Date: Sat, 12 Jun 2021 10:12:34 -0500
From: Justin Pryzby <pryzby@telsasoft.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers@lists.postgresql.org
Subject: Re: pg_filenode_relation(0,0) elog
Message-ID: <20210612151234.GW16435@telsasoft.com>
References: <20210612023324.GT16435@telsasoft.com>
 <2146777.1623469895@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <2146777.1623469895@sss.pgh.pa.us>
User-Agent: Mutt/1.9.4 (2018-02-28)
Archived-At: 
 <https://www.postgresql.org/message-id/20210612151234.GW16435%40telsasoft.com>
Precedence: bulk

On Fri, Jun 11, 2021 at 11:51:35PM -0400, Tom Lane wrote:
> Justin Pryzby <pryzby@telsasoft.com> writes:
> > Per sqlsmith.
> > postgres=# SELECT pg_filenode_relation(0,0);
> > ERROR:  unexpected duplicate for tablespace 0, relfilenode 0
> 
> Ugh.
> 
> > The usual expectation is that sql callable functions should return null rather
> > than hitting elog().
> 
> Agreed, but you should put the short-circuit into the SQL-callable
> function, ie pg_filenode_relation.  Lower-level callers ought not be
> passing junk data.

Right.  I spent inadequate time reading output of git grep.

> Likely it should check the reltablespace, too.

I don't think so.  The docs say:
https://www.postgresql.org/docs/current/functions-admin.html#FUNCTIONS-ADMIN-DBLOCATION
|For a relation in the database's default tablespace, the tablespace can be specified as zero.

Also, that would breaks expected/alter_table.out for the same reason.

diff --git a/src/backend/utils/adt/dbsize.c b/src/backend/utils/adt/dbsize.c
index 3c70bb5943..144aca1099 100644
--- a/src/backend/utils/adt/dbsize.c
+++ b/src/backend/utils/adt/dbsize.c
@@ -905,6 +905,9 @@ pg_filenode_relation(PG_FUNCTION_ARGS)
 	Oid			relfilenode = PG_GETARG_OID(1);
 	Oid			heaprel = InvalidOid;
 
+	if (!OidIsValid(relfilenode))
+		PG_RETURN_NULL();
+
 	heaprel = RelidByRelfilenode(reltablespace, relfilenode);
 
 	if (!OidIsValid(heaprel))