database/postgres
database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-05-05 09:19:17 +03:00
Code

Fix ndistinct estimates with system attributes

Browse Source 
When estimating the number of groups using extended statistics, the code
was discarding information about system attributes. This led to strange
situation that

    SELECT 1 FROM t GROUP BY ctid;

could have produced higher estimate (equal to pg_class.reltuples) than

    SELECT 1 FROM t GROUP BY a, b, ctid;

with extended statistics on (a,b). Fixed by retaining information about
the system attribute.

Backpatch all the way to 10, where extended statistics were introduced.

Author: Tomas Vondra
Backpatch-through: 10
This commit is contained in:
Tomas Vondra 2021-03-26 22:34:53 +01:00
parent a14a0118a1
commit 33e52ad9a3
2 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/backend/utils/adt/selfuncs.c
View File

@ -3987,11 +3987,11 @@ estimate_multivariate_ndistinct(PlannerInfo *root, RelOptInfo *rel,
attnum = ((Var *) varinfo->var)->varattno; attnum = ((Var *) varinfo->var)->varattno;
if (!AttrNumberIsForUserDefinedAttr(attnum)) if (AttrNumberIsForUserDefinedAttr(attnum) &&
bms_is_member(attnum, matched))
continue; continue;
if (!bms_is_member(attnum, matched)) newlist = lappend(newlist, varinfo);
newlist = lappend(newlist, varinfo);
} }
*varinfos = newlist; *varinfos = newlist;

2
src/test/regress/expected/stats_ext.out
View File

@ -260,7 +260,7 @@ SELECT s.stxkind, d.stxdndistinct
SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY ctid, a, b'); SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY ctid, a, b');
estimated | actual estimated | actual
-----------+-------- -----------+--------
11 | 1000 1000 | 1000
(1 row) (1 row)
-- Hash Aggregate, thanks to estimates improved by the statistic -- Hash Aggregate, thanks to estimates improved by the statistic