Re: pg9.6 segfault using simple query (related to use fk for join estimates) - Mailing list pgsql-hackers

From Julien Rouhaud
Subject Re: pg9.6 segfault using simple query (related to use fk for join estimates)
Date
Msg-id 572384D7.1020602@dalibo.com
Whole thread Raw
In response to Re: pg9.6 segfault using simple query (related to use fk for join estimates)  (Michael Paquier <michael.paquier@gmail.com>)
Responses Re: pg9.6 segfault using simple query (related to use fk for join estimates)  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On 29/04/2016 13:20, Michael Paquier wrote:
> On Fri, Apr 29, 2016 at 7:25 PM, Stefan Huehner <stefan@huehner.org> wrote:
>> If you need any more info or testing done just let me know.
>
> The information you are providing is sufficient to reproduce the
> problem, thanks! I have added this bug to the list of open items for
> 9.6.
>

The segfault is caused by quals_match_foreign_key() calling get_leftop()
and get_rightop() on a ScalarArrayOpExpr node.

Reordering the common fields of OpExpr and ScalarArrayOpExpr at the
beginning of the struct so the get_*op() work with either (as in
attached patch) fixes the issue.

I'm not sure that assuming this compatibility is the right way to fix
this though.

--
Julien Rouhaud
http://dalibo.com - http://dalibo.org

Attachment

pgsql-hackers by date:

Previous
From: Rodney Lott
Date:
Subject: postgresql 9.3.10, FIPS mode and DRBG issues.
Next
From: Simon Riggs
Date:
Subject: Re: Lets (not) break all the things. Was: [pgsql-advocacy] 9.6 -> 10.0