BUG #3301: explain crashes server on simple (?) query - Mailing list pgsql-bugs

From Leszek Trenkner
Subject BUG #3301: explain crashes server on simple (?) query
Date
Msg-id 200705240118.l4O1Iha0047413@wwwmaster.postgresql.org
Whole thread Raw
Responses Re: BUG #3301: explain crashes server on simple (?) query  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
The following bug has been logged online:

Bug reference:      3301
Logged by:          Leszek Trenkner
Email address:      olaf@post.pl
PostgreSQL version: 8.3devel
Operating system:   Linux (Ubuntu Fiesty 7.04)
Description:        explain crashes server on simple (?) query
Details:

When I try to EXPLAIN or EXPLAIN analyze following query, server reports
crash and restarts (output from psql session follows):
------------------------------------------------
crash=# explain select t.id, foo.name from t join
(select ta.id, ta.name from ta
union select tb.id, tb.name from tb
) foo on foo.id = any (t.ids);
server closed the connection unexpectedly
        This probably means the server terminated abnormally
        before or while processing the request.
The connection to the server was lost. Attempting reset: Failed.
!>
------------------------------------------------
But, if I change UNION to UNION ALL  it works as expected:
------------------------------------------------
crash=# explain select t.id, foo.name from t join
(select ta.id, ta.name from ta
union all select tb.id, tb.name from tb
) foo on foo.id = any (t.ids);
                          QUERY PLAN
-------------------------------------------------------
 Nested Loop  (cost=1.05..4.09 rows=1 width=10)
   Join Filter: (foo.id = ANY (t.ids))
   ->  Append  (cost=0.00..2.06 rows=6 width=10)
         ->  Seq Scan on ta  (cost=0.00..1.03 rows=3 width=10)
         ->  Seq Scan on tb  (cost=0.00..1.03 rows=3 width=10)
   ->  Materialize  (cost=1.05..1.10 rows=5 width=47)
         ->  Seq Scan on t  (cost=0.00..1.05 rows=5 width=47)
(7 rows)
-------------------------------------------------------
This all on today's (2007-05-24) CSV 8.3 development version. The same bug
was in SVN version build from CSV on 2007-04-27. The server's log says:
-----------------------------------------
2007-05-24 02:27:44 CEST <@ > : LOG:  server process (PID 18536) was
terminated by signal 11: Segmentation fault
2007-05-24 02:27:44 CEST <@ > : LOG:  terminating any other active server
processes
2007-05-24 02:27:44 CEST <postgres@xxx postgres> idle: WARNING:  terminating
connection because of crash of another server process
2007-05-24 02:27:44 CEST <postgres@xxx postgres> idle: DETAIL:  The
postmaster has commanded this server process to roll back the current
transaction and exit, because another server process exited abnormally and
possibly corrupted shared memory.
2007-05-24 02:27:44 CEST <postgres@xxx postgres> idle: HINT:  In a moment
you should be able to reconnect to the database and repeat your command.
.
.
.
2007-05-24 02:27:44 CEST <@ > : LOG:  all server processes terminated;
reinitializing
---------------------------------------------------
I'll post schema for bug recreation on pgsql-bugs.

Greetings,
Leszek Trenkner

pgsql-bugs by date:

Previous
From: "Cstdenis"
Date:
Subject: BUG #3302: Crash on gist ltree - PANIC: failed to add item to index page
Next
From: "filipe paiva"
Date:
Subject: ...