Re: PATCH: Unlogged tables re-initialization tests

From: Thomas Munro <thomas(dot)munro(at)enterprisedb(dot)com>
To: David Steele <david(at)pgmasters(dot)net>
Cc: Pg Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: PATCH: Unlogged tables re-initialization tests
Date: 2018-03-01 05:27:19
Message-ID: CAEepm=1o=Dqvhad1KX6b==O3Z3jU4X3ZO+fOYbxN_s2WW4G7WA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, Mar 1, 2018 at 9:24 AM, David Steele <david(at)pgmasters(dot)net> wrote:
> These tests were originally included in the exclude unlogged tables
> patch [1] to provide coverage for the refactoring of reinit.c.

Hi David,

+# The following tests test symlinks. Windows doesn't have symlinks, so
+# skip on Windows.

Could you please explain this a bit more? I don't see any symlinks
being used directly in this test. I do see CREATE TABLESPACE and that
uses symlink(), but win32_port.h converts that to "junctions",
apparently the Windows equivalent. Is there some reason that doesn't
work with this test?

If, by any chance, you are talking about whatever dark force prevents
the "tablespace" regression test from succeeding when run as a user
with administrative privileges on Windows, then I would *love* to hear
an explanation for that phenomenon and how to fix it, because it
currently prevents me from automatically testing all Commitfest
patches on the appveyor.com Windows build farm. I know that it passes
as a non-administrative user.

--
Thomas Munro
http://www.enterprisedb.com

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Arseny Sher 2018-03-01 05:39:34 Two-phase update of restart_lsn in LogicalConfirmReceivedLocation
Previous Message Andres Freund 2018-03-01 05:26:26 Re: Why chain of snapshots is used in ReorderBufferCommit?