Re: Explanation for bug #13908: hash joins are badly broken

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: Explanation for bug #13908: hash joins are badly broken
Date: 2016-02-06 21:18:43
Message-ID: 24074.1454793523@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com> writes:
> I believe the attached patch should fix this by actually copying the
> tuples into the densely allocated chunks. Haven't tested it though, will
> do in a few hours.

BTW, I confirmed that this patch fixes the wrong-number-of-output-tuples
issue in the test case from bug #13908. So that shows that the diagnosis
is correct. We still need to clean up the patch, but this way does work
to fix the problem.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2016-02-06 21:22:26 Re: Explanation for bug #13908: hash joins are badly broken
Previous Message Tomas Vondra 2016-02-06 21:17:52 Re: Explanation for bug #13908: hash joins are badly broken