Re: pgsql: Improve performance of timezone loading, especially pg_timezone_

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>
Cc: David Rowley <david(dot)rowley(at)2ndquadrant(dot)com>, pgsql-committers <pgsql-committers(at)postgresql(dot)org>
Subject: Re: pgsql: Improve performance of timezone loading, especially pg_timezone_
Date: 2017-05-07 15:17:08
Message-ID: 20119.1494170228@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> writes:
> On Sun, May 7, 2017 at 8:27 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> Ah-hah. Still, the log message should have been there before.

> No, it didn't exist before because dirname passed to scan_directory_ci
> didn't contain the filename appended to it.

Gotcha. On closer look, the "fall through" path added by this commit
wasn't really exercised at all on Unix machines, because calls with
no canonname only occur when we expect to find the zone file with name
as stated. It's accidental (and fortunate) that the bug in the MSVC
build support allowed us to find this bug promptly ;-)

I concur with your thought that these are separate bugs and deserve
separate commits. Will do that in a minute.

regards, tom lane

In response to

Responses

Browse pgsql-committers by date

  From Date Subject
Next Message Tom Lane 2017-05-07 15:35:30 pgsql: Restore fullname[] contents before falling through in pg_open_tz
Previous Message Amit Kapila 2017-05-07 15:03:36 Re: pgsql: Improve performance of timezone loading, especially pg_timezone_