MIME-Version: 1.0
References: <CAPmGK169vQ83PQwQkoxO-AK2EeK1EsgsxixedM+BLWEAhZ_AqQ@mail.gmail.com>
 <D7DDEE03-A38F-42E3-B946-7206B1D2B78A@yesql.se>
In-Reply-To: <D7DDEE03-A38F-42E3-B946-7206B1D2B78A@yesql.se>
From: Etsuro Fujita <etsuro.fujita@gmail.com>
Date: Thu, 16 Mar 2023 18:28:41 +0900
Message-ID: <CAPmGK15TZ9KsqBemCkCVdKZq8x0Uzi4ARUYiV8UCHYKOKznjDQ@mail.gmail.com>
Subject: Re: postgres_fdw: Useless if-test in GetConnection()
To: Daniel Gustafsson <daniel@yesql.se>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://www.postgresql.org/message-id/CAPmGK15TZ9KsqBemCkCVdKZq8x0Uzi4ARUYiV8UCHYKOKznjDQ%40mail.gmail.com>
Precedence: bulk

On Wed, Mar 15, 2023 at 7:58=E2=80=AFPM Daniel Gustafsson <daniel@yesql.se>=
 wrote:
> > On 15 Mar 2023, at 11:18, Etsuro Fujita <etsuro.fujita@gmail.com> wrote=
:
> > While working on something else, I noticed that the =E2=80=9Cif (entry-=
>conn
> > =3D=3D NULL)=E2=80=9D test after doing disconnect_pg_server() when re-e=
stablishing
> > a given connection in GetConnection() is pointless, because the former
> > function ensures that entry->conn is NULL.  So I removed the if-test.
> > Attached is a patch for that.
>
> LGTM, nice catch.
>
> > I think we could instead add an assertion, but I did not, because we al=
ready
> > have it in make_new_connection().
>
> Agreed, the assertion in make_new_connection is enough (and is needed the=
re).

Great!  Thanks for looking!

Best regards,
Etsuro Fujita