Re: pgsql: Fix pgstattuple_approx handling of unknown tuple estimates

From: Melanie Plageman <melanieplageman(at)gmail(dot)com>
To: pgsql-committers(at)lists(dot)postgresql(dot)org
Subject: Re: pgsql: Fix pgstattuple_approx handling of unknown tuple estimates
Date: 2026-11-18 18:18:49
Message-ID: CAAKRu_bvcmBSUreBH2VkN8KS6MhBJbAaJfROO32QHL=Oj9t3Rg@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-committers

On Wed, Nov 18, 2026 at 10:43 AM Melanie Plageman
<melanieplageman(at)gmail(dot)com> wrote:
>
> Fix pgstattuple_approx handling of unknown tuple estimates
>
> vac_estimate_reltuples() can return -1 when there is not enough
> information to estimate the number of live tuples. pgstattuple_approx()
> assigned that value to an unsigned counter before attempting to clamp
> it, making the clamp ineffective and exposing -1 through the bigint
> result.
>
> Clamp the signed estimate before assigning it to the unsigned counter.

Looks like I broke the buildfarm with the new test

# @@ -257,11 +257,7 @@
# commit;
# select scanned_percent, approx_tuple_count
# from pgstattuple_approx('test_approx_unknown');
# - scanned_percent | approx_tuple_count
# ------------------+--------------------
# - 0 | 0
# -(1 row)
# -
# +ERROR: cannot access temporary tables during a parallel operation
# -- not for the index calls though, of course
# select pgstatindex('test_partition');
# ERROR: relation "test_partition" is not a btree index
# 1 of 1 tests failed.

Something about debug_parallel_query setting that didn't fail locally.
I will investigate and push a fix.

- Melanie

In response to

Browse pgsql-committers by date

  From Date Subject
Next Message Masahiko Sawada 2026-11-18 18:30:29 pgsql: Fall back to serializing when parallel apply runs out of DSM seg
Previous Message Tom Lane 2026-11-18 17:45:44 pgsql: Fix formatting of to_char's EEEE output for numeric and bigint.