Re: creating objects in pg_catalog

Robert Haas <robertmhaas@gmail.com>

From: Robert Haas <robertmhaas@gmail.com>

To: Tom Lane <tgl@sss.pgh.pa.us>

Cc: pgsql-hackers@postgresql.org

Date: 2012-06-07T12:44:26Z

Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

Checking to decide whether relations are system relations now depends
- 9999f5a10e72 7.3.1 cited

Attachments

remove-heap-create-check.patch (application/octet-stream) patch

On Wed, Jun 6, 2012 at 5:21 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> rhaas=# create table pg_catalog.tom (a int);
>> ERROR:  permission denied to create "pg_catalog.tom"
>
>> The offending error check is in heap_create(), and based on what
>> you're saying here it seems like we should just rip it out.
>
> Hmm.  Yeah, it seems like the regular permissions tests on the schemas
> in question should be enough to keep Joe User from making tables there,
> and I do not see a reason why the backend would care if there are
> non-catalog tables laying about in pg_catalog.
>
> Checking the commit history, it seems this was originally a test to
> prevent people from creating tables named "pg_xxx":
>
> http://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=9999f5a10e722c052006886b678995695001958a#patch3
>
> which may or may not have been critical once upon a time, but surely is
> not any more.
>
> So no objection to removing that particular test.

Patch attached.  Barring objections, I'll apply this to 9.3 once we branch.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company