = systable_getnext(scan))) - { - Oid inhparent; - - inhparent = ((Form_pg_inherits) GETSTRUCT(inheritsTuple))->inhparent; - if (inhparent == RelationGetRelid(parent_rel)) - { - CatalogTupleDelete(catalogRelation, &inheritsTuple->t_self); - found = true; - break; - } - } - - systable_endscan(scan); - heap_close(catalogRelation, RowExclusiveLock); - + found = DeleteInheritsTuple(RelationGetRelid(child_rel), + RelationGetRelid(parent_rel)); if (!found) { if (child_is_partition) @@ -13226,7 +13287,8 @@ RangeVarCallbackForAlterRelation(const RangeVar *rv, Oid relid, Oid oldrelid, (errcode(ERRCODE_WRONG_OBJECT_TYPE), errmsg("\"%s\" is not a composite type", rv->relname))); - if (reltype == OBJECT_INDEX && relkind != RELKIND_INDEX + if (reltype == OBJECT_INDEX && relkind != RELKIND_INDEX && + relkind != RELKIND_PARTITIONED_INDEX && !IsA(stmt, RenameStmt)) ereport(ERROR, (errcode(ERRCODE_WRONG_OBJECT_TYPE), @@ -13946,6 +14008,9 @@ ATExecAttachPartition(List **wqueue, Relation rel, PartitionCmd *cmd) /* Update the pg_class entry. */ StorePartitionBound(attachrel, rel, cmd->bound); + /* Ensure there exists a correct set of indexes in the partition. */ + AttachPartitionEnsureIndexes(rel, attachrel); + /* * Generate partition constraint from the partition bound specification. * If the parent itself is a partition, make sure to include its @@ -14016,6 +14081,127 @@ ATExecAttachPartition(List **wqueue, Relation rel, PartitionCmd *cmd) } /* + * AttachPartitionEnsureIndexes + * subroutine for ATExecAttachPartition to create/match indexes + * + * Enforce the indexing rule for partitioned tables during ALTER TABLE / ATTACH + * PARTITION: every partition must have an index attached to each index on the + * partitioned table. + */ +static void +AttachPartitionEnsureIndexes(Relation rel, Relation attachrel) +{ + List *idxes; + List *attachRelIdxs; + Relation *attachrelIdxRels; + IndexInfo **attachInfos; + int i; + ListCell *cell; + MemoryContext cxt; + MemoryContext oldcxt; + + cxt = AllocSetContextCreate(CurrentMemoryContext, + "AttachPartitionEnsureIndexes", + ALLOCSET_DEFAULT_SIZES); + oldcxt = MemoryContextSwitchTo(cxt); + + idxes = RelationGetIndexList(rel); + attachRelIdxs = RelationGetIndexList(attachrel); + attachrelIdxRels = palloc(sizeof(Relation) * list_length(attachRelIdxs)); + attachInfos = palloc(sizeof(IndexInfo *) * list_length(attachRelIdxs)); + + /* Build arrays of all existing indexes and their IndexInfos */ + i = 0; + foreach(cell, attachRelIdxs) + { + Oid cldIdxId = lfirst_oid(cell); + + attachrelIdxRels[i] = index_open(cldIdxId, AccessShareLock); + attachInfos[i] = BuildIndexInfo(attachrelIdxRels[i]); + i++; + } + + /* + * For each index on the partitioned table, find a matching one in the + * partition-to-be; if one is not found, create one. + */ + foreach(cell, idxes) + { + Oid idx = lfirst_oid(cell); + Relation idxRel = index_open(idx, AccessShareLock); + IndexInfo *info; + AttrNumber *attmap; + bool found = false; + + /* + * Ignore indexes in the partitioned table other than partitioned + * indexes. + */ + if (idxRel->rd_rel->relkind != RELKIND_PARTITIONED_INDEX) + { + index_close(idxRel, AccessShareLock); + continue; + } + + /* construct an indexinfo to compare existing indexes against */ + info = BuildIndexInfo(idxRel); + attmap = convert_tuples_by_name_map(RelationGetDescr(attachrel), + RelationGetDescr(rel), + gettext_noop("could not convert row type")); + + /* + * Scan the list of existing indexes in the partition-to-be, and mark + * the first matching, unattached one we find, if any, as partition of + * the parent index. If we find one, we're done. + */ + for (i = 0; i < list_length(attachRelIdxs); i++) + { + /* does this index have a parent? if so, can't use it */ + if (has_superclass(RelationGetRelid(attachrelIdxRels[i]))) + continue; + + if (CompareIndexInfo(attachInfos[i], info, + attachrelIdxRels[i]->rd_indcollation, + idxRel->rd_indcollation, + attachrelIdxRels[i]->rd_opfamily, + idxRel->rd_opfamily, + attmap, + RelationGetDescr(rel)->natts)) + { + /* bingo. */ + IndexSetParentIndex(attachrelIdxRels[i], idx); + found = true; + break; + } + } + + /* + * If no suitable index was found in the partition-to-be, create one + * now. + */ + if (!found) + { + IndexStmt *stmt; + + stmt = generateClonedIndexStmt(NULL, RelationGetRelid(attachrel), + idxRel, attmap, + RelationGetDescr(rel)->natts); + DefineIndex(RelationGetRelid(attachrel), stmt, InvalidOid, + RelationGetRelid(idxRel), + false, false, false, false, false); + } + + index_close(idxRel, AccessShareLock); + } + + /* Clean up. */ + for (i = 0; i < list_length(attachRelIdxs); i++) + index_close(attachrelIdxRels[i], AccessShareLock); + MemoryContextSwitchTo(oldcxt); + MemoryContextDelete(cxt); +} + +/* * ALTER TABLE DETACH PARTITION * * Return the address of the relation that is no longer a partition of rel. @@ -14033,6 +14219,8 @@ ATExecDetachPartition(Relation rel, RangeVar *name) new_repl[Natts_pg_class]; ObjectAddress address; Oid defaultPartOid; + List *indexes; + ListCell *cell; /* * We must lock the default partition, because detaching this partition @@ -14094,6 +14282,24 @@ ATExecDetachPartition(Relation rel, RangeVar *name) } } + /* detach indexes too */ + indexes = RelationGetIndexList(partRel); + foreach(cell, indexes) + { + Oid idxid = lfirst_oid(cell); + Relation idx; + + if (!has_superclass(idxid)) + continue; + + Assert((IndexGetRelation(get_partition_parent(idxid), false) == + RelationGetRelid(rel))); + + idx = index_open(idxid, AccessExclusiveLock); + IndexSetParentIndex(idx, InvalidOid); + relation_close(idx, AccessExclusiveLock); + } + /* * Invalidate the parent's relcache so that the partition is no longer * included in its partition descriptor. @@ -14107,3 +14313,328 @@ ATExecDetachPartition(Relation rel, RangeVar *name) return address; } + +/* + * Before acquiring lock on an index, acquire the same lock on the owning + * table. + */ +struct AttachIndexCallbackState +{ + Oid partitionOid; + Oid parentTblOid; + bool lockedParentTbl; +}; + +static void +RangeVarCallbackForAttachIndex(const RangeVar *rv, Oid relOid, Oid oldRelOid, + void *arg) +{ + struct AttachIndexCallbackState *state; + Form_pg_class classform; + HeapTuple tuple; + + state = (struct AttachIndexCallbackState *) arg; + + if (!state->lockedParentTbl) + { + LockRelationOid(state->parentTblOid, AccessShareLock); + state->lockedParentTbl = true; + } + + /* + * If we previously locked some other heap, and the name we're looking up + * no longer refers to an index on that relation, release the now-useless + * lock. XXX maybe we should do *after* we verify whether the index does + * not actually belong to the same relation ... + */ + if (relOid != oldRelOid && OidIsValid(state->partitionOid)) + { + UnlockRelationOid(state->partitionOid, AccessShareLock); + state->partitionOid = InvalidOid; + } + + /* Didn't find a relation, so no need for locking or permission checks. */ + if (!OidIsValid(relOid)) + return; + + tuple = SearchSysCache1(RELOID, ObjectIdGetDatum(relOid)); + if (!HeapTupleIsValid(tuple)) + return; /* concurrently dropped, so nothing to do */ + classform = (Form_pg_class) GETSTRUCT(tuple); + if (classform->relkind != RELKIND_PARTITIONED_INDEX && + classform->relkind != RELKIND_INDEX) + ereport(ERROR, + (errcode(ERRCODE_INVALID_OBJECT_DEFINITION), + errmsg("\"%s\" is not an index", rv->relname))); + ReleaseSysCache(tuple); + + /* + * Since we need only examine the heap's tupledesc, an access share lock + * on it (preventing any DDL) is sufficient. + */ + state->partitionOid = IndexGetRelation(relOid, false); + LockRelationOid(state->partitionOid, AccessShareLock); +} + +/* + * ALTER INDEX i1 ATTACH PARTITION i2 + */ +static ObjectAddress +ATExecAttachPartitionIdx(List **wqueue, Relation parentIdx, RangeVar *name) +{ + Relation partIdx; + Relation partTbl; + Relation parentTbl; + ObjectAddress address; + Oid partIdxId; + Oid currParent; + struct AttachIndexCallbackState state; + + /* + * We need to obtain lock on the index 'name' to modify it, but we also + * need to read its owning table's tuple descriptor -- so we need to lock + * both. To avoid deadlocks, obtain lock on the table before doing so on + * the index. Furthermore, we