Re: Use virtual tuple slot for Unique node

On Thu, 12 Oct 2023 at 23:06, Ashutosh Bapat
<ashutosh(dot)bapat(dot)oss(at)gmail(dot)com> wrote:
> Q7 select distinct a,b from (select string_agg(left(a, 100), ', ')
> over (order by a rows 2 preceding) a, b from t_text) q
> HEAD: 16070.62 ms
> patched: 16182.16 ms

Did you time the SELECT or EXPLAIN ANALYZE?

With SELECT, I'm unable to recreate this slowdown. Using your setup:

$ cat bench.sql
set enable_hashagg=0;
set work_mem='10GB';
select distinct a,b from (select string_agg(left(a, 100), ', ') over
(order by a rows 2 preceding) a, b from t_text) q;

Master @ 13d00729d
$ pgbench -n -f bench.sql -T 300 postgres | grep latency
latency average = 7739.250 ms

Master + use_subnode_slot_type_for_nodeunique.patch
$ pgbench -n -f bench.sql -T 300 postgres | grep latency
latency average = 7718.007 ms

It's hard to imagine why there would be a slowdown as this query uses
a TTSOpsMinimalTuple slot type in the patch and the unpatched version.

David