Re: [HACKERS] path toward faster partition pruning

On 2018/02/22 20:28, David Rowley wrote:
> On 22 February 2018 at 22:48, Amit Langote
> <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp> wrote:
>>> I'm having to add some NULL handling there for the run-time pruning
>>> patch but wondered if it was also required for your patch.
>>
>> Hmm, not sure why. Can you explain a bit more?
>
> hmm, yeah, but perhaps we should be discussing on the other thread...
>
> With a prepared statement the Param will be unavailable until
> execution, in which case we don't do the const folding.

Ah right.

> A simple case is:
>
> create table listp (a int) partition by list (a);
> create table listp1 partition of listp for values in(1);
> prepare q1 (int) as select * from listp where a = $1;
> explain analyze execute q1(1); -- repeat 5 times.
> explain analyze execute q1(null); -- partkey_datum_from_expr() gets a
> NULL param via the call from nodeAppend.c

I wonder if NULLs should somehow be managed at a higher level, resulting
in the same behavior as const-folding in the optimizer produces? In any
case, I suppose that would be something for the run-time pruning patch to
handle.

Thanks,
Amit