Horton

Re: Bug in nbtree optimization to skip > operator comparisons (or < comparisons in backwards scans)

Alexander Korotkov <aekorotkov@gmail.com>

View thread View on postgr.es
From: Alexander Korotkov <aekorotkov@gmail.com>
To: Peter Geoghegan <pg@bowt.ie>
Cc: Heikki Linnakangas <hlinnaka@iki.fi>, Matthias van de Meent <boekewurm+postgres@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2023-12-24T22:50:52Z
Lists: pgsql-hackers

Attachments

On Tue, Dec 12, 2023 at 3:22 PM Alexander Korotkov <aekorotkov@gmail.com> wrote:
>
> On Mon, Dec 11, 2023 at 6:16 PM Peter Geoghegan <pg@bowt.ie> wrote:
> > Will you be in Prague this week? If not this might have to wait.
>
> Sorry, I wouldn't be in Prague this week.  Due to my current
> immigration status, I can't travel.
> I wish you to have a lovely time in Prague.  I'm OK to wait, review
> once you can.  I will probably provide a more polished version
> meanwhile.

Please find the revised patchset attached.  It comes with revised
comments and commit messages.  Besides bug fixing the second patch
makes optimization easier to understand.  Now the flag used for
skipping checks of same direction required keys is named
continuescanPrechecked and means exactly that *continuescan flag is
known to be true for the last item on the page.

Any objections to pushing these two patches?

------
Regards,
Alexander Korotkov

Commits

  1. Fix an oversight in refactoring in 06b10f80ba4.

  2. Make RangeTblEntry dump order consistent

  3. Improvements and fixes for e0b1ee17dc

  4. Remove BTScanOpaqueData.firstPage

  5. Optimize nbtree backward scan boundary cases.

  6. Skip checking of scan keys required for directional scan in B-tree