Re: add_partial_path() may remove dominated path but still in use

From: Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>
To: robertmhaas(at)gmail(dot)com
Cc: kaigai(at)heterodb(dot)com, tgl(at)sss(dot)pgh(dot)pa(dot)us, pgsql-hackers(at)postgresql(dot)org
Subject: Re: add_partial_path() may remove dominated path but still in use
Date: 2019-01-17 09:28:48
Message-ID: 20190117.182848.186931872.horiguchi.kyotaro@lab.ntt.co.jp
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

Hello, sorry for the absence.

At Fri, 11 Jan 2019 11:36:43 -0500, Robert Haas <robertmhaas(at)gmail(dot)com> wrote in <CA+TgmoYyxBgkfN_APBdxdutFMukb=P-EgGNY-NbauRcL7mGnmA(at)mail(dot)gmail(dot)com>
> On Thu, Jan 10, 2019 at 9:10 PM Kohei KaiGai <kaigai(at)heterodb(dot)com> wrote:
> > 2019年1月11日(金) 5:52 Robert Haas <robertmhaas(at)gmail(dot)com>:
> > > On Wed, Jan 9, 2019 at 12:44 AM Kohei KaiGai <kaigai(at)heterodb(dot)com> wrote:
> > > > So, is it sufficient if set_rel_pathlist_hook is just relocated in
> > > > front of the generate_gather_paths?
> > > > If we have no use case for the second hook, here is little necessity
> > > > to have the post_rel_pathlist_hook() here.
> > > > (At least, PG-Strom will use the first hook only.)
> > >
> > > +1. That seems like the best way to be consistent with the principle
> > > that we need to have all the partial paths before generating any
> > > Gather paths.
> > >
> > Patch was updated, just for relocation of the set_rel_pathlist_hook.
> > Please check it.
>
> Seems reasonable to me.

Also seems reasonable to me. The extension can call
generate_gather_paths redundantly as is but it almost doesn't
harm, so it is acceptable even in a minor release.

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Fabien COELHO 2019-01-17 09:46:34 Re: PSA: we lack TAP test coverage on NetBSD and OpenBSD
Previous Message David Rowley 2019-01-17 09:08:40 Re: Delay locking partitions during query execution