Re: pgsql: Blind attempt at fixing the non-MSVC Windows builds

From: Andres Freund <andres(at)2ndquadrant(dot)com>
To: Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
Cc: pgsql-committers(at)postgresql(dot)org
Subject: Re: pgsql: Blind attempt at fixing the non-MSVC Windows builds
Date: 2013-02-22 15:38:28
Message-ID: 20130222153827.GH19114@awork2.anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

On 2013-02-22 12:34:19 -0300, Alvaro Herrera wrote:
> Andres Freund wrote:
> > On 2013-02-22 14:55:06 +0000, Alvaro Herrera wrote:
> > > Blind attempt at fixing the non-MSVC Windows builds
> > >
> > > Apparently, they need -DBUILDING_DLL for the Assert() declarations to
> > > work correctly.
> >
> > More specifically, they need that do access any PGDLLIMPORT symbols. Of
> > which
> > extern PGDLLIMPORT bool assert_enabled;
> > is one. So that seems like the appropriate fix.
>
> Yeah, I'm not really sure if this is okay for both build environments or
> is only needed in the backend. Right now there should be no difference
> because we're not using relpath.c in frontend programs; but if we add
> pg_xlogdump there might be further problems.

I don't think it will matter for relpath itself because that doesn't
seem to access anything but assert_enabled - and that only in backend
builds.
But if we add further stuff which really might access 'extern'
variables, it should also be needed in frontend programs unless I miss
something.

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

In response to

Browse pgsql-committers by date

  From Date Subject
Next Message Tom Lane 2013-02-22 15:56:54 pgsql: Fix some planning oversights in postgres_fdw.
Previous Message Alvaro Herrera 2013-02-22 15:34:19 Re: pgsql: Blind attempt at fixing the non-MSVC Windows builds