need to examine the parent table of the + * partition, so lock that one too. + */ + state.partitionOid = InvalidOid; + state.parentTblOid = parentIdx->rd_index->indrelid; + state.lockedParentTbl = false; + partIdxId = + RangeVarGetRelidExtended(name, AccessExclusiveLock, false, false, + RangeVarCallbackForAttachIndex, + (void *) &state); + /* Not there? */ + if (!OidIsValid(partIdxId)) + ereport(ERROR, + (errcode(ERRCODE_UNDEFINED_OBJECT), + errmsg("index \"%s\" does not exist", name->relname))); + + /* no deadlock risk: RangeVarGetRelidExtended already acquired the lock */ + partIdx = relation_open(partIdxId, AccessExclusiveLock); + + /* we already hold locks on both tables, so this is safe: */ + parentTbl = relation_open(parentIdx->rd_index->indrelid, AccessShareLock); + partTbl = relation_open(partIdx->rd_index->indrelid, NoLock); + + ObjectAddressSet(address, RelationRelationId, RelationGetRelid(partIdx)); + + /* Silently do nothing if already in the right state */ + currParent = !has_superclass(partIdxId) ? InvalidOid : + get_partition_parent(partIdxId); + if (currParent != RelationGetRelid(parentIdx)) + { + IndexInfo *childInfo; + IndexInfo *parentInfo; + AttrNumber *attmap; + bool found; + int i; + PartitionDesc partDesc; + + /* + * If this partition already has an index attached, refuse the operation. + */ + refuseDupeIndexAttach(parentIdx, partIdx, partTbl); + + if (OidIsValid(currParent)) + ereport(ERROR, + (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE), + errmsg("cannot attach index \"%s\" as a partition of index \"%s\"", + RelationGetRelationName(partIdx), + RelationGetRelationName(parentIdx)), + errdetail("Index \"%s\" is already attached to another index.", + RelationGetRelationName(partIdx)))); + + /* Make sure it indexes a partition of the other index's table */ + partDesc = RelationGetPartitionDesc(parentTbl); + found = false; + for (i = 0; i < partDesc->nparts; i++) + { + if (partDesc->oids[i] == state.partitionOid) + { + found = true; + break; + } + } + if (!found) + ereport(ERROR, + (errmsg("cannot attach index \"%s\" as a partition of index \"%s\"", + RelationGetRelationName(partIdx), + RelationGetRelationName(parentIdx)), + errdetail("Index \"%s\" is not an index on any partition of table \"%s\".", + RelationGetRelationName(partIdx), + RelationGetRelationName(parentTbl)))); + + /* Ensure the indexes are compatible */ + childInfo = BuildIndexInfo(partIdx); + parentInfo = BuildIndexInfo(parentIdx); + attmap = convert_tuples_by_name_map(RelationGetDescr(partTbl), + RelationGetDescr(parentTbl), + gettext_noop("could not convert row type")); + if (!CompareIndexInfo(childInfo, parentInfo, + partIdx->rd_indcollation, + parentIdx->rd_indcollation, + partIdx->rd_opfamily, + parentIdx->rd_opfamily, + attmap, + RelationGetDescr(partTbl)->natts)) + ereport(ERROR, + (errcode(ERRCODE_INVALID_OBJECT_DEFINITION), + errmsg("cannot attach index \"%s\" as a partition of index \"%s\"", + RelationGetRelationName(partIdx), + RelationGetRelationName(parentIdx)), + errdetail("The index definitions do not match."))); + + /* All good -- do it */ + IndexSetParentIndex(partIdx, RelationGetRelid(parentIdx)); + pfree(attmap); + + CommandCounterIncrement(); + + validatePartitionedIndex(parentIdx, parentTbl); + } + + relation_close(parentTbl, AccessShareLock); + /* keep these locks till commit */ + relation_close(partTbl, NoLock); + relation_close(partIdx, NoLock); + + return address; +} + +/* + * Verify whether the given partition already contains an index attached + * to the given partitioned index. If so, raise an error. + */ +static void +refuseDupeIndexAttach(Relation parentIdx, Relation partIdx, Relation partitionTbl) +{ + Relation pg_inherits; + ScanKeyData key; + HeapTuple tuple; + SysScanDesc scan; + + pg_inherits = heap_open(InheritsRelationId, AccessShareLock); + ScanKeyInit(&key, Anum_pg_inherits_inhparent, + BTEqualStrategyNumber, F_OIDEQ, + ObjectIdGetDatum(RelationGetRelid(parentIdx))); + scan = systable_beginscan(pg_inherits, InheritsParentIndexId, true, + NULL, 1, &key); + while (HeapTupleIsValid(tuple = systable_getnext(scan))) + { + Form_pg_inherits