Ensure we have a snapshot when updating various system catalogs.

nathan-bossart · nathan-bossart · commit 24135398f1e1 · 2025-05-30T15:17:28.000-05:00
A few places that access system catalogs don't set up an active snapshot before potentially accessing their TOAST tables. To fix, push an active snapshot just before each section of code that might require accessing one of these TOAST tables, and pop it shortly afterwards. While at it, this commit adds some rather strict assertions in an attempt to prevent such issues in the future. Commit 16bf24e recently removed pg_replication_origen's TOAST table in order to fix the same problem for that catalog. On the back-branches, those bugs are left in place. We cannot easily remove a catalog's TOAST table on released major versions, and only replication origens with extremely long names are affected. Given the low severity of the issue, fixing older versions doesn't seem worth the trouble of significantly modifying the patch. Also, on v13 and v14, the aforementioned strict assertions have been omitted because commit 2776922, which added HaveRegisteredOrActiveSnapshot(), was not back-patched. While we could probably back-patch it now, I've opted against it because it seems unlikely that new TOAST snapshot issues will be introduced in the oldest supported versions. Reported-by: Alexander Lakhin <exclusion@gmail.com> Reviewed-by: Michael Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/18127-fe54b6a667f29658%40postgresql.org Discussion: https://postgr.es/m/18309-c0bf914950c46692%40postgresql.org Discussion: https://postgr.es/m/ZvMSUPOqUU-VNADN%40nathan Backpatch-through: 13
diff --git a/src/backend/access/heap/heapam.c b/src/backend/access/heap/heapam.c
@@ -54,6 +54,7 @@
 #include "catalog/catalog.h"
 #include "catalog/pg_database.h"
 #include "catalog/pg_database_d.h"
+#include "catalog/pg_replication_origen.h"
 #include "commands/vacuum.h"
 #include "miscadmin.h"
 #include "pgstat.h"
@@ -231,6 +232,38 @@ static const int MultiXactStatusLock[MaxMultiXactStatus + 1] =
 #define TUPLOCK_from_mxstatus(status) \
 			(MultiXactStatusLock[(status)])
 
+/*
+ * Check that we have a valid snapshot if we might need TOAST access.
+ */
+static inline void
+AssertHasSnapshotForToast(Relation rel)
+{
+#ifdef USE_ASSERT_CHECKING
+
+	/* bootstrap mode in particular breaks this rule */
+	if (!IsNormalProcessingMode())
+		return;
+
+	/* if the relation doesn't have a TOAST table, we are good */
+	if (!OidIsValid(rel->rd_rel->reltoastrelid))
+		return;
+
+	/*
+	 * Commit 16bf24e fixed accesses to pg_replication_origen without a
+	 * an active snapshot by removing its TOAST table.  On older branches,
+	 * these bugs are left in place.  Its only varlena column is roname (the
+	 * replication origen name), so this is only a problem if the name
+	 * requires out-of-line storage, which seems unlikely.  In any case,
+	 * fixing it doesn't seem worth extra code churn on the back-branches.
+	 */
+	if (RelationGetRelid(rel) == ReplicationOriginRelationId)
+		return;
+
+	Assert(HaveRegisteredOrActiveSnapshot());
+
+#endif							/* USE_ASSERT_CHECKING */
+}
+
 /* ----------------------------------------------------------------
  *						 heap support routines
  * ----------------------------------------------------------------
@@ -1858,6 +1891,8 @@ heap_insert(Relation relation, HeapTuple tup, CommandId cid,
 	Assert(HeapTupleHeaderGetNatts(tup->t_data) <=
 		   RelationGetNumberOfAttributes(relation));
 
+	AssertHasSnapshotForToast(relation);
+
 	/*
 	 * Fill in tuple header fields and toast the tuple if necessary.
 	 *
@@ -2135,6 +2170,8 @@ heap_multi_insert(Relation relation, TupleTableSlot **slots, int ntuples,
 	/* currently not needed (thus unsupported) for heap_multi_insert() */
 	Assert(!(options & HEAP_INSERT_NO_LOGICAL));
 
+	AssertHasSnapshotForToast(relation);
+
 	needwal = RelationNeedsWAL(relation);
 	saveFreeSpace = RelationGetTargetPageFreeSpace(relation,
 												   HEAP_DEFAULT_FILLFACTOR);
@@ -2557,6 +2594,8 @@ heap_delete(Relation relation, ItemPointer tid,
 
 	Assert(ItemPointerIsValid(tid));
 
+	AssertHasSnapshotForToast(relation);
+
 	/*
 	 * Forbid this during a parallel operation, lest it allocate a combo CID.
 	 * Other workers might need that combo CID for visibility checks, and we
@@ -3052,6 +3091,8 @@ heap_update(Relation relation, ItemPointer otid, HeapTuple newtup,
 	Assert(HeapTupleHeaderGetNatts(newtup->t_data) <=
 		   RelationGetNumberOfAttributes(relation));
 
