Re: pgsql: Fix broken multibyte regression tests.

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Tatsuo Ishii <ishii(at)postgresql(dot)org>
Cc: pgsql-committers(at)postgresql(dot)org
Subject: Re: pgsql: Fix broken multibyte regression tests.
Date: 2015-11-28 18:55:09
Message-ID: 27771.1448736909@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Tatsuo Ishii <ishii(at)postgresql(dot)org> writes:
> Fix broken multibyte regression tests.
> commit 9043Fe390f4f0b4586cfe59cbd22314b9c3e2957 broke multibyte
> regression tests because the commit removes the warning message when
> temporary hash indexes is created, which has been added by commit
> 07af523870bcfe930134054febd3a6a114942e5b.

> Back patched to 9.5 stable tree.

AFAICT this patch was incorrect and should be reverted, because the
src/test/mb tests all fail for me, in both HEAD and 9.5, as a consequence
of getting hash-index WARNINGs that are not in the expected-files. The
commit you mention (which is 9043ef390f4f0b4586cfe59cbd22314b9c3e2957 not
what's cited in this commit message) disabled the WARNING for temporary
and unlogged hash indexes, but the indexes created by the test scripts
are neither.

Please recheck it.

regards, tom lane

In response to

Browse pgsql-committers by date

  From Date Subject
Next Message Tom Lane 2015-11-29 23:18:54 pgsql: Avoid caching expression state trees for domain constraints acro
Previous Message Tom Lane 2015-11-28 18:42:41 pgsql: Avoid doing encoding conversions by double-conversion via MULE_I