inhForm; + Oid tab; + + inhForm = (Form_pg_inherits) GETSTRUCT(tuple); + tab = IndexGetRelation(inhForm->inhrelid, false); + if (tab == RelationGetRelid(partitionTbl)) + ereport(ERROR, + (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE), + errmsg("cannot attach index \"%s\" as a partition of index \"%s\"", + RelationGetRelationName(partIdx), + RelationGetRelationName(parentIdx)), + errdetail("Another index is already attached for partition \"%s\".", + RelationGetRelationName(partitionTbl)))); + } + + systable_endscan(scan); + heap_close(pg_inherits, AccessShareLock); +} + +/* + * Verify whether the set of attached partition indexes to a parent index on + * a partitioned table is complete. If it is, mark the parent index valid. + * + * This should be called each time a partition index is attached. + */ +static void +validatePartitionedIndex(Relation partedIdx, Relation partedTbl) +{ + Relation inheritsRel; + SysScanDesc scan; + ScanKeyData key; + int tuples = 0; + HeapTuple inhTup; + bool updated = false; + + Assert(partedIdx->rd_rel->relkind == RELKIND_PARTITIONED_INDEX); + + /* + * Scan pg_inherits for this parent index. Count each valid index we find + * (verifying the pg_index entry for each), and if we reach the total + * amount we expect, we can mark this parent index as valid. + */ + inheritsRel = heap_open(InheritsRelationId, AccessShareLock); + ScanKeyInit(&key, Anum_pg_inherits_inhparent, + BTEqualStrategyNumber, F_OIDEQ, + ObjectIdGetDatum(RelationGetRelid(partedIdx))); + scan = systable_beginscan(inheritsRel, InheritsParentIndexId, true, + NULL, 1, &key); + while ((inhTup = systable_getnext(scan)) != NULL) + { + Form_pg_inherits inhForm = (Form_pg_inherits) GETSTRUCT(inhTup); + HeapTuple indTup; + Form_pg_index indexForm; + + indTup = SearchSysCache1(INDEXRELID, + ObjectIdGetDatum(inhForm->inhrelid)); + if (!indTup) + elog(ERROR, "cache lookup failed for index %u", + inhForm->inhrelid); + indexForm = (Form_pg_index) GETSTRUCT(indTup); + if (IndexIsValid(indexForm)) + tuples += 1; + ReleaseSysCache(indTup); + } + + /* Done with pg_inherits */ + systable_endscan(scan); + heap_close(inheritsRel, AccessShareLock); + + /* + * If we found as many inherited indexes as the partitioned table has + * partitions, we're good; update pg_index to set indisvalid. + */ + if (tuples == RelationGetPartitionDesc(partedTbl)->nparts) + { + Relation idxRel; + HeapTuple newtup; + + idxRel = heap_open(IndexRelationId, RowExclusiveLock); + + newtup = heap_copytuple(partedIdx->rd_indextuple); + ((Form_pg_index) GETSTRUCT(newtup))->indisvalid = true; + updated = true; + + CatalogTupleUpdate(idxRel, &partedIdx->rd_indextuple->t_self, newtup); + + heap_close(idxRel, RowExclusiveLock); + } + + /* + * If this index is in turn a partition of a larger index, validating it + * might cause the parent to become valid also. Try that. + */ + if (updated && + has_superclass(RelationGetRelid(partedIdx))) + { + Oid parentIdxId, + parentTblId; + Relation parentIdx, + parentTbl; + + /* make sure we see the validation we just did */ + CommandCounterIncrement(); + + parentIdxId = get_partition_parent(RelationGetRelid(partedIdx)); + parentTblId = get_partition_parent(RelationGetRelid(partedTbl)); + parentIdx = relation_open(parentIdxId, AccessExclusiveLock); + parentTbl = relation_open(parentTblId, AccessExclusiveLock); + Assert(!parentIdx->rd_index->indisvalid); + + validatePartitionedIndex(parentIdx, parentTbl); + + relation_close(parentIdx, AccessExclusiveLock); + relation_close(parentTbl, AccessExclusiveLock); + } +} diff --git a/src/backend/nodes/copyfuncs.c b/src/backend/nodes/copyfuncs.c index ddbbc79823e..65d8c77d7aa 100644 --- a/src/backend/nodes/copyfuncs.c +++ b/src/backend/nodes/copyfuncs.c @@ -3379,6 +3379,7 @@ _copyIndexStmt(const IndexStmt *from) COPY_STRING_FIELD(idxname); COPY_NODE_FIELD(relation); + COPY_SCALAR_FIELD(relationId); COPY_STRING_FIELD(accessMethod); COPY_STRING_FIELD(tableSpace); COPY_NODE_FIELD(indexParams); diff --git a/src/backend/nodes/equalfuncs.c b/src/backend/nodes/equalfuncs.c index 30ccc9c5ae7..0bd12e862e5 