Message-ID: <d2b460a9c28b8931c35a165d32717ff736dd1a68.camel@j-davis.com>
Subject: Re: Minimal logical decoding on standbys
From: Jeff Davis <pgsql@j-davis.com>
To: Andres Freund <andres@anarazel.de>
Cc: "Drouvot, Bertrand" <bertranddrouvot.pg@gmail.com>, Robert Haas
 <robertmhaas@gmail.com>, Thomas Munro <thomas.munro@gmail.com>, Alvaro
 Herrera <alvherre@2ndquadrant.com>, Ibrar Ahmed <ibrar.ahmad@gmail.com>,
 Amit Khandekar <amitdkhan.pg@gmail.com>, fabriziomello@gmail.com, tushar
 <tushar.ahuja@enterprisedb.com>, Rahila Syed <rahila.syed@2ndquadrant.com>,
  pgsql-hackers <pgsql-hackers@postgresql.org>
Date: Sun, 02 Apr 2023 16:04:42 -0700
In-Reply-To: <20230402222904.d3va24dwpr3mgbpj@awork3.anarazel.de>
References: <061eeb26-156e-6011-02b6-b4b1422e2aa6@gmail.com>
	 <ec744999afcdbd6bec5d9e6572aba875060019cb.camel@j-davis.com>
	 <8a7da0c2-b7fe-1c78-4e47-a16239bf4839@gmail.com>
	 <d29e4458e7406b276c50197d9f3b31a5247fe915.camel@j-davis.com>
	 <6ab352de1553c1e3d0c7d97c038dc470ca51e772.camel@j-davis.com>
	 <4df984c8-3adb-3b69-0e4e-a491097862ce@gmail.com>
	 <c188da88-a3d5-15c5-b6b5-f925abccc873@gmail.com>
	 <bda73929475de2cb791a69f88d5d1e373e449478.camel@j-davis.com>
	 <20230402213506.ocprj6ky45s4ydtr@awork3.anarazel.de>
	 <d8f1f66dfe845a7d18eba7b6e5a27b2f1a887b57.camel@j-davis.com>
	 <20230402222904.d3va24dwpr3mgbpj@awork3.anarazel.de>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu1 
MIME-Version: 1.0
Archived-At: <https://www.postgresql.org/message-id/d2b460a9c28b8931c35a165d32717ff736dd1a68.camel%40j-davis.com>
Precedence: bulk

On Sun, 2023-04-02 at 15:29 -0700, Andres Freund wrote:
> I agree that the *wait* has to go through condition_variable.c, but
> it doesn't
> seem right that creation of the WES needs to go through
> condition_variable.c.

The kind of WES required by a CV is an implementation detail, so I was
concerned about making too many assumptions across different APIs.

But what I ended up with is arguably not better, so perhaps I should do
it your way and then just have some comments about what assumptions are
being made?

> The only thing that ConditionVariableEventSleep() seems to require is
> that the
> WES is waiting for MyLatch. You don't even need a separate WES for
> that, the
> already existing WES should suffice.

By "already existing" WES, I assume you mean FeBeWaitSet? Yes, that
mostly matches, but it uses WL_POSTMASTER_DEATH instead of
WL_EXIT_ON_PM_DEATH, so I'd need to handle PM death in
condition_variable.c. That's trivial to do, though.

Regards,
	Jeff Davis