Correct obsolete nbtree recovery comments.

petergeoghegan · petergeoghegan · commit 741c4c41429b · 2018-12-18T16:59:49.000-08:00
Commit 40dae7e, which made the handling of interrupted nbtree page splits more robust, removed an nbtree-specific end-of-recovery cleanup step. This meant that it was no longer possible to complete an interrupted page split during recovery. However, a reference to recovery as a reason for using a NULL stack while inserting into a parent page was missed. Remove the reference. Remove a similar obsolete reference to recovery that was introduced much more recently, as part of the btree fastpath optimization enhancement that made it into Postgres 11 (commit 2b27273, and follow-up commits). Backpatch: 11-, where the fastpath optimization was introduced.
diff --git a/src/backend/access/nbtree/nbtinsert.c b/src/backend/access/nbtree/nbtinsert.c
@@ -887,10 +887,10 @@ _bt_insertonpg(Relation rel,
 		 * all the required conditions, including the fact that this page has
 		 * enough freespace. Note that this routine can in theory deal with
 		 * the situation where a NULL stack pointer is passed (that's what
-		 * would happen if the fastpath is taken), like it does during crash
-		 * recovery. But that path is much slower, defeating the very purpose
-		 * of the optimization.  The following assertion should protect us
-		 * from any future code changes that invalidate those assumptions.
+		 * would happen if the fastpath is taken). But that path is much
+		 * slower, defeating the very purpose of the optimization.  The
+		 * following assertion should protect us from any future code changes
+		 * that invalidate those assumptions.
 		 *
 		 * Note that whenever we fail to take the fastpath, we clear the
 		 * cached block. Checking for a valid cached block at this point is
@@ -1818,8 +1818,8 @@ _bt_checksplitloc(FindSplitData *state,
  * and it'd be possible for some other process to try to split or delete
  * one of these pages, and get confused because it cannot find the downlink.)
  *
- * stack - stack showing how we got here.  May be NULL in cases that don't
- *			have to be efficient (concurrent ROOT split, WAL recovery)
+ * stack - stack showing how we got here.  Will be NULL when splitting true
+ *			root, or during concurrent root split, where we can be inefficient
  * is_root - we split the true root
  * is_only - we split a page alone on its level (might have been fast root)
  */
