Re: pg_get__*_ddl consolidation
| От | Jeff Davis |
|---|---|
| Тема | Re: pg_get__*_ddl consolidation |
| Дата | |
| Msg-id | 4efc426cde8eed9734af3a4552c3be29188f3661.camel@j-davis.com обсуждение |
| Ответ на | Re: pg_get__*_ddl consolidation (Andrew Dunstan <andrew@dunslane.net>) |
| Ответы |
Re: pg_get__*_ddl consolidation
|
| Список | pgsql-hackers |
On Sun, 2026-04-05 at 11:06 -0400, Andrew Dunstan wrote:
> Pushed. I have moved the remaining get_*_ddl items to PG20-1
The line:
role_settings = DatumGetArrayTypeP(datum);
should be DatumGetArrayTypePCopy(), because it's being pfree()d later.
The existing code will sometimes make a copy and sometimes not, e.g.:
-- settings are contrived to make the datum inline
CREATE USER u1;
ALTER ROLE u1 SET search_path = 'public, pg_catalog, pg_temp';
ALTER ROLE u1 SET work_mem='64MB';
ALTER ROLE u1 SET statement_timeout='30s';
ALTER ROLE u1 SET lock_timeout='10s';
ALTER ROLE u1 SET idle_in_transaction_session_timeout = '60s';
SELECT pg_get_role_ddl('u1');
ERROR: pfree called with invalid pointer 0x7986dd0c7cc8 (header
0x0000400600000000)
Also, it looks like the scan key in pg_get_role_ddl_internal() uses
only the second attribute of the index, which might be worth a comment.
Regards,
Jeff Davis
В списке pgsql-hackers по дате отправления: