Re: BUG #18055: logical decoding core on AllocateSnapshotBuilder()

On Sat, Sep 9, 2023 at 5:09 PM Amit Kapila <amit.kapila16@gmail.com> wrote:
>
> On Thu, Sep 7, 2023 at 5:40 PM Zhijie Hou (Fujitsu)
> <houzj.fnst@fujitsu.com> wrote:
> >
> > On Tuesday, September 5, 2023 7:31 PM Amit Kapila <amit.kapila16@gmail.com> wrote:
> > >
> > > On Fri, Aug 25, 2023 at 9:36 AM Zhijie Hou (Fujitsu) <houzj.fnst@fujitsu.com>
> > > wrote:
> > > >
> > > > On Wednesday, August 23, 2023 4:47 PM Amit Kapila
> > > <amit.kapila16@gmail.com> wrote:
> > > > >
> > > > > On Mon, Aug 21, 2023 at 11:57 PM Andres Freund <andres@anarazel.de>
> > > > > wrote:
> > > > > >
> > > > > > -1. This makes things global concerns that shouldn't be.
> > > > > >
> > > > > > If we really need something to clean this up, I'd look at
> > > > > > MemoryContextRegisterResetCallback().
> > > > > >
> > > > >
> > > > > +1. This sounds like a better idea.
> > > >
> > > > Here is a small patch based on this idea.
> > > > I registered the callback under snapshot builder context and the
> > > > callback will reset the variables.
> > > >
> > >
> > > LGTM. Unless Andres or Sawada-San thinks otherwise, I would like to proceed
> > > with this fix.
> >
> > Thanks, attach the patches for all support branches. I have confirmed that they can
> > pass the regression test and fix the issue.
> >
>
> I have also verified and it works for me as well. I'll push this (from
> PG11-PG15) next week by Tuesday unless there are more comments.
>

Pushed.

--
With Regards,
Amit Kapila.