Re: Fix of fake unlogged LSN initialization

From: Michael Paquier <michael(at)paquier(dot)xyz>
To: "tsunakawa(dot)takay(at)fujitsu(dot)com" <tsunakawa(dot)takay(at)fujitsu(dot)com>
Cc: 'Simon Riggs' <simon(at)2ndquadrant(dot)com>, "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Fix of fake unlogged LSN initialization
Date: 2019-10-25 06:58:14
Message-ID: 20191025065814.GH8671@paquier.xyz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Fri, Oct 25, 2019 at 02:07:04AM +0000, tsunakawa(dot)takay(at)fujitsu(dot)com wrote:
> From: Simon Riggs <simon(at)2ndquadrant(dot)com>
>> From xlogdefs.h added by 9155580:
>> /*
>> * First LSN to use for "fake" LSNs.
>> *
>> * Values smaller than this can be used for special per-AM purposes.
>> */
>> #define FirstNormalUnloggedLSN ((XLogRecPtr) 1000)
>
> Yeah, I had seen it, but I didn't understand what kind of usage is assumed.

There is an explanation in the commit message of 9155580: that's to
make an interlocking logic in GiST able to work where a valid LSN
needs to be used. So a magic value was just wanted.

Your patch looks fine to me by the way after a second look, so I think
that we had better commit it and back-patch sooner than later. If
there are any objections or more comments, please feel free..
--
Michael

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Dongming Liu 2019-10-25 07:18:34 Problem with synchronous replication
Previous Message Arseny Sher 2019-10-25 06:56:23 Re: ERROR: subtransaction logged without previous top-level txn record