Re: [HACKERS] current- crash - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: [HACKERS] current- crash
Date
Msg-id 199809210220.WAA16980@candle.pha.pa.us
Whole thread Raw
In response to current- crash  (Michael Reifenberger <root@totum.plaut.de>)
List pgsql-hackers
> Hi,
> I can easily crash the backend of current- postgres on current- FreeBSD with
> loading a database with test.sql and test.dmp an executing the script x.sql.
> Anyone else?
>
> Bye!
> ----
> Michael Reifenberger
> Plaut Software GmbH, R/3 Basis
Content-Description:

> CREATE TABLE b (begt datetime, kid int4);
> CREATE TABLE a (kid int4);
> CREATE TABLE c (a int4, b varchar(30), c int4);
> CREATE  INDEX b_0 on b using btree ( begt datetime_ops );
> CREATE  INDEX b_1 on b using btree ( kid int4_ops );
> CREATE  INDEX a_0 on a using btree ( kid int4_ops );
> Content-Description:

> COPY b FROM stdin;
> \.
> COPY a FROM stdin;
> \.
> COPY c FROM stdin;
> 1    foo    1
> 2    foo bar    2
> 3    \N    3
> 4    \\serverla    4
> \.
> Content-Description:

>   explain SELECT a.kid as foo
>     FROM a, b WHERE
>     a.kid = b.kid AND
>     ( b.kid = 23 OR
>       b.kid = 36 );
>

This is a known problem, and is going on the open items list.  The
problem is the new OR indexing code, and an attempt to test for a
MERGEJOIN where it really should not be.  I can see why it is happening,
and will have to research it.

--
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
http://www.op.net/~candle              |  (610) 353-9879(w)
  +  If your life is a hard drive,     |  (610) 853-3000(h)
  +  Christ can be your backup.        |

pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: [HACKERS] Transaction system (proposal for 6.5)
Next
From: Bruce Momjian
Date:
Subject: Re: Results of port of Sept 18 port of PostgreSQL