NotificationsYou must be signed in to change notification settings
Fork6
Star31

Commit895267a

committed

Remove unneeded nbtree latestRemovedXid comments.

Discussing the low level issue of nbtree VACUUM and recovery conflictsin btvacuumpage() now seems inappropriate. The same issue is discussedin nbtxlog.h, as well as in a comment block above _bt_delitems_vacuum().The comment block made more sense when it was part of a broaderdiscussion of nbtree VACUUM "pin scans". These were removed by commit9f83468.

1 parente6a7600 commit895267aCopy full SHA for 895267a

File tree

1 file changed

-26

lines changed

src/backend/access/nbtree
- nbtree.c

1 file changed

-26

lines changed

`‎src/backend/access/nbtree/nbtree.c‎`

Lines changed: 1 addition & 26 deletions

Original file line number	Diff line number	Diff line change
`@@ -1183,12 +1183,6 @@ btvacuumpage(BTVacState *vstate, BlockNumber scanblkno)`
`1183`	`1183`	`opaque->btpo_next<scanblkno)`
`1184`	`1184`	`backtrack_to=opaque->btpo_next;`
`1185`	`1185`
`1186`		`-/*`
`1187`		`- * When each VACUUM begins, it determines an OldestXmin cutoff value.`
`1188`		`- * Tuples before the cutoff are removed by VACUUM. Scan over all`
`1189`		`- * items to see which ones need to be deleted according to cutoff`
`1190`		`- * point using callback.`
`1191`		`- */`
`1192`	`1186`	`ndeletable=0;`
`1193`	`1187`	`nupdatable=0;`
`1194`	`1188`	`minoff=P_FIRSTDATAKEY(opaque);`
`@@ -1197,6 +1191,7 @@ btvacuumpage(BTVacState *vstate, BlockNumber scanblkno)`
`1197`	`1191`	`nhtidslive=0;`
`1198`	`1192`	`if (callback)`
`1199`	`1193`	`{`
	`1194`	`+/* btbulkdelete callback tells us what to delete (or update) */`
`1200`	`1195`	`for (offnum=minoff;`
`1201`	`1196`	`offnum <=maxoff;`
`1202`	`1197`	`offnum=OffsetNumberNext(offnum))`
`@@ -1206,26 +1201,6 @@ btvacuumpage(BTVacState *vstate, BlockNumber scanblkno)`
`1206`	`1201`	`itup= (IndexTuple)PageGetItem(page,`
`1207`	`1202`	`PageGetItemId(page,offnum));`
`1208`	`1203`
`1209`		`-/*`
`1210`		`- * Hot Standby assumes that it's okay that XLOG_BTREE_VACUUM`
`1211`		`- * records do not produce their own conflicts. This is safe`
`1212`		`- * as long as the callback function only considers whether the`
`1213`		`- * index tuple refers to pre-cutoff heap tuples that were`
`1214`		`- * certainly already pruned away during VACUUM's initial heap`
`1215`		`- * scan by the time we get here. (heapam's XLOG_HEAP2_PRUNE`
`1216`		`- * records produce conflicts using a latestRemovedXid value`
`1217`		`- * for the pointed-to heap tuples, so there is no need to`
`1218`		`- * produce our own conflict now.)`
`1219`		`- *`
`1220`		`- * Backends with snapshots acquired after a VACUUM starts but`
`1221`		`- * before it finishes could have visibility cutoff with a`
`1222`		`- * later xid than VACUUM's OldestXmin cutoff. These backends`
`1223`		`- * might happen to opportunistically mark some index tuples`
`1224`		`- * LP_DEAD before we reach them, even though they may be after`
`1225`		`- * our cutoff. We don't try to kill these "extra" index`
`1226`		`- * tuples in _bt_delitems_vacuum(). This keep things simple,`
`1227`		`- * and allows us to always avoid generating our own conflicts.`
`1228`		`- */`
`1229`	`1204`	`Assert(!BTreeTupleIsPivot(itup));`
`1230`	`1205`	`if (!BTreeTupleIsPosting(itup))`
`1231`	`1206`	`{`

0 commit comments

Comments

(0)

Movatterモバイル変換

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Commit895267a

File tree

1 file changed

1 file changed

`‎src/backend/access/nbtree/nbtree.c‎`

0 commit comments