From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <kostja@tarantool.org>
Date: Fri, 21 Jun 2019 10:26:24 +0300
From: Konstantin Osipov <kostja@tarantool.org>
Subject: Re: [tarantool-patches] Re: [PATCH v4 6/9] wal: introduce a journal
 entry finalization callback
Message-ID: <20190621072624.GH18958@atlas>
References: <cover.1560978655.git.georgy@tarantool.org>
 <079a143aa17e185c44c5fe9c53ef207d7988fe56.1560978655.git.georgy@tarantool.org>
 <20190620140848.77crsufqebxhf4q2@esperanza>
 <3009341.sCQpP4uagq@home.lan>
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <3009341.sCQpP4uagq@home.lan>
To: tarantool-patches@freelists.org
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
List-ID: <tarantool-patches.dev.tarantool.org>

* Георгий Кириченко <georgy@tarantool.org> [19/06/20 23:23]:
> > > +	/*
> > > +	 * A trigger could be processed by the wal scheduler fiber
> > > +	 * so steal the latch first.
> > > +	 */
> > 
> > Not "could be", but "is processed", I guess.
> There are two cases: for non-yielding journal (wal none or recovery) it would 
> be the same fiber but for yielding one - this would be a tx_prio callback.

I think it's OK to mention both in the comment.


-- 
Konstantin Osipov, Moscow, Russia