Commit9abbf47

committed

Another fix for single-page hash index vacuum.

The WAL consistency checking code needed to be updated for the newpage status bit, but that didn't get done previously.Ashutosh Sharma, reviewed by Amit KapilaDiscussion:http://postgr.es/m/CAA4eK1LP_oz4EfMen14OjJuzN5CqPdfRkFFuA-MfkcfeE8zGyg@mail.gmail.com

1 parent953477c commit9abbf47Copy full SHA for 9abbf47

File tree

-1

lines changed

-1

lines changed

Lines changed: 7 additions & 1 deletion

Original file line number	Diff line number	Diff line change
`@@ -1250,8 +1250,14 @@ hash_mask(char *pagedata, BlockNumber blkno)`
`1250`	`1250`	`/*`
`1251`	`1251`	`* In hash bucket and overflow pages, it is possible to modify the`
`1252`	`1252`	`* LP_FLAGS without emitting any WAL record. Hence, mask the line`
`1253`		`- * pointer flags. See hashgettuple() for details.`
	`1253`	`+ * pointer flags. See hashgettuple(), _hash_kill_items() for details.`
`1254`	`1254`	`*/`
`1255`	`1255`	`mask_lp_flags(page);`
`1256`	`1256`	`}`
	`1257`	`+`
	`1258`	`+/*`
	`1259`	`+ * It is possible that the hint bit LH_PAGE_HAS_DEAD_TUPLES may remain`
	`1260`	`+ * unlogged. So, mask it. See _hash_kill_items() for details.`
	`1261`	`+ */`
	`1262`	`+opaque->hasho_flag &= ~LH_PAGE_HAS_DEAD_TUPLES;`
`1257`	`1263`	`}`

Comments

(0)