Re: Extensible executor nodes for preparation of SQL/MED

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Itagaki Takahiro <itagaki(dot)takahiro(at)gmail(dot)com>
Cc: PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Extensible executor nodes for preparation of SQL/MED
Date: 2010-11-16 18:28:50
Message-ID: 2673.1289932130@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Itagaki Takahiro <itagaki(dot)takahiro(at)gmail(dot)com> writes:
> Here is a WIP patch to extensible executor nodes.

I am of the opinion that a run-time-extensible set of plan node types
is merest fantasy. We will never have that, so putting in place 5%
of the infrastructure for it is a waste of time and notational
complexity.

I would support replacing the switch in ExecProcNode with a
function-pointer call on performance grounds. The implementation of
that IMHO should match what's done for runtime expression state nodes:
set up the function pointer in the InitNode function that creates the
executor state node. But the rest of this patch, particularly the
addition of vtables to plan nodes, is pointless complication.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2010-11-16 18:34:55 Re: unlogged tables
Previous Message Heikki Linnakangas 2010-11-16 18:22:22 Re: GiST insert algorithm rewrite