Feedback-ID: iecce486f:Fastmail
MIME-Version: 1.0
Date: Wed, 15 Apr 2026 20:44:38 -0400
From: "Jack Bonatakis" <jack@bonatak.is>
To: pgsql-hackers@lists.postgresql.org
Message-Id: <28b82ab2-5721-4e7c-bf71-979c3f198a2e@app.fastmail.com>
In-Reply-To: <573E45C1-31A4-4885-A00C-1A2171159A2A@gmail.com>
References: <573E45C1-31A4-4885-A00C-1A2171159A2A@gmail.com>
Subject: Re: Fix a server crash problem from pg_get_database_ddl
Content-Type: multipart/alternative;
 boundary=bf67c97b0ae7b25bc7d0be880e97593e4efb553a
Archived-At: 
 <https://www.postgresql.org/message-id/28b82ab2-5721-4e7c-bf71-979c3f198a2e%40app.fastmail.com>
Precedence: bulk

--bf67c97b0ae7b25bc7d0be880e97593e4efb553a
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

I have reproduced this error against the current master:

```
CREATE TABLESPACE ts1 LOCATION '/workspace/tablespaces/pg_bug_ts1';
CREATE DATABASE db1 TABLESPACE ts1;
DELETE FROM pg_tablespace WHERE spcname = 'ts1';
SELECT * FROM pg_get_database_ddl('db1'::regdatabase);

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.
```
Backend logs show:

```
[1] LOG:  client backend (PID 15420) was terminated by signal 11: Segmentation fault
[1] DETAIL:  Failed process was running: SELECT * FROM pg_get_database_ddl('db1'::regdatabase);
[1] LOG:  terminating any other active server processes
```
After applying the patch:

```
SELECT * FROM pg_get_database_ddl('db1'::regdatabase);
ERROR:  tablespace with OID 16393 does not exist
HINT:  To recover, try ALTER DATABASE ... SET TABLESPACE ... to a valid tablespace.
```
and backend logs show:

```
[56] ERROR:  tablespace with OID 16393 does not exist
[56] HINT:  To recover, try ALTER DATABASE ... SET TABLESPACE ... to a valid tablespace.
[56] STATEMENT:  SELECT * FROM pg_get_database_ddl('db1'::regdatabase);
```
All tests pass.

The only note I'd have on the code change is that there is no accompanying test. It seems like a TAP test would be reasonable, but I am quite new and will defer to whether you think that's the right call or even necessary. 

Jack
--bf67c97b0ae7b25bc7d0be880e97593e4efb553a
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div>I have repro=
duced this error against the current master:</div><div><br></div><div>``=
`</div><div>CREATE TABLESPACE ts1 LOCATION '/workspace/tablespaces/pg_bu=
g_ts1';<br></div><div>CREATE DATABASE db1 TABLESPACE ts1;<br></div><div>=
DELETE FROM pg_tablespace WHERE spcname =3D 'ts1';<br></div><div>SELECT =
* FROM pg_get_database_ddl('db1'::regdatabase);</div><div><br></div><div=
>server closed the connection unexpectedly</div><div>This probably means=
 the server terminated abnormally</div><div>before or while processing t=
he request.</div><div>The connection to the server was lost. Attempting =
reset: Failed.<br></div><div>```</div><div>Backend logs show:</div><div>=
<br></div><div>```</div><div>[1] LOG:&nbsp; client backend (PID 15420) w=
as terminated by signal 11: Segmentation fault<br></div><div>[1] DETAIL:=
&nbsp; Failed process was running: SELECT * FROM pg_get_database_ddl('db=
1'::regdatabase);<br></div><div>[1] LOG:&nbsp; terminating any other act=
ive server processes</div><div>```<br></div><div>After applying the patc=
h:</div><div><br></div><div>```</div><div>SELECT * FROM pg_get_database_=
ddl('db1'::regdatabase);</div><div>ERROR:&nbsp; tablespace with OID 1639=
3 does not exist</div><div>HINT:&nbsp; To recover, try ALTER DATABASE ..=
. SET TABLESPACE ... to a valid tablespace.</div><div>```</div><div>and =
backend logs show:</div><div><br></div><div>```</div><div>[56] ERROR:&nb=
sp; tablespace with OID 16393 does not exist</div><div>[56] HINT:&nbsp; =
To recover, try ALTER DATABASE ... SET TABLESPACE ... to a valid tablesp=
ace.</div><div>[56] STATEMENT:&nbsp; SELECT * FROM pg_get_database_ddl('=
db1'::regdatabase);<br></div><div>```</div><div>All tests pass.</div><di=
v><br></div><div>The only note I'd have on the code change is that there=
 is no accompanying test. It seems like a TAP test would be reasonable, =
but I am quite new and will defer to whether you think that's the right =
call or even necessary.&nbsp;</div><div><br></div><div>Jack</div></body>=
</html>
--bf67c97b0ae7b25bc7d0be880e97593e4efb553a--