Fix assorted bugs in CREATE/DROP INDEX CONCURRENTLY.

Commit 8cb53654dbdb4c386369eb988062d0bbb6de725e, which introduced DROP
INDEX CONCURRENTLY, managed to break CREATE INDEX CONCURRENTLY via a poor
choice of catalog state representation.  The pg_index state for an index
that's reached the final pre-drop stage was the same as the state for an
index just created by CREATE INDEX CONCURRENTLY.  This meant that the
(necessary) change to make RelationGetIndexList ignore about-to-die indexes
also made it ignore freshly-created indexes; which is catastrophic because
the latter do need to be considered in HOT-safety decisions.  Failure to
do so leads to incorrect index entries and subsequently wrong results from
queries depending on the concurrently-created index.

To fix, make the final state be indisvalid = true and indisready = false,
which is otherwise nonsensical.  This is pretty ugly but we can't add
another column without forcing initdb, and it's too late for that in 9.2.
(There's a cleaner fix in HEAD.)

In addition, change CREATE/DROP INDEX CONCURRENTLY so that the pg_index
flag changes they make without exclusive lock on the index are made via
heap_inplace_update() rather than a normal transactional update.  The
latter is not very safe because moving the pg_index tuple could result in
concurrent SnapshotNow scans finding it twice or not at all, thus possibly
resulting in index corruption.  This is a pre-existing bug in CREATE INDEX
CONCURRENTLY, which was copied into the DROP code.

In addition, fix various places in the code that ought to check to make
sure that the indexes they are manipulating are valid and/or ready as
appropriate.  These represent bugs that have existed since 8.2, since
a failed CREATE INDEX CONCURRENTLY could leave a corrupt or invalid
index behind, and we ought not try to do anything that might fail with
such an index.

Also fix RelationReloadIndexInfo to ensure it copies all the pg_index
columns that are allowed to change after initial creation.  Previously we
could have been left with stale values of some fields in an index relcache
entry.  It's not clear whether this actually had any user-visible
consequences, but it's at least a bug waiting to happen.

In addition, do some code and docs review for DROP INDEX CONCURRENTLY;
some cosmetic code cleanup but mostly addition and revision of comments.

Portions of this need to be back-patched even further, but I'll work
on that separately.

Problem reported by Amit Kapila, diagnosis by Pavan Deolasee,
fix by Tom Lane and Andres Freund.

1 files changed, 12 insertions, 18 deletions

diff --git a/src/backend/commands/tablecmds.c b/src/backend/commands/tablecmds.c
index d88e3dee3fa..ab5ab940ade 100644
--- a/src/backend/commands/tablecmds.c
+++ b/src/backend/commands/tablecmds.c
@@ -743,10 +743,13 @@ RemoveRelations(DropStmt *drop)
 	int			flags = 0;
 	LOCKMODE	lockmode = AccessExclusiveLock;
 
+	/* DROP CONCURRENTLY uses a weaker lock, and has some restrictions */
 	if (drop->concurrent)
 	{
+		flags |= PERFORM_DELETION_CONCURRENTLY;
 		lockmode = ShareUpdateExclusiveLock;
-		if (list_length(drop->objects) > 1)
+		Assert(drop->removeType == OBJECT_INDEX);
+		if (list_length(drop->objects) != 1)
 			ereport(ERROR,
 					(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
 					 errmsg("DROP INDEX CONCURRENTLY does not support dropping multiple objects")));
@@ -838,19 +841,6 @@ RemoveRelations(DropStmt *drop)
 		add_exact_object_address(&obj, objects);
 	}
 
-	/*
-	 * Set options and check further requirements for concurrent drop
-	 */
-	if (drop->concurrent)
-	{
-		/*
-		 * Confirm that concurrent behaviour is restricted in grammar.
-		 */
-		Assert(drop->removeType == OBJECT_INDEX);
-
-		flags |= PERFORM_DELETION_CONCURRENTLY;
-	}
-
 	performMultipleDeletions(objects, drop->behavior, flags);
 
 	free_object_addresses(objects);
@@ -917,7 +907,7 @@ RangeVarCallbackForDropRelation(const RangeVar *rel, Oid relOid, Oid oldRelOid,
 	 * locking the index.  index_drop() will need this anyway, and since
 	 * regular queries lock tables before their indexes, we risk deadlock if
 	 * we do it the other way around.  No error if we don't find a pg_index
-	 * entry, though --- the relation may have been droppd.
+	 * entry, though --- the relation may have been dropped.
 	 */
 	if (relkind == RELKIND_INDEX && relOid != oldRelOid)
 	{
@@ -4783,6 +4773,8 @@ ATExecDropNotNull(Relation rel, const char *colName, LOCKMODE lockmode)
 
 	/*
 	 * Check that the attribute is not in a primary key
+	 *
+	 * Note: we'll throw error even if the pkey index is not valid.
 	 */
 
 	/* Loop over all indexes on the relation */
@@ -6317,7 +6309,7 @@ transformFkeyGetPrimaryKey(Relation pkrel, Oid *indexOid,
 	/*
 	 * Get the list of index OIDs for the table from the relcache, and look up
 	 * each one in the pg_index syscache until we find one marked primary key
-	 * (hopefully there isn't more than one such).
+	 * (hopefully there isn't more than one such).  Insist it's valid, too.
 	 */
 	*indexOid = InvalidOid;
 
@@ -6331,7 +6323,7 @@ transformFkeyGetPrimaryKey(Relation pkrel, Oid *indexOid,
 		if (!HeapTupleIsValid(indexTuple))
 			elog(ERROR, "cache lookup failed for index %u", indexoid);
 		indexStruct = (Form_pg_index) GETSTRUCT(indexTuple);
-		if (indexStruct->indisprimary)
+		if (indexStruct->indisprimary && IndexIsValid(indexStruct))
 		{
 			/*
 			 * Refuse to use a deferrable primary key.	This is per SQL spec,
@@ -6429,10 +6421,12 @@ transformFkeyCheckAttrs(Relation pkrel,
 
 		/*
 		 * Must have the right number of columns; must be unique and not a
-		 * partial index; forget it if there are any expressions, too
+		 * partial index; forget it if there are any expressions, too. Invalid
+		 * indexes are out as well.
 		 */
 		if (indexStruct->indnatts == numattrs &&
 			indexStruct->indisunique &&
+			IndexIsValid(indexStruct) &&
 			heap_attisnull(indexTuple, Anum_pg_index_indpred) &&
 			heap_attisnull(indexTuple, Anum_pg_index_indexprs))
 		{