100644 --- a/src/backend/nodes/equalfuncs.c +++ b/src/backend/nodes/equalfuncs.c @@ -1332,6 +1332,7 @@ _equalIndexStmt(const IndexStmt *a, const IndexStmt *b) { COMPARE_STRING_FIELD(idxname); COMPARE_NODE_FIELD(relation); + COMPARE_SCALAR_FIELD(relationId); COMPARE_STRING_FIELD(accessMethod); COMPARE_STRING_FIELD(tableSpace); COMPARE_NODE_FIELD(indexParams); diff --git a/src/backend/nodes/outfuncs.c b/src/backend/nodes/outfuncs.c index 5e72df137e7..b1cdfc36a62 100644 --- a/src/backend/nodes/outfuncs.c +++ b/src/backend/nodes/outfuncs.c @@ -2650,6 +2650,7 @@ _outIndexStmt(StringInfo str, const IndexStmt *node) WRITE_STRING_FIELD(idxname); WRITE_NODE_FIELD(relation); + WRITE_OID_FIELD(relationId); WRITE_STRING_FIELD(accessMethod); WRITE_STRING_FIELD(tableSpace); WRITE_NODE_FIELD(indexParams); diff --git a/src/backend/parser/gram.y b/src/backend/parser/gram.y index e42b7caff6b..93e67e8adcc 100644 --- a/src/backend/parser/gram.y +++ b/src/backend/parser/gram.y @@ -290,7 +290,7 @@ static Node *makeRecursiveViewSelect(char *relname, List *aliases, Node *query); %type <ival> add_drop opt_asc_desc opt_nulls_order %type <node> alter_table_cmd alter_type_cmd opt_collate_clause - replica_identity partition_cmd + replica_identity partition_cmd index_partition_cmd %type <list> alter_table_cmds alter_type_cmds %type <list> alter_identity_column_option_list %type <defelt> alter_identity_column_option @@ -1891,6 +1891,15 @@ AlterTableStmt: n->missing_ok = true; $$ = (Node *)n; } + | ALTER INDEX qualified_name index_partition_cmd + { + AlterTableStmt *n = makeNode(AlterTableStmt); + n->relation = $3; + n->cmds = list_make1($4); + n->relkind = OBJECT_INDEX; + n->missing_ok = false; + $$ = (Node *)n; + } | ALTER INDEX ALL IN_P TABLESPACE name SET TABLESPACE name opt_nowait { AlterTableMoveAllStmt *n = @@ -2025,6 +2034,22 @@ partition_cmd: } ; +index_partition_cmd: + /* ALTER INDEX <name> ATTACH PARTITION <index_name> */ + ATTACH PARTITION qualified_name + { + AlterTableCmd *n = makeNode(AlterTableCmd); + PartitionCmd *cmd = makeNode(PartitionCmd); + + n->subtype = AT_AttachPartition; + cmd->name = $3; + cmd->bound = NULL; + n->def = (Node *) cmd; + + $$ = (Node *) n; + } + ; + alter_table_cmd: /* ALTER TABLE <name> ADD <coldef> */ ADD_P columnDef @@ -7330,7 +7355,7 @@ defacl_privilege_target: *****************************************************************************/ IndexStmt: CREATE opt_unique INDEX opt_concurrently opt_index_name - ON qualified_name access_method_clause '(' index_params ')' + ON relation_expr access_method_clause '(' index_params ')' opt_reloptions OptTableSpace where_clause { IndexStmt *n = makeNode(IndexStmt); @@ -7338,6 +7363,7 @@ IndexStmt: CREATE opt_unique INDEX opt_concurrently opt_index_name n->concurrent = $4; n->idxname = $5; n->relation = $7; + n->relationId = InvalidOid; n->accessMethod = $8; n->indexParams = $10; n->options = $12; @@ -7356,7 +7382,7 @@ IndexStmt: CREATE opt_unique INDEX opt_concurrently opt_index_name $$ = (Node *)n; } | CREATE opt_unique INDEX opt_concurrently IF_P NOT EXISTS index_name - ON qualified_name access_method_clause '(' index_params ')' + ON relation_expr access_method_clause '(' index_params ')' opt_reloptions OptTableSpace where_clause { IndexStmt *n = makeNode(IndexStmt); @@ -7364,6 +7390,7 @@ IndexStmt: CREATE opt_unique INDEX opt_concurrently opt_index_name n->concurrent = $4; n->idxname = $8; n->relation = $10; + n->relationId = InvalidOid; n->accessMethod = $11; n->indexParams = $13; n->options = $15; diff --git a/src/backend/parser/parse_utilcmd.c b/src/backend/parser/parse_utilcmd.c index 128f1679c6b..90bb356df85 100644 --- a/src/backend/parser/parse_utilcmd.c +++ b/src/backend/parser/parse_utilcmd.c @@ -118,9 +118,6 @@ static void transformTableLikeClause(CreateStmtContext *cxt, TableLikeClause *table_like_clause); static void transformOfType(CreateStmtContext *cxt, TypeName *ofTypename); -static IndexStmt *generateClonedIndexStmt(CreateStmtContext *cxt, - Relation source_idx, - const AttrNumber *attmap, int attmap_length); static List *get_collation(Oid collation, Oid actual_datatype); static