Re: BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal - Mailing list pgsql-bugs

From Andrew Gierth
Subject Re: BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal
Date
Msg-id 87pnstgde6.fsf@news-spur.riddles.org.uk
Whole thread Raw
In response to BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal  (PG Bug reporting form <noreply@postgresql.org>)
Responses Re: BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal  (Alexander Lakhin <exclusion@gmail.com>)
List pgsql-bugs
>>>>> "PG" == PG Bug reporting form <noreply@postgresql.org> writes:

 PG> Bug reference:      15599
 PG> Logged by:          Alexander Lakhin
 PG> Email address:      exclusion@gmail.com
 PG> PostgreSQL version: 11.1
 PG> Operating system:   Ubuntu 18.04
 PG> Description:        

 PG> The PostgreSQL server crashes with the following query (excerpt
 PG> from regression test, run it after `make check`):

Not a bug.

 PG> CREATE FUNCTION overpaid(emp)
 PG>    RETURNS hobbies_r
 PG>    AS '/src/postgres/src/test/regress/regress.so'
 PG>    LANGUAGE C STRICT;

"overpaid" in regress.so returns a boolean scalar, not a row of any
kind. It is the responsibility of the person issuing CREATE FUNCTION for
a C-language function to get the types right; if they are wrong, it can
(as you observed) crash the server, which is why this action is
restricted to database superusers.

-- 
Andrew (irc:RhodiumToad)


pgsql-bugs by date:

Previous
From: PG Bug reporting form
Date:
Subject: BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal
Next
From: Alexander Lakhin
Date:
Subject: Re: BUG #15599: Segfault at MakeExpandedObjectReadOnlyInternal