Re: Missing hook for UPPERREL_PARTIAL_GROUP_AGG rels? - Mailing list pgsql-hackers

From Erik Nordström
Subject Re: Missing hook for UPPERREL_PARTIAL_GROUP_AGG rels?
Date
Msg-id CACAa4VJqiSWWxXzS7h5KGuhT2pEApHpFpJY6TQpRzrATEtgRMQ@mail.gmail.com
Whole thread Raw
In response to Re: Missing hook for UPPERREL_PARTIAL_GROUP_AGG rels?  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Thanks for the quick response. I do think this might be a separate issue because in the set_rel_pathlist_hook case, that hook is actually called at one point. In this case there is simply no place in the PostgreSQL code where a call is made to create_upper_paths_hook for the UPPERREL_PARTIAL_GROUP_AGG upper rel kind. See create_partial_grouping_paths().

Regards,

Erik

On Thu, Jun 27, 2019 at 8:58 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
Erik Nordström <erik@timescale.com> writes:
> I noticed that the create_upper_paths_hook is never called for partially
> grouped rels. Is this intentional or a bug? Only the FDW routine's
> GetForeignUpperPaths hook is called for partially grouped rels. This seems
> odd since the regular create_upper_paths_hook gets called for all other
> upper rels.

This seems possibly related to the discussion re set_rel_pathlist_hook
at

https://www.postgresql.org/message-id/flat/CADsUR0AaPx4sVgmnuVJ_bOkocccQZGubv6HajzW826rbSmFpCg%40mail.gmail.com

I don't think we've quite resolved what to do there, but maybe
create_upper_paths_hook needs to be looked at at the same time.

                        regards, tom lane

pgsql-hackers by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: Hypothetical indexes using BRIN broken since pg10
Next
From: Tom Lane
Date:
Subject: Re: [HACKERS] Regression tests vs existing users in an installation