On 4/12/21 7:04 PM, Tom Lane wrote:
> Ranier Vilela <ranier.vf@gmail.com> writes:
>> Em seg., 12 de abr. de 2021 às 03:04, Tom Lane <tgl@sss.pgh.pa.us> escreveu:
>>> It would be wrong, though, or at least not have the same effect.
>
>> I think that you speak about fill pointers with 0 is not the same as fill
>> pointers with NULL.
>
> No, I mean that InvalidBlockNumber isn't 0.
>
>> I was confused here, does the patch follow the pattern and fix the problem
>> or not?
>
> Your patch seems fine. Justin's proposed improvement isn't.
>
Pushed.
> (I'm not real sure whether there's any *actual* bug here --- would we
> really be looking at either ctid or tableoid of this temporary tuple?
> But it's probably best to ensure that they're valid anyway.)>
Yeah, the tuple is only built so that we can pass it to the various
selectivity estimators. I don't think anything will be actually looking
at those fields, but initializing them seems easy enough.
regards
--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
