Re: Support prepared statement invalidation when result types change

Another similar issue was reported on the PgBouncer prepared statement PR[1].

It's related to an issue that was already reported on the
mailinglist[2]. It turns out that invalidation of the argument types
is also important in some cases. The newly added 3rd patch in this
series addresses that issue.

[1]: https://github.com/pgbouncer/pgbouncer/pull/845#discussion_r1309454695
[2]: https://www.postgresql.org/message-id/flat/CA%2Bmi_8YAGf9qibDFTRNKgaTwaBa1OUcteKqLAxfMmKFbo3GHZg%40mail.gmail.com