Re: [PATCH] Improve geometric types

From: Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>
To: Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>
Cc: emre(at)hasegeli(dot)com, thomas(dot)munro(at)enterprisedb(dot)com, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: [PATCH] Improve geometric types
Date: 2018-07-29 10:35:26
Message-ID: def5855d-9864-1d65-01af-bf06ee5b2d61@2ndquadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 07/29/2018 03:54 AM, Tomas Vondra wrote:
>
>
> On 07/27/2018 10:20 AM, Kyotaro HORIGUCHI wrote:
>> Thank you for taking this.
>>
>> At Thu, 26 Jul 2018 17:12:50 +0200, Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com> wrote in <672f4c42-6742-c1ec-e9a4-1994b815acc7(at)2ndquadrant(dot)com>
>>> On 07/11/2018 07:13 PM, Emre Hasegeli wrote:
>>>> New versions are attached after the <float.h> patch got in. I noticed
>>>> tests failing on Windows [1] and added alternative .out file.
>>>> [1]
>>>> https://ci.appveyor.com/project/postgresql-cfbot/postgresql/build/1.0.5235
>>>>
>>>
>>> The version posted about two weeks ago is slightly broken - AFAICS the
>>> float.h includes in geo_ops.c and gistproc.c need to be part of 0002,
>>> not 0003. Attached is a version fixing that.
>>>
>>> Barring objections, I'll get this committed over the next few days,
>>> once I review all the individual parts once more. I'm planning to
>>> commit the 6 parts separately, as submitted. No backpatching, as
>>> discussed before.
>>
>> +1 for no backpatching, and not merging into single big patch.
>>
>
> I've committed the first two parts, after a review and testing.
>
> As these two parts were primarily refactoring (and quite extensive),
> this seems like a good place to wait if the buildfarm is happy with it.
> If yes, I'll continue applying the patches that do fix/change the
> behavior in various ways.
>

Seems there's a couple of failures like this, so far:

*** 42,48 ****
s
---------------------------------------------
{1,-1,1}
! {1,-1,0}
{-0.4,-1,-6}
{-0.000184615384615385,-1,15.3846153846154}
{1,-1,11}
--- 42,48 ----
s
---------------------------------------------
{1,-1,1}
! {1,-1,-0}
{-0.4,-1,-6}
{-0.000184615384615385,-1,15.3846153846154}
{1,-1,11}
***************

It's always 0/-0 difference, and it's limited to power machines. I'll
try to get access to such system and see what's wrong.

regards

--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Thomas Munro 2018-07-29 10:42:32 Re: Removing unneeded self joins
Previous Message Thomas Munro 2018-07-29 10:11:17 Re: [HACKERS] WIP: long transactions on hot standby feedback replica / proof of concept