From: Tom Lane <tgl@sss.pgh.pa.us>
To: Robert Haas <robertmhaas@gmail.com>
cc: Arthur Zakirov <a.zakirov@postgrespro.ru>,
        Tomas Vondra <tomas.vondra@2ndquadrant.com>,
        Pavel Stehule <pavel.stehule@gmail.com>,
        Andres Freund <andres@anarazel.de>,
        Ildus Kurbangaliev <i.kurbangaliev@postgrespro.ru>,
        pgsql-hackers <pgsql-hackers@postgresql.org>
Subject: Re: [PROPOSAL] Shared Ispell dictionaries
In-reply-to: 
 <CA+Tgmobe8d4Bx+ZVXoYqGm--ASWfyTGRwOxyjatKmKtkPt888Q@mail.gmail.com>
References: <20180322105603.GA23544@zakirov.localdomain>
 <28250.1521924996@sss.pgh.pa.us>
 <ada73d07-6c88-acab-7e31-75b2eb724d24@2ndquadrant.com>
 <25186.1521951490@sss.pgh.pa.us> <20180325205408.GA19457@arthur.localdomain>
 <27296.1522078068@sss.pgh.pa.us> <20180327121954.GA12726@zakirov.localdomain>
 <CA+TgmoaR_fVd_u60mA-PHO=J1u0PmdVGOt9WaxZ1UYuGq276oQ@mail.gmail.com>
 <20180516113631.GA29544@zakirov.localdomain>
 <CA+TgmoY=NVOA9YqoOVWzSkZcw2ammQpA0PjMyaZ8j-oJZY3dXA@mail.gmail.com>
 <20180516204230.GA2397@artur-book.localdomain>
 <CA+Tgmobe8d4Bx+ZVXoYqGm--ASWfyTGRwOxyjatKmKtkPt888Q@mail.gmail.com>
Comments: In-reply-to Robert Haas <robertmhaas@gmail.com>
	message dated "Thu, 17 May 2018 09:57:59 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <27787.1526566736.1@sss.pgh.pa.us>
Date: Thu, 17 May 2018 10:18:56 -0400
Message-ID: <27788.1526566736@sss.pgh.pa.us>
Precedence: bulk

Robert Haas <robertmhaas@gmail.com> writes:
> ... Assuming that we can
> convince ourselves that that much is OK, I don't see why using a
> syscache callback to help ensure that the mappings are blown away in
> an at-least-somewhat-timely fashion is worse than any other approach.

I think the point you've not addressed is that "syscache callback
occurred" does not equate to "object was dropped".  Can the code
survive having this occur at any invalidation point?
(CLOBBER_CACHE_ALWAYS testing would soon expose any fallacy there.)

			regards, tom lane