Re: PGXS "check" target forcing an install ?

On Tue, Jun 23, 2015 at 1:31 AM, Michael Paquier
<michael.paquier@gmail.com> wrote:
>> I tracked the dangerous -rf to come from Makefile.global and it's empty
>> string being due to abs_top_builddir not being define in my own Makefile.
>> But beside that, which I can probably fix, it doesn't sound correct
>> that a "check" rule insists in finding an "install" rule.
>
> Oops, this is a regression, and a dangerous one indeed. This is caused
> by dcae5fac.
>
> One fix is to use NO_TEMP_INSTALL=yes in Makefile.global in the
> context of PGXS, like in the patch attached, this variable needing to
> be set before Makefile.global is loaded. We could as well use directly
> PGXS in the section "Testing", but that does not sound appealing for
> Makefile.global's readability.

Gulp.  I certainly agree that emitting rm -rf /tmp_install is a scary
thing for a PostgreSQL Makefile to be doing.  Fortunately, people
aren't likely to have a directory under / by that name, and maybe not
permissions on it even if they did, but all the same it's not good.  I
propose trying to guard against that a bit more explicitly, as in the
attached.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company