Horton

Thread

Commits

  1. Make heap TID a tiebreaker nbtree index column.

  1. nbtree: reduce code redundancy by using _bt_getmeta() in _bt_gettrueroot()

    Chao Li <li.evan.chao@gmail.com> — 2026-01-15T08:10:42Z

    Hi Hackers,

I noticed this while reviewing [1]. _bt_gettrueroot() performs the same
metapage checks as _bt_getmeta(), so using _bt_getmeta() here removes the
duplicated code.

Build passed without warning and "make check" passed.

[1]
https://postgr.es/m/fvfmkr5kk4nyex56ejgxj3uzi63isfxovp2biecb4bspbjrze7@az2pljabhnff

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/

  2. Re: nbtree: reduce code redundancy by using _bt_getmeta() in _bt_gettrueroot()

    Neil Chen <carpenter.nail.cz@gmail.com> — 2026-01-16T08:09:03Z

    Hi Chao,

On Thu, Jan 15, 2026 at 4:11 PM Chao Li <li.evan.chao@gmail.com> wrote:

>
> I noticed this while reviewing [1]. _bt_gettrueroot() performs the same
> metapage checks as _bt_getmeta(), so using _bt_getmeta() here removes the
> duplicated code.
>
>
It looks good to me. These two code snippets are exactly identical,
so removing the redundant code is reasonable.

As can be seen from the contextual comments, this is not a
performance-critical path either,
so we don’t need to worry about the performance overhead caused by function
calls.

I also checked the changes made in commit dd299df, which essentially does
the same thing for the function _bt_getrootheight.

  3. Re: nbtree: reduce code redundancy by using _bt_getmeta() in _bt_gettrueroot()

    Heikki Linnakangas <hlinnaka@iki.fi> — 2026-07-07T21:32:21Z

    On 16/01/2026 10:09, Neil Chen wrote:
> Hi Chao,
> 
> On Thu, Jan 15, 2026 at 4:11 PM Chao Li <li.evan.chao@gmail.com 
> <mailto:li.evan.chao@gmail.com>> wrote:
> 
> 
>     I noticed this while reviewing [1]. _bt_gettrueroot() performs the
>     same metapage checks as _bt_getmeta(), so using _bt_getmeta() here
>     removes the duplicated code.
> 
> 
> It looks good to me. These two code snippets are exactly identical,
> so removing the redundant code is reasonable.
> 
> As can be seen from the contextual comments, this is not a performance- 
> critical path either,
> so we don’t need to worry about the performance overhead caused by 
> function calls.
> 
> I also checked the changes made in commit dd299df, which essentially 
> does the same thing for the function _bt_getrootheight.

Ok, committed.

- Heikki