Re: Why extra -I in "make test"?
by Michael G Schwern other posts by this author
Jul 2 2001 4:40AM messages near this date
Re: Why extra -I in "make test"?
|
Re: Why extra -I in "make test"?
On Sat, Jun 30, 2001 at 09:10:55AM -0700, Gurusamy Sarathy wrote:
> On Sat, 30 Jun 2001 11:34:29 EDT, schwern@[...].com wrote:
> >My only concern is that modules which live dual-lives on CPAN and the
> >core will have to edit their Makefile.PL every time they sync.
>
> INSTALLDIRS=perl is actually passed on the command line when building
> extensions. If you want to a PERL_CORE thingy for this, it should
> _not_ be hardwired into Makefile.PL. Edit ext/util/make_ext instead.
Ok, I'm trying to do this but I don't understand what's going on. A
few questions about how perl builds:
What is calling Makefile.PL for each of the ext/ directories?
What exactly am I supposed to do in make_ext to allow Makefile.PL to see
a PERL_CORE flag?
Basically, I'm changing ExtUtils::MM_Unix so it spits out a $RUNPERL
that it uses instead of all the hardwired $(PERL) -I$(PERL_ARCH)
-I$(PERL_LIB). If PERL_CORE is set, $RUNPERL will have the extra
-I's. Otherwise its just $PERL. I just can't figure out how to get
the ext/*/Makefile.PL's to see the PERL_CORE flag.
--
Michael G. Schwern <schwern@[...].com> http://www.pobox.com/~schwern/
Perl6 Quality Assurance <perl-qa@[...].org> Kwalitee Is Job One
BOFH excuse #331:
those damn racoons!
Thread:
Jarkko Hietaniemi
Jarkko Hietaniemi
Gurusamy Sarathy
Michael G Schwern
Gurusamy Sarathy
Michael G Schwern
Gurusamy Sarathy
Gurusamy Sarathy
Jarkko Hietaniemi
Nick Ing-Simmons
|