Re: Parallel Aggregate

On Wed, Mar 16, 2016 at 6:49 AM, David Rowley
<david.rowley@2ndquadrant.com> wrote:
> On 16 March 2016 at 15:04, Robert Haas <robertmhaas@gmail.com> wrote:
>> I don't think I'd be objecting if you made PartialAggref a real
>> alternative to Aggref.  But that's not what you've got here.  A
>> PartialAggref is just a wrapper around an underlying Aggref that
>> changes the interpretation of it - and I think that's not a good idea.
>> If you want to have Aggref and PartialAggref as truly parallel node
>> types, that seems cool, and possibly better than what you've got here
>> now.  Alternatively, Aggref can do everything.  But I don't think we
>> should go with this wrapper concept.
>
> Ok, I've now gotten rid of the PartialAggref node, and I'm actually
> quite happy with how it turned out. I made
> search_indexed_tlist_for_partial_aggref() to follow-on the series of
> other search_indexed_tlist_for_* functions and have made it behave the
> same way, by returning the newly created Var instead of doing that in
> fix_combine_agg_expr_mutator(), as the last version did.
>
> Thanks for the suggestion.
>
> New patch attached.

Cool!  Why not initialize aggpartialtype always?

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