Horton

Re: Crash in BRIN minmax-multi indexes

Zhihong Yu <zyu@yugabyte.com>

View thread View on postgr.es
From: Zhihong Yu <zyu@yugabyte.com>
To: Jaime Casanova <jcasanov@systemguards.com.ec>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Tomas Vondra <tomas.vondra@enterprisedb.com>
Date: 2021-03-31T18:20:20Z
Lists: pgsql-hackers 
Hi,
In build_distances():

        a1 = eranges[i].maxval;
        a2 = eranges[i + 1].minval;

It seems there was overlap between the successive ranges, leading to
delta = -6785000000

FYI

On Wed, Mar 31, 2021 at 10:30 AM Jaime Casanova <
jcasanov@systemguards.com.ec> wrote:

> Hi,
>
> Just found $SUBJECT involving time with time zone and a subselect. I
> still don't have narrowed to the exact table/index minimal schema but
> if you run this query on the regression database it will creash.
>
> ```
> update public.brintest_multi set
>   timetzcol = (select tz from generate_series('2021-01-01'::timestamp
> with time zone, '2021-01-31', '5 days') tz limit 1)
> ;
> ```
>
> attached a backtrace. Let me know if you need extra information.
>
> --
> Jaime Casanova
> Director de Servicios Profesionales
> SYSTEMGUARDS - Consultores de PostgreSQL
>

Commits

  1. Fix assert in BRIN build_distances

  2. Fix bug in brin_minmax_multi_union

  3. Fix order of parameters in BRIN minmax-multi calls

  4. Fix BRIN minmax-multi distance for inet type

  5. Fix BRIN minmax-multi distance for timetz type

  6. Fix BRIN minmax-multi distance for interval type