Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 20 Oct 2005 04:39:30 -0400
From:      Kris Kennaway <kris@obsecurity.org>
To:        Volker Stolz <stolz@i2.informatik.rwth-aachen.de>
Cc:        cvs-ports@FreeBSD.org, ports-committers@FreeBSD.org, Emanuel Haupt <ehaupt@FreeBSD.org>, Kris Kennaway <kris@obsecurity.org>
Subject:   Re: cvs commit: ports/japanese/xdtp Makefile
Message-ID:  <20051020083930.GA24992@xor.obsecurity.org>
In-Reply-To: <20051020082453.GC73991@i2.informatik.rwth-aachen.de>
References:  <200510191912.j9JJCruT013822@repoman.freebsd.org> <20051019191435.GA26002@xor.obsecurity.org> <20051020080639.GB73991@i2.informatik.rwth-aachen.de> <20051020081305.GA22079@xor.obsecurity.org> <20051020082453.GC73991@i2.informatik.rwth-aachen.de>

next in thread | previous in thread | raw e-mail | index | archive | help

--lrZ03NoBR/3+SXJZ
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Oct 20, 2005 at 10:24:53AM +0200, Volker Stolz wrote:
> On Thu, Oct 20, 2005 at 04:13:05AM -0400, Kris Kennaway wrote:
> > > > >   . compile with -fPIC, this fixes build on amd64
> > > >=20
> > > > Please don't pessimize all architectures like this.  You shouldn't
> > > > even be pessimizing all files on amd64 with this flag, only those t=
hat
> > > > need them.
> > >=20
> > > Is this really such a strong pessimisation?
> > > http://www.ukuug.org/events/linux2002/papers/pdf/dsohowto.pdf
> > > "How To Write Shared Libraries" claims that this is almost a non-issu=
e.
> > > (Please rub my nose the archives I missed any discussion about the pr=
os/cons.)
> >=20
> > I don't know how much of an issue it is, but there's no reason not to
> > do it correctly.
>=20
> Increasing the number of conditionals in a port is another form of
> pessimisation in my book.
>
> Maybe this could be done globally for CFLAGS on amd64?

As I said, the correct thing to do is to patch the vendor sources to
only add -fPIC to those compilations that require it (shared library
objects).

Whether or not it's usually a significant runtime pessimization, in
FreeBSD we don't take the lazy way out.

Kris

--lrZ03NoBR/3+SXJZ
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (FreeBSD)

iD8DBQFDV1fCWry0BWjoQKURAl84AJ90glOeQwk7Wh47x2m+78CASNGAOgCgvqRT
bpaI4CS4h1S5hYGCQGUrHXM=
=wI9d
-----END PGP SIGNATURE-----

--lrZ03NoBR/3+SXJZ--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051020083930.GA24992>