Date: Wed, 14 Mar 2018 12:43:52 -0300
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Peter Eisentraut <peter.eisentraut@2ndquadrant.com>,
	Stephen Frost <sfrost@snowman.net>,
	Magnus Hagander <magnus@hagander.net>,
	PostgreSQL WWW <pgsql-www@lists.postgresql.org>
Subject: Re: Bogus reports from coverage.postgresql.org
Message-ID: <20180314154352.ao4psdohgxd4pw65@alvherre.pgsql>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <27631.1521041578@sss.pgh.pa.us>
User-Agent: NeoMutt/20170306-137-4415bd-dirty (1.8.0)
Precedence: bulk

Tom Lane wrote:
> Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes:
> > On 3/13/18 13:45, Tom Lane wrote:
> >> Yeah.  I'd written that off as too low-probability to worry about,
> >> but maybe not.  I'll poke further.  Peter, which lcov did you test?
> 
> > lcov: LCOV version 1.13
> > The one that comes with Debian stable, presumably the same that is
> > running on coverage.p.o.
> 
> OK, now we're getting into the *really* low-probability cases.  Like,
> maybe it's time to remove and reinstall lcov and/or gcc on that box?

I checked on my own Debian box, which shows

$ gcov --version
gcov (Debian 6.3.0-18+deb9u1) 6.3.0 20170516

$ lcov --version
lcov: LCOV version 1.13

and what I get is:
     492           0 :     elog(ERROR, "predicate_classify returned a bogus value");
     803           0 :     elog(ERROR, "predicate_classify returned a bogus value");

so.

In the coverage.pg.org box, we have:

$ cat .lcovrc 
lcov_branch_coverage = 1

I'm going to do a test run with that disabled and see what happens.

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services