Commit41fdcf7

committed

Correct comment in Hot Standby nbtree handling

Logic is correct, matching handling of LP_DEAD elsewhere.

1 parenta7a3e71 commit41fdcf7Copy full SHA for 41fdcf7

File tree

-5

lines changed

-5

lines changed

Lines changed: 5 additions & 5 deletions

Original file line number	Diff line number	Diff line change
`@@ -719,11 +719,11 @@ btree_xlog_delete_get_latestRemovedXid(xl_btree_delete *xlrec)`
`719`	`719`	`UnlockReleaseBuffer(ibuffer);`
`720`	`720`
`721`	`721`	`/*`
`722`		`- *XXXIf all heap tuples were LP_DEAD then we will be returning`
`723`		`- * InvalidTransactionId here,causing conflict for all HS transactions.`
`724`		`- *That should happen very rarely (reasoning please?). Also note that`
`725`		`- *caller can't tellthedifference between this case andthefast path`
`726`		`- *exit above. May need to change that in future.`
	`722`	`+ * If all heap tuples were LP_DEAD then we will be returning`
	`723`	`+ * InvalidTransactionId here,which avoids conflicts. This matches`
	`724`	`+ *existing logic which assumes that LP_DEAD tuples must already be`
	`725`	`+ *older thanthelatestRemovedXid onthecleanup record that`
	`726`	`+ *set them as LP_DEAD, hence must already have generated a conflict.`
`727`	`727`	`*/`
`728`	`728`	`returnlatestRemovedXid;`
`729`	`729`	`}`

Comments

(0)