On Sat, 8 May 2021 at 20:29, David Rowley <dgrowle...@gmail.com> wrote:
>
> On Sat, 8 May 2021 at 20:17, Zhihong Yu <z...@yugabyte.com> wrote:
> > +       if (!OidIsValid(saop->negfuncid))
> > +           record_plan_function_dependency(root, saop->hashfuncid);
> >
> > Is there a typo in the second line ? (root, saop->negfuncid)
>
> Yeah, that's a mistake. Thanks for checking it.

I've attached a patch which fixes the mistake mentioned above.

Also, dropped the RMT from the thread. I only introduced them when I
wanted some input about if hashing NOT IN should be included in PG14.
Nobody seems to think that should be done.

David

Attachment: v4-0001-Speedup-NOT-IN-with-a-set-of-Consts.patch
Description: Binary data

Reply via email to