Re: Expending the use of xlog_internal.h's macros

From: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>
To: Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
Cc: PostgreSQL mailing lists <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Expending the use of xlog_internal.h's macros
Date: 2015-07-01 12:53:08
Message-ID: CAHGQGwES6q21T61UqxHPuO1ga76=pvXfebW0imeS1wYZNw-3SA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, Jul 1, 2015 at 8:58 PM, Michael Paquier
<michael(dot)paquier(at)gmail(dot)com> wrote:
> On Wed, Jul 1, 2015 at 8:18 PM, Fujii Masao wrote:
>> On Wed, Jul 1, 2015 at 8:16 PM, Fujii Masao wrote:
>>> I updated the patch as follows. Patch attached.
>>>
>>> +#define XLogFileNameExtended(fname, tli, log, seg)
>>>
>>> Move this macro to xlog_internal.h because it's used both in
>>> pg_standby and pg_archivecleanup. There seems no need to
>>> define it independently.
>
> OK for me.
>
>>> -#define MAXFNAMELEN 64
>>> +#define MAXFNAMELEN 64
>>>
>>> Revert this unnecessary change.
>
> Yes, thanks.
>
>>>
>>> +/* Length of XLog file name */
>>> +#define XLOG_DATA_FNAME_LEN 24
>>>
>>> Shorten the name of this macro variable, to XLOG_FNAME_LEN,
>>> for more code readability.
>
> Thanks. You have more talent for naming than I do.
>
>>> Comments?
>
> Just reading it again, I think that XLogFileNameById should use
> MAXFNAMELEN, and that XLogFileName should call directly
> XLogFileNameById as both are doing the same operation like in the
> attached.

We can refactor the code that way, but which looks a bit messy
at least to me. The original coding looks simpler and easier-readable,
so I'd like to adopt the original one here.

> It seems also safer to use MAXFNAMELEN instead of MAXPGPATH
> for exclusiveCleanupFileName in pg_standby.c and pg_archivecleanup.c.

Yep.

Regards,

--
Fujii Masao

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Fujii Masao 2015-07-01 14:15:19 Re: Support for N synchronous standby servers - take 2
Previous Message Michael Paquier 2015-07-01 12:37:55 Re: pg_basebackup and replication slots