Re: Patching for increasing the number of columns - Mailing list pgsql-hackers

Tom wrote:> You would have to show us the actual failure diffs to get much useful 
comment, but in general increasing the size of tuple headers could 
easily lead to> changes in plan choices

Thank you Tom. So there is some hope! In effect the query plan is 
different for the join and the view tests. The result set is different 
only for the 'without_oid' test.
A side question: Are these tests comprehensive, or should I run other 
tests just to be sure? Hints on where to find those tests are welcome.
Thanks!
(diff below)
Mayeul
*** 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/expected/join.out 
2014-03-17 19:35:47.000000000 +0000
--- 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/results/join.out 
2014-08-20 15:40:56.248603754 +0100
***************
*** 2791,2814 ****    join int4_tbl i1 on b.thousand = f1    right join int4_tbl i2 on i2.f1 = b.tenthous    order by
1;
!                                        QUERY PLAN
! 
-----------------------------------------------------------------------------------------   Sort     Sort Key:
b.unique1
!    ->  Nested Loop Left Join
!          ->  Seq Scan on int4_tbl i2           ->  Nested Loop Left Join                 Join Filter: (b.unique1 =
42)                ->  Nested Loop                       ->  Nested Loop                             ->  Seq Scan on
int4_tbli1
 
!                            ->  Index Scan using tenk1_thous_tenthous 
on tenk1 b
!                                  Index Cond: ((thousand = i1.f1) AND 
(i2.f1 = tenthous))                       ->  Index Scan using tenk1_unique1 on tenk1 a
IndexCond: (unique1 = b.unique2)                 ->  Index Only Scan using tenk1_thous_tenthous on tenk1 c
        Index Cond: (thousand = a.thousand)
 
! (15 rows)
  select b.unique1 from    tenk1 a join tenk1 b on a.unique1 = b.unique2
--- 2791,2818 ----    join int4_tbl i1 on b.thousand = f1    right join int4_tbl i2 on i2.f1 = b.tenthous    order by
1;
!                                   QUERY PLAN
! 
-------------------------------------------------------------------------------   Sort     Sort Key: b.unique1
!    ->  Hash Right Join
!          Hash Cond: (b.tenthous = i2.f1)           ->  Nested Loop Left Join                 Join Filter: (b.unique1
=42)                 ->  Nested Loop                       ->  Nested Loop                             ->  Seq Scan on
int4_tbli1
 
!                            ->  Bitmap Heap Scan on tenk1 b
!                                  Recheck Cond: (thousand = i1.f1)
!                                  ->  Bitmap Index Scan on 
tenk1_thous_tenthous
!                                        Index Cond: (thousand = i1.f1)                       ->  Index Scan using
tenk1_unique1on tenk1 a                             Index Cond: (unique1 = b.unique2)                 ->  Index Only
Scanusing tenk1_thous_tenthous on tenk1 c                       Index Cond: (thousand = a.thousand)
 
!          ->  Hash
!                ->  Seq Scan on int4_tbl i2
! (19 rows)
  select b.unique1 from    tenk1 a join tenk1 b on a.unique1 = b.unique2

======================================================================

*** 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/expected/select_views_1.out 
2014-03-17 19:35:47.000000000 +0000
--- 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/results/select_views.out 
2014-08-20 15:41:01.212603532 +0100
***************
*** 1413,1423 ****         WHERE f_leak(cnum) AND ymd >= '2011-10-01' AND ymd < '2011-11-01';
        QUERY PLAN
 
------------------------------------------------------------------------------
!  Nested Loop
!    Join Filter: (l.cid = r.cid)     ->  Seq Scan on credit_usage r           Filter: ((ymd >= '10-01-2011'::date) AND
(ymd< 
 
'11-01-2011'::date))
!    ->  Materialize           ->  Subquery Scan on l                 Filter: f_leak(l.cnum)                 ->  Hash
Join
--- 1413,1423 ----         WHERE f_leak(cnum) AND ymd >= '2011-10-01' AND ymd < '2011-11-01';
        QUERY PLAN
 
------------------------------------------------------------------------------
!  Hash Join
!    Hash Cond: (r.cid = l.cid)     ->  Seq Scan on credit_usage r           Filter: ((ymd >= '10-01-2011'::date) AND
(ymd< 
 
'11-01-2011'::date))
!    ->  Hash           ->  Subquery Scan on l                 Filter: f_leak(l.cnum)                 ->  Hash Join
***************
*** 1446,1456 ****
------------------------------------------------------------------------------------   Subquery Scan on
my_credit_card_usage_secure    Filter: f_leak(my_credit_card_usage_secure.cnum)
 
!    ->  Nested Loop
!          Join Filter: (l.cid = r.cid)           ->  Seq Scan on credit_usage r                 Filter: ((ymd >=
'10-01-2011'::date)AND (ymd < 
 
'11-01-2011'::date))
!          ->  Materialize                 ->  Hash Join                       Hash Cond: (r_1.cid = l.cid)
         ->  Seq Scan on credit_card r_1
 
--- 1446,1456 ----
------------------------------------------------------------------------------------   Subquery Scan on
my_credit_card_usage_secure    Filter: f_leak(my_credit_card_usage_secure.cnum)
 
!    ->  Hash Join
!          Hash Cond: (r.cid = l.cid)           ->  Seq Scan on credit_usage r                 Filter: ((ymd >=
'10-01-2011'::date)AND (ymd < 
 
'11-01-2011'::date))
!          ->  Hash                 ->  Hash Join                       Hash Cond: (r_1.cid = l.cid)
  ->  Seq Scan on credit_card r_1
 

======================================================================

*** 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/expected/without_oid.out 
2014-03-17 19:35:47.000000000 +0000
--- 
~/bin/postgresql_9.3.4/patched_3_64/postgresql-9.3.4/src/test/regress/results/without_oid.out 
2014-08-20 15:41:02.068603494 +0100
***************
*** 53,59 ****   WHERE relname IN ('wi', 'wo');   ?column? | ?column?  ----------+----------
!  t        |        0  (1 row)
  DROP TABLE wi;
--- 53,59 ----   WHERE relname IN ('wi', 'wo');   ?column? | ?column?  ----------+----------
!  f        |        0  (1 row)
  DROP TABLE wi;

======================================================================