Re: "make check" changes have caused buildfarm deterioration.

On 07/24/2015 01:21 PM, Jeff Janes wrote:
> On Tue, Jul 21, 2015 at 6:31 PM, Michael Paquier
> <michael(dot)paquier(at)gmail(dot)com <mailto:michael(dot)paquier(at)gmail(dot)com>> wrote:
>
> On Wed, Jul 22, 2015 at 10:04 AM, Peter Eisentraut
> <peter_e(at)gmx(dot)net <mailto:peter_e(at)gmx(dot)net>> wrote:
> > On 7/21/15 10:00 AM, Tom Lane wrote:
> >> I agree; this change may have seemed like a good idea at the
> time, but
> >> it was not. Failures during "make check"'s install step are
> rare enough
> >> that you don't really need all that output in your face to help
> with the
> >> rare situation where it fails. And for the buildfarm's
> purposes, it is
> >> surely desirable to segregate that output from the actual check
> step.
> >
> > It wasn't really an idea; it was just not necessary anymore. We
> can put
> > it [directing the make install output into a file] back if
> that's what
> > people prefer.
>
> OK... Attached are two patches (please merge them into a single
> commit, I am just separating them as they are separate issues):
> - 0001 adds a missing entry in test_ddl_deparse's .gitignore. I
> mentioned that upthread.
> - 0002 redirects the installation logs into
> abs_top_builddir/tmp_install/log/install.log. We could redirect it
> only to abs_top_builddir/log/ but tmp_install is not removed after a
> run of a regression make target.
>
>
> If I run 'make check' on an unbuilt tree, any compiler warnings
> emitted during the build phase now get directed into the install log.
> Was that intentional or a side effect?
>
>

Probably not, but you could get around it easily by doing "make && make
check". Getting around the previous behavior was not nearly so easy.

cheers

andrew