Re: FETCH FIRST clause WITH TIES option

Tomas Vondra <tomas.vondra@2ndquadrant.com>

From: Tomas Vondra <tomas.vondra@2ndquadrant.com>
To: Surafel Temesgen <surafel3000@gmail.com>
Cc: David Steele <david@pgmasters.net>, Michael Paquier <michael@paquier.xyz>, Robert Haas <robertmhaas@gmail.com>, andrew@tao11.riddles.org.uk, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2019-04-03T19:02:57Z
Lists: pgsql-hackers
Hi,

Unfortunately this got broken again, this time by aef65db676 :-(

I've tried to fix the merge conflict (essentially by moving some of the
code to adjust_limit_rows_costs(), but I'm wondering if the code added to
create_limit_path is actually correct

    if (count_est != 0)
    {
        double        count_rows;

        if (count_est > 0)
            count_rows = (double) count_est;
        else
            count_rows = clamp_row_est(subpath->rows * 0.10);

        if (limitOption == WITH_TIES)
        {
            ...
            count_rows = Max(avgGroupSize, count_est + (...));
        }
        ...
    }

Firstly, this seriously needs some comment explaining why we do this. But
more importantly - shouldn't it really be

    count_rows = Max(avgGroupSize, count_rows + (...));

instead of using count_est again (which might easily be -1 anyway)?


regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services




Commits

  1. Support FETCH FIRST WITH TIES