On Sat, Dec 20, 2025 at 7:32 AM Kirill Reshke <reshkekirill@gmail.com> wrote:
>
> Hi! I checked v29-0009, about HeapTupleSatisfiesVacuumHorizon. Origins
> of this code track down to fdf9e21196a6 which was committed as part of
> [0], at which point
> there was no HeapTupleSatisfiesVacuumHorizon function. I guess this is
> the reason this optimization was not performed earlier.
Thanks for taking a look into this!
> I also think this patch is correct, because we do similar things for
> HEAPTUPLE_DEAD & HEAPTUPLE_RECENTLY_DEAD, and
> HeapTupleSatisfiesVacuumHorizon is just a proxy to
> HeapTupleSatisfiesVacuumHorizon with only difference in DEAD VS
> RECENTLY_DEAD handling.
>
> Similar change could be done at heapam_scan_analyze_next_tuple
>
> ...
> case HEAPTUPLE_DEAD:
> case HEAPTUPLE_RECENTLY_DEAD:
> /* Count dead and recently-dead rows */
> *deadrows += 1;
> break;
In v30 sent here [1], I did end up making this change in 0010. I just
realized that I should have also changed
table_scan_analyze_next_tuple() and removed the call to
GetOldestRemovableTransactionId(). I've done that in attached v31.
I'm not sure we should change the table AM API (by removing
OldestXmin), though. I looked for table AMs implementing
scan_analyze_next_tuple() to see if they use OldestXmin. I found two:
OrioleDB [2] and Citus columnar [3], which both implement
scan_analyze_next_tuple() and neither of them use OldestXmin. I
couldn't easily find other table AMs implementing
scan_analyze_next_tuple(). I don't have a strong sense of whether or
not I should make this change. Changing it is churn to a public API
and doesn't specifically enable us to do something.
I could also just leave it unused by heapam's implementation. I
haven't checked what, if any, other table AMs callbacks have
parameters completely unused by their heap implementation.
So, I'm on the fence about whether or not to make the change at all,
and, if I do, whether or not to change the table AM callback. That is
done in v31, though, so we can discuss.
- Melanie
[1] https://www.postgresql.org/message-id/CAAKRu_ZCjHoRPfQ8AbMrFY8TOMCPAvZ0_m9SX7yg0edfTk45-g%40mail.gmail.com
[2]
https://github.com/orioledb/orioledb/blob/acff65984d106dabf708a179e2c6694297e08c02/src/tableam/handler.c#L978C68-L978C78
[3]
https://github.com/citusdata/citus/blob/ee3812d267db3ab007efb6f5f432c82c1f448695/src/backend/columnar/columnar_tableam.c#L1418
