Re: Win32 regression test status

[resurrecting this thread from the dead]

Tom Lane wrote:
> It's not mandated but we need to know why this platform acts
> differently from the rest. The join failures look like it may be an issue

> of the qsort() implementation acting differently for equal keys than most
do.
> Not sure whether the same applies to rules.

Any "off the top of the head" ideas on how to confirm/test this? Or how to
proceed otherwise?

Now that the code to get stats working under win32 has been submitted, I'd
like to round these two failures out (not to mention looking into replacing
localtime et al with tzcode/zic), before cleaning up the remaining items for
win32.

Cheers,
Claudio

---
Certain disclaimers and policies apply to all email sent from Memetrics.
For the full text of these disclaimers and policies see
<a
href="http://www.memetrics.com/emailpolicy.html">http://www.memetrics.com/em
ailpolicy.html</a>

Claudio Natoli wrote:
>
> [resurrecting this thread from the dead]
>
> Tom Lane wrote:
> > It's not mandated but we need to know why this platform acts
> > differently from the rest. The join failures look like it may be an issue
>
> > of the qsort() implementation acting differently for equal keys than most
> do.
> > Not sure whether the same applies to rules.
>
> Any "off the top of the head" ideas on how to confirm/test this? Or how to
> proceed otherwise?

Uh, I think you just add ORDER BY to the query and see if the output now
matches the regression test. If so, submit a patch to add the ORDER BY.

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073

Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> Uh, I think you just add ORDER BY to the query and see if the output now
> matches the regression test. If so, submit a patch to add the ORDER BY.

I'd really prefer not to add ORDER BY to all the join tests, as that
would restrict the set of join plan types being tested.

regards, tom lane

Tom Lane wrote:
> Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> > Uh, I think you just add ORDER BY to the query and see if the output now
> > matches the regression test. If so, submit a patch to add the ORDER BY.
>
> I'd really prefer not to add ORDER BY to all the join tests, as that
> would restrict the set of join plan types being tested.

Agreed, we don't add them to all joins, but don't we historically add
ORDER BY to queries that are shown to be sensitive to qsort
implementations on various platforms? Are you suggesting we create a
separate regression output file for that platform?

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073

Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> Agreed, we don't add them to all joins, but don't we historically add
> ORDER BY to queries that are shown to be sensitive to qsort
> implementations on various platforms? Are you suggesting we create a
> separate regression output file for that platform?

I'm not suggesting any particular solution yet, just pointing out that
there are reasons to look for alternatives rather than automatically
adding ORDER BY.

regards, tom lane