Re: Deleting prepared statements from libpq.
Jelte Fennema <me@jeltef.nl>
From: Jelte Fennema <me@jeltef.nl>
To: Michael Paquier <michael@paquier.xyz>
Cc: jian he <jian.universality@gmail.com>,
Dmitry Igrishin <dmitigr@gmail.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>, Daniele Varrazzo <daniele.varrazzo@gmail.com>
Date: 2023-06-23T07:39:00Z
Lists: pgsql-hackers
Attachments
- v6-0002-Support-sending-Close-messages-from-libpq.patch (application/octet-stream) patch v6-0002
- v6-0001-Correct-error-message-in-test_prepared.patch (application/octet-stream) patch v6-0001
On Fri, 23 Jun 2023 at 05:59, Michael Paquier <michael@paquier.xyz> wrote:
> [...]
> res = PQgetResult(conn);
> if (res == NULL)
> - pg_fatal("expected NULL result");
> + pg_fatal("expected non-NULL result");
>
> This should check for the NULL-ness of the result returned for
> PQclosePrepared() rather than changing the behavior of the follow-up
> calls?
To be clear, it didn't actually change the behaviour. I only changed
the error message, since it said the exact opposite of what it was
expecting. I split this minor fix into its own commit now to clarify
that. I think it would even make sense to commit this small patch to
the PG16 branch, since it's a bugfix in the tests (and possibly even
back-patch to others if that's not a lot of work). I changed the error
message to be in line with one from earlier in the test.
I addressed all of your other comments.
Commits
-
libpq: Add support for Close on portals and statements
- 28b572656184 17.0 landed
-
Fix incorrect error message in libpq_pipeline
- 451ca5c1e699 14.9 landed
- 4fd633df5080 15.4 landed
- f69a7f08fd01 16.0 landed