From: Tom Lane <tgl@sss.pgh.pa.us>
To: Etsuro Fujita <etsuro.fujita@gmail.com>
cc: eric.cyr@gmail.com, pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #17889: Invalid cursor direction for a foreign scan that
 reached the fetch_size (MOVE BACKWARD ALL IN cX)
In-reply-to: 
 <CAPmGK14u-jGV-4B2_T90Lf5U5Az2d5wH0hV3vfUbWkw2d4UjBA@mail.gmail.com>
References: <17889-e8c39a251d258dda@postgresql.org>
 <CAPmGK17vjjy0Yb=jyK_iCL5rpGW7w1imv=kQB7rkk9xMO4cEjA@mail.gmail.com>
 <CAPmGK16DNvnReiNyKprn4O-j5S2uGzUrcDANaF49md1-ejjjQw@mail.gmail.com>
 <CAPmGK14u-jGV-4B2_T90Lf5U5Az2d5wH0hV3vfUbWkw2d4UjBA@mail.gmail.com>
Comments: In-reply-to Etsuro Fujita <etsuro.fujita@gmail.com>
	message dated "Sat, 29 Apr 2023 20:42:02 +0900"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1852634.1682808624.1@sss.pgh.pa.us>
Date: Sat, 29 Apr 2023 18:50:24 -0400
Message-ID: <1852635.1682808624@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1852635.1682808624%40sss.pgh.pa.us>
Precedence: bulk

Etsuro Fujita <etsuro.fujita@gmail.com> writes:
>> I think that the root cause is in commit d844cd75a, which disallowed
>> rewinding and then re-fetching forwards in a NO SCROLL cursor.  I am
>> not sure what to do about this issue, but I am wondering whether that
>> commit is too restrictive, because 1) these examples would work just
>> fine without that commit, and 2) we still allow
>> rewind-and-fetch-forwards in a SCROLL cursor even when the query
>> includes volatile functions.

Well, the short answer here is that postgres_fdw is depending on
something that has squishy semantics, for the reasons enumerated
in d844cd75a (see also [1]).  Maybe we can carve out an exception
that's narrow enough that we can convince ourselves it's not squishy,
but I'm not very sure what the rules should be.

			regards, tom lane

[1] https://www.postgresql.org/message-id/201144.1682350828%40sss.pgh.pa.us