Commit2e54d88

committed

Correct comment in Hot Standby nbtree handling

Logic is correct, matching handling of LP_DEAD elsewhere.

1 parent284c464 commit2e54d88Copy full SHA for 2e54d88

File tree

-5

lines changed

-5

lines changed

Lines changed: 5 additions & 5 deletions

Original file line number	Diff line number	Diff line change
`@@ -714,11 +714,11 @@ btree_xlog_delete_get_latestRemovedXid(xl_btree_delete *xlrec)`
`714`	`714`	`UnlockReleaseBuffer(ibuffer);`
`715`	`715`
`716`	`716`	`/*`
`717`		`- *XXXIf all heap tuples were LP_DEAD then we will be returning`
`718`		`- * InvalidTransactionId here,causing conflict for all HS transactions.`
`719`		`- *That should happen very rarely (reasoning please?). Also note that`
`720`		`- *caller can't tellthedifference between this case andthefast path`
`721`		`- *exit above. May need to change that in future.`
	`717`	`+ * If all heap tuples were LP_DEAD then we will be returning`
	`718`	`+ * InvalidTransactionId here,which avoids conflicts. This matches`
	`719`	`+ *existing logic which assumes that LP_DEAD tuples must already be`
	`720`	`+ *older thanthelatestRemovedXid onthecleanup record that`
	`721`	`+ *set them as LP_DEAD, hence must already have generated a conflict.`
`722`	`722`	`*/`
`723`	`723`	`returnlatestRemovedXid;`
`724`	`724`	`}`

Comments

(0)