+	AssertHasSnapshotForToast(relation);
+
 	/*
 	 * Forbid this during a parallel operation, lest it allocate a combo CID.
 	 * Other workers might need that combo CID for visibility checks, and we
diff --git a/src/backend/commands/indexcmds.c b/src/backend/commands/indexcmds.c
@@ -4064,12 +4064,20 @@ ReindexRelationConcurrently(Oid relationOid, ReindexParams *params)
 									 get_rel_namespace(oldidx->tableId),
 									 false);
 
+		/*
+		 * Swapping the indexes might involve TOAST table access, so ensure we
+		 * have a valid snapshot.
+		 */
+		PushActiveSnapshot(GetTransactionSnapshot());
+
 		/*
 		 * Swap old index with the new one.  This also marks the new one as
 		 * valid and the old one as not valid.
 		 */
 		index_concurrently_swap(newidx->indexId, oldidx->indexId, oldName);
 
+		PopActiveSnapshot();
+
 		/*
 		 * Invalidate the relcache for the table, so that after this commit
 		 * all sessions will refresh any cached plans that might reference the
diff --git a/src/backend/commands/tablecmds.c b/src/backend/commands/tablecmds.c
@@ -18789,9 +18789,17 @@ ATExecDetachPartition(List **wqueue, AlteredTableInfo *tab, Relation rel,
 		tab->rel = rel;
 	}
 
+	/*
+	 * Detaching the partition might involve TOAST table access, so ensure we
+	 * have a valid snapshot.
+	 */
+	PushActiveSnapshot(GetTransactionSnapshot());
+
 	/* Do the final part of detaching */
 	DetachPartitionFinalize(rel, partRel, concurrent, defaultPartOid);
 
+	PopActiveSnapshot();
+
 	ObjectAddressSet(address, RelationRelationId, RelationGetRelid(partRel));
 
 	/* keep our lock until commit */
diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c
@@ -2372,6 +2372,12 @@ do_autovacuum(void)
 						get_namespace_name(classForm->relnamespace),
 						NameStr(classForm->relname))));
 
+		/*
+		 * Deletion might involve TOAST table access, so ensure we have a
+		 * valid snapshot.
+		 */
+		PushActiveSnapshot(GetTransactionSnapshot());
+
 		object.classId = RelationRelationId;
 		object.objectId = relid;
 		object.objectSubId = 0;
@@ -2384,6 +2390,7 @@ do_autovacuum(void)
 		 * To commit the deletion, end current transaction and start a new
 		 * one.  Note this also releases the locks we took.
 		 */
+		PopActiveSnapshot();
 		CommitTransactionCommand();
 		StartTransactionCommand();
 
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
@@ -4725,8 +4725,16 @@ ApplyWorkerMain(Datum main_arg)
 			walrcv_startstreaming(LogRepWorkerWalRcvConn, &options);
 
 			StartTransactionCommand();
+
+			/*
+			 * Updating pg_subscription might involve TOAST table access, so
+			 * ensure we have a valid snapshot.
+			 */
+			PushActiveSnapshot(GetTransactionSnapshot());
+
 			UpdateTwoPhaseState(MySubscription->oid, LOGICALREP_TWOPHASE_STATE_ENABLED);
 			MySubscription->twophasestate = LOGICALREP_TWOPHASE_STATE_ENABLED;
+			PopActiveSnapshot();
 			CommitTransactionCommand();
 		}
 		else
@@ -4779,7 +4787,15 @@ DisableSubscriptionAndExit(void)
 
 	/* Disable the subscription */
 	StartTransactionCommand();
+
+	/*
+	 * Updating pg_subscription might involve TOAST table access, so ensure we
+	 * have a valid snapshot.
+	 */
+	PushActiveSnapshot(GetTransactionSnapshot());
+
 	DisableSubscription(MySubscription->oid);
+	PopActiveSnapshot();
 	CommitTransactionCommand();
 
 	/* Ensure we remove no-longer-useful entry for worker's start time */
@@ -4883,6 +4899,12 @@ clear_subscription_skip_lsn(XLogRecPtr finish_lsn)
 		started_tx = true;
 	}
 
+	/*
+	 * Updating pg_subscription might involve TOAST table access, so ensure we
+	 * have a valid snapshot.
+	 */
+	PushActiveSnapshot(GetTransactionSnapshot());
+
 	/*
 	 * Protect subskiplsn of pg_subscription from being concurrently updated
 	 * while clearing it.
@@ -4941,6 +4963,8 @@ clear_subscription_skip_lsn(XLogRecPtr finish_lsn)
 	heap_freetuple(tup);
 	table_close(rel, NoLock);
 
+	PopActiveSnapshot();
+
 	if (started_tx)
 		CommitTransactionCommand();
 }
