Re: Performance regression with PostgreSQL 11 and partitioning

On Fri, May 25, 2018 at 1:53 PM, Amit Langote <amitlangote09@gmail.com> wrote:
> Seems here that we call find_appinfos_by_relids here for *all*
> partitions, even if all but one partition may have been pruned.  I
> haven't studied this code in detail, but I suspect it might be
> unnecessary, although I might be wrong.

Uggh.  It might be possible to skip it for dummy children.  That would
leave the dummy child rels generating a different pathtarget than the
non-dummy children, but I guess if we never use them again maybe it
wouldn't matter.

> Fwiw, I'm not sure why the new pruning code would call here, at least
> git grep find_appinfos_by_relids doesn't turn up anything interesting
> in that regard.

Hmm, OK.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company