Re: pgsql: Improve support for building PGXS modules with VPATH.

From: Cédric Villemain <cedric(at)2ndquadrant(dot)com>
To: pgsql-committers(at)postgresql(dot)org
Cc: Andrew Dunstan <andrew(at)dunslane(dot)net>, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
Subject: Re: pgsql: Improve support for building PGXS modules with VPATH.
Date: 2013-07-01 21:13:55
Message-ID: 201307012313.59438.cedric@2ndquadrant.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Le lundi 1 juillet 2013 21:22:55, Andrew Dunstan a écrit :
> On 07/01/2013 02:40 PM, Andrew Dunstan wrote:
> > On 07/01/2013 01:39 PM, Alvaro Herrera wrote:
> >> Andrew Dunstan wrote:
> >>> Improve support for building PGXS modules with VPATH.
> >>>
> >>> A VPATH build will be performed when the module's make file path is not
> >>> the current directory or when USE_VPATH is set.
> >>
> >> So what's the USE_VPATH thing for? I mean, can't you just use VPATH
> >> instead?
> >
> > Well, I was distinguishing between things set by the Makefile and
> > things set outisde, but you're probably right. I'll test it and tweak.
>
> Actually, when I try to get rid of it I get issues with srcdir being
> recursively defined.

As for VPATH, the srcdir was previously hard-coded, the patch just improve
most of the use cases by setting a more appropriate value..

> And in fact, VPATH is defined by Makefile.global when the Postgres build
> is a VPATH build, and unrelated to the module's build setup.

pgxs.mk already (I mean before the last commit) set VPATH, the patch around
that just try to fix it to be correct instead of hardcoded to NULL.. What's in
Makefile.global is not be affected.

--
Cédric Villemain +33 (0)6 20 30 22 52
http://2ndQuadrant.fr/
PostgreSQL: Support 24x7 - Développement, Expertise et Formation

In response to

Responses

Browse pgsql-committers by date

  From Date Subject
Next Message Peter Eisentraut 2013-07-02 01:13:42 pgsql: Fix regression test make dependencies
Previous Message Simon Riggs 2013-07-01 19:46:12 Re: pgsql: Add a convenience routine makeFuncCall to reduce duplication.