| From: | Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com> |
|---|---|
| To: | "Andrey M(dot) Borodin" <x4mmm(at)yandex-team(dot)ru> |
| Cc: | "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: allow sorted builds for btree_gist |
| Date: | 2024-05-18 10:22:58 |
| Message-ID: | b949d6c0-2b33-4854-beca-2b73d5b408a2@enterprisedb.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 5/18/24 08:51, Andrey M. Borodin wrote:
>
>
>> On 18 May 2024, at 00:41, Tomas Vondra
>> <tomas(dot)vondra(at)enterprisedb(dot)com> wrote:
>>
>> if the opclass supports sorted builds, because then we could
>> parallelize the sort.
>
> Hi Tomas!
>
> Yup, I'd also be glad to see this feature. PostGIS folks are using
> their geometry (sortsupport was developed for this) with object id
> (this disables sort build).
>
> It was committed once [0], but then reverted, vardata opclasses were
> implemented wrong. Now it's on CF[1], Bernd is actively responding in
> the thread, but currently patch lacks tests.
>
> Thanks for raising this!
>
Oh, damn! I didn't notice the CF already has a patch doing this, and
that it was committed/reverted in 2021. I was completely oblivious to
that. Apologies.
Let's continue working on that patch/thread, I'll take a look in the
next CF.
regards
--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Alvaro Herrera | 2024-05-18 11:00:46 | Re: Speed up clean meson builds by ~25% |
| Previous Message | Jingxian Li | 2024-05-18 09:10:38 | Re: [PATCH] LockAcquireExtended improvement |