List *get_opclass(Oid opclass, Oid actual_datatype); static void transformIndexConstraints(CreateStmtContext *cxt); @@ -1185,7 +1182,8 @@ transformTableLikeClause(CreateStmtContext *cxt, TableLikeClause *table_like_cla parent_index = index_open(parent_index_oid, AccessShareLock); /* Build CREATE INDEX statement to recreate the parent_index */ - index_stmt = generateClonedIndexStmt(cxt, parent_index, + index_stmt = generateClonedIndexStmt(cxt->relation, InvalidOid, + parent_index, attmap, tupleDesc->natts); /* Copy comment on index, if requested */ @@ -1263,10 +1261,12 @@ transformOfType(CreateStmtContext *cxt, TypeName *ofTypename) /* * Generate an IndexStmt node using information from an already existing index - * "source_idx". Attribute numbers should be adjusted according to attmap. + * "source_idx", for the rel identified either by heapRel or heapRelid. + * + * Attribute numbers should be adjusted according to attmap. */ -static IndexStmt * -generateClonedIndexStmt(CreateStmtContext *cxt, Relation source_idx, +IndexStmt * +generateClonedIndexStmt(RangeVar *heapRel, Oid heapRelid, Relation source_idx, const AttrNumber *attmap, int attmap_length) { Oid source_relid = RelationGetRelid(source_idx); @@ -1287,6 +1287,9 @@ generateClonedIndexStmt(CreateStmtContext *cxt, Relation source_idx, Datum datum; bool isnull; + Assert((heapRel == NULL && OidIsValid(heapRelid)) || + (heapRel != NULL && !OidIsValid(heapRelid))); + /* * Fetch pg_class tuple of source index. We can't use the copy in the * relcache entry because it doesn't include optional fields. @@ -1322,7 +1325,8 @@ generateClonedIndexStmt(CreateStmtContext *cxt, Relation source_idx, /* Begin building the IndexStmt */ index = makeNode(IndexStmt); - index->relation = cxt->relation; + index->relation = heapRel; + index->relationId = heapRelid; index->accessMethod = pstrdup(NameStr(amrec->amname)); if (OidIsValid(idxrelrec->reltablespace)) index->tableSpace = get_tablespace_name(idxrelrec->reltablespace); @@ -3289,18 +3293,39 @@ transformPartitionCmd(CreateStmtContext *cxt, PartitionCmd *cmd) { Relation parentRel = cxt->rel; - /* the table must be partitioned */ - if (parentRel->rd_rel->relkind != RELKIND_PARTITIONED_TABLE) - ereport(ERROR, - (errcode(ERRCODE_INVALID_OBJECT_DEFINITION), - errmsg("\"%s\" is not partitioned", - RelationGetRelationName(parentRel)))); - - /* transform the partition bound, if any */ - Assert(RelationGetPartitionKey(parentRel) != NULL); - if (cmd->bound != NULL) - cxt->partbound = transformPartitionBound(cxt->pstate, parentRel, - cmd->bound); + switch (parentRel->rd_rel->relkind) + { + case RELKIND_PARTITIONED_TABLE: + /* transform the partition bound, if any */ + Assert(RelationGetPartitionKey(parentRel) != NULL); + if (cmd->bound != NULL) + cxt->partbound = transformPartitionBound(cxt->pstate, parentRel, + cmd->bound); + break; + case RELKIND_PARTITIONED_INDEX: + /* nothing to check */ + Assert(cmd->bound == NULL); + break; + case RELKIND_RELATION: + /* the table must be partitioned */ + ereport(ERROR, + (errcode(ERRCODE_INVALID_OBJECT_DEFINITION), + errmsg("table \"%s\" is not partitioned", + RelationGetRelationName(parentRel)))); + break; + case RELKIND_INDEX: + /* the index must be partitioned */ + ereport(ERROR, + (errcode(ERRCODE_INVALID_OBJECT_DEFINITION), + errmsg("index \"%s\" is not partitioned", + RelationGetRelationName(parentRel)))); + break; + default: + /* parser shouldn't let this case through */ + elog(ERROR, "\"%s\" is not a partitioned table or index", + RelationGetRelationName(parentRel)); + break; + } } /* diff --git a/src/backend/tcop/utility.c b/src/backend/tcop/utility.c index ec98a612ec9..9cccc8d39de 100644 --- a/src/backend/tcop/utility.c +++ b/src/backend/tcop/utility.c @@ -23,6 +23,7 @@ #include "access/xlog.h" #include "catalog/catalog.h" #include "catalog/namespace.h" +#include "catalog/pg_inherits_fn.h" #include "catalog/toasting.h" #include "commands/alter.h" #include "commands/async.h" @@ -1300,6 +1301,7 @@ ProcessUtilitySlow(ParseState *pstate, IndexStmt *stmt = (IndexStmt *) parsetree; Oid relid; LOCKMODE lockmode; + List *inheritors = NIL; if (stmt->concurrent) PreventTransactionChain(isTopLevel, @@ -1322,6 +1324,23 @@ ProcessUtilitySlow(ParseState *pstate, RangeVarCallbackOwnsRelation, NULL); + /* + * CREATE INDEX on partitioned tables (but not regular + * inherited tables) recurses to partitions, so we must + * acquire locks early to avoid deadlocks. + */ + if (stmt->relation->inh) + { + Relation rel; + + /* already locked by RangeVarGetRelidExtended */ + rel = heap_open(relid, NoLock); + if (rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE) + inheritors = find_all_inheritors(relid, lockmode, + NULL); + heap_close(rel, NoLock); + } + /* Run parse analysis ... */ stmt = transformIndexStmt(relid, stmt, queryString); @@ -1331,6 +1350,7 @@ ProcessUtilitySlow(ParseState *pstate, DefineIndex(relid, /* OID of heap relation */ stmt, InvalidOid, /* no predefined OID */ + InvalidOid, /* no parent index */ false, /* is_alter_table */ true, /* check_rights */ true, /* check_not_in_use */ @@ -1346,6 +1366,8 @@ ProcessUtilitySlow(ParseState *pstate, parsetree); commandCollected = true; EventTriggerAlterTableEnd(); + + list_free(inheritors); } break; diff --git a/src/backend/utils/adt/amutils.c b/src/backend/utils/adt/amutils.c index a6d8feea5b4..0f7ceb62eb5 100644 --- a/src/backend/utils/adt/amutils.c +++ b/src/backend/utils/adt/amutils.c @@ -183,7 +183,8 @@ indexam_property(FunctionCallInfo fcinfo, if (!HeapTupleIsValid(tuple)) PG_RETURN_NULL(); rd_rel = (Form_pg_class) GETSTRUCT(tuple); - if (rd_rel->relkind != RELKIND_INDEX) + if (rd_rel->relkind != RELKIND_INDEX && + rd_rel->relkind != RELKIND_PARTITIONED_INDEX) { ReleaseSysCache(tuple); PG_RETURN_NULL(); diff --git a/src/backend/utils/adt/ruleutils.c b/src/backend/utils/adt/ruleutils.c index 9cdbb06add7..c5f5a1ca3f9 100644 --- a/src/backend/utils/adt/ruleutils.c +++ b/src/backend/utils/adt/ruleutils.c @@ -317,7 +317,7 @@ static void decompile_column_index_array(Datum column_index_array, Oid relId, static char *pg_get_ruledef_worker(Oid ruleoid, int prettyFlags); static char *pg_get_indexdef_worker(Oid indexrelid, int colno, const Oid *excludeOps, - bool attrsOnly, bool showTblSpc, + bool attrsOnly, bool showTblSpc, bool inherits, int prettyFlags, bool missing_ok); static char *pg_get_statisticsobj_worker(Oid statextid, bool missing_ok); static char *pg_get_partkeydef_worker(Oid relid, int prettyFlags, @@ -1086,7 +1086,7 @@ pg_get_indexdef(PG_FUNCTION_ARGS) prettyFlags = PRETTYFLAG_INDENT; - res = pg_get_indexdef_worker(indexrelid, 0, NULL, false, false, + res = pg_get_indexdef_worker(indexrelid, 0, NULL, false, false, false, prettyFlags, true); if (res == NULL) @@ -1107,7 +1107,7 @@ pg_get_indexdef_ext(PG_FUNCTION_ARGS) prettyFlags = pretty ? PRETTYFLAG_PAREN | PRETTYFLAG_INDENT : PRETTYFLAG_INDENT; res = pg_get_indexdef_worker(indexrelid, colno, NULL, colno != 0, false, - prettyFlags, true); + false, prettyFlags, true); if (res == NULL) PG_RETURN_NULL(); @@ -1123,7 +1123,7 @@ pg_get_indexdef_ext(PG_FUNCTION_ARGS) char * pg_get_indexdef_string(Oid indexrelid) { - return pg_get_indexdef_worker(indexrelid, 0, NULL, false, true, 0, false); + return pg_get_indexdef_worker(indexrelid, 0, NULL, false, true, true, 0, false); } /* Internal version that just reports the column definitions */ @@ -1133,7 +1133,7 @@ pg_get_indexdef_columns(Oid indexrelid, bool pretty) int prettyFlags; prettyFlags = pretty ? PRETTYFLAG_PAREN | PRETTYFLAG_INDENT : PRETTYFLAG_INDENT; - return pg_get_indexdef_worker(indexrelid, 0, NULL, true, false, + return pg_get_indexdef_worker(indexrelid, 0, NULL, true, false, false, prettyFlags, false); } @@ -1146,7 +1146,7 @@ pg_get_indexdef_columns(Oid indexrelid, bool pretty) static char * pg_get_indexdef_worker(Oid indexrelid, int colno, const Oid *excludeOps, - bool attrsOnly, bool showTblSpc, + bool attrsOnly, bool showTblSpc, bool inherits, int prettyFlags, bool missing_ok) { /* might want a separate isConstraint parameter later */ @@ -1259,9 +1259,11 @@ pg_get_indexdef_worker(Oid indexrelid, int colno, if (!attrsOnly) { if (!isConstraint) - appendStringInfo(&buf, "CREATE %sINDEX %s ON %s USING %s (", + appendStringInfo(&buf, "CREATE %sINDEX %s ON %s%s USING %s (", idxrec->indisunique ? "UNIQUE " : "", quote_identifier(NameStr(idxrelrec->relname)), + idxrelrec->relkind == RELKIND_PARTITIONED_INDEX + && !inherits ? "ONLY " : "", generate_relation_name(indrelid, NIL), quote_identifier(NameStr(amrec->amname))); else /* currently, must be EXCLUDE constraint */ @@ -2148,6 +2150,7 @@ pg_get_constraintdef_worker(Oid constraintId, bool fullCommand, operators, false, false, + false, prettyFlags, false)); break; diff --git a/src/backend/utils/cache/relcache.c b/src/backend/utils/cache/relcache.c index 00ba33bfb44..c081b88b733 100644 --- a/src/backend/utils/cache/relcache.c +++ b/src/backend/utils/cache/relcache.c @@ -430,18 +430,26 @@ static void RelationParseRelOptions(Relation relation, HeapTuple tuple) { bytea *options; + amoptions_function amoptsfn; relation->rd_options = NULL; - /* Fall out if relkind should not have options */ + /* + * Look up any AM-specific parse function; fall out if relkind should not + * have options. + */ switch (relation->rd_rel->relkind) { case RELKIND_RELATION: case RELKIND_TOASTVALUE: - case RELKIND_INDEX: case RELKIND_VIEW: case RELKIND_MATVIEW: case RELKIND_PARTITIONED_TABLE: + amoptsfn = NULL; + break; + case RELKIND_INDEX: + case RELKIND_PARTITIONED_INDEX: + amoptsfn = relation->rd_amroutine->amoptions; break; default: return; @@ -452,10 +460,7 @@ RelationParseRelOptions(Relation relation, HeapTuple tuple) * we might not have any other for pg_class yet (consider executing this * code for pg_class itself) */ - options = extractRelOptions(tuple, - GetPgClassDescriptor(), - relation->rd_rel->relkind == RELKIND_INDEX ? - relation->rd_amroutine->amoptions : NULL); + options = extractRelOptions(tuple, GetPgClassDescriptor(), amoptsfn); /* * Copy parsed data into CacheMemoryContext. To guard against the @@ -2053,7 +2058,8 @@ RelationIdGetRelation(Oid relationId) * and we don't want to use the full-blown procedure because it's * a headache for indexes that reload itself depends on. */ - if (rd->rd_rel->relkind == RELKIND_INDEX) + if (rd->rd_rel->relkind == RELKIND_INDEX || + rd->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) RelationReloadIndexInfo(rd); else RelationClearRelation(rd, true); @@ -2167,7 +2173,8 @@ RelationReloadIndexInfo(Relation relation) Form_pg_class relp; /* Should be called only for invalidated indexes */ - Assert(relation->rd_rel->relkind == RELKIND_INDEX && + Assert((relation->rd_rel->relkind == RELKIND_INDEX || + relation->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) && !relation->rd_isvalid); /* Ensure it's closed at smgr level */ @@ -2387,7 +2394,8 @@ RelationClearRelation(Relation relation, bool rebuild) { RelationInitPhysicalAddr(relation); - if (relation->rd_rel->relkind == RELKIND_INDEX) + if (relation->rd_rel->relkind == RELKIND_INDEX || + relation->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) { relation->rd_isvalid = false; /* needs to be revalidated */ if (relation->rd_refcnt > 1 && IsTransactionState()) @@ -2403,7 +2411,8 @@ RelationClearRelation(Relation relation, bool rebuild) * re-read the pg_class row to handle possible physical relocation of the * index, and we check for pg_index updates too. */ - if (relation->rd_rel->relkind == RELKIND_INDEX && + if ((relation->rd_rel->relkind == RELKIND_INDEX || + relation->rd_rel->relkind == RELKIND_PARTITIONED_INDEX) && relation->rd_refcnt > 0 && relation->rd_indexcxt != NULL) { @@ -5461,7 +5470,10 @@ load_relcache_init_file(bool shared) rel->rd_att->constr = constr; } - /* If it's an index, there's more to do */ + /* + * If it's an index, there's more to do. Note we explicitly ignore + * partitioned indexes here. + */ if (rel->rd_rel->relkind == RELKIND_INDEX) { MemoryContext indexcxt; @@ -5825,7 +5837,10 @@ write_relcache_init_file(bool shared) (rel->rd_options ? VARSIZE(rel->rd_options) : 0), fp); - /* If it's an index, there's more to do */ + /* + * If it's an index, there's more to do. Note we explicitly ignore + * partitioned indexes here. + */ if (rel->rd_rel->relkind == RELKIND_INDEX) { /* write the pg_index tuple */ |