Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 9 Oct 2005 01:10:34 -0400
From:      Kris Kennaway <kris@obsecurity.org>
To:        ray@redshift.com
Cc:        freebsd-amd64@freebsd.org, Kris Kennaway <kris@obsecurity.org>
Subject:   Re: /etc/make.conf
Message-ID:  <20051009051034.GB93028@xor.obsecurity.org>
In-Reply-To: <3.0.1.32.20051008220300.00a74ba8@pop.redshift.com>
References:  <3.0.1.32.20051008215252.00a74ba8@pop.redshift.com> <3.0.1.32.20051008202606.00a74ba8@pop.redshift.com> <3.0.1.32.20051008202606.00a74ba8@pop.redshift.com> <3.0.1.32.20051008215252.00a74ba8@pop.redshift.com> <3.0.1.32.20051008220300.00a74ba8@pop.redshift.com>

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

--OwLcNYc0lM97+oe1
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Sat, Oct 08, 2005 at 10:03:00PM -0700, ray@redshift.com wrote:

> I've been running those flags for years - if you would like to supply some logic
> as to why they won't work on the AMD platform, then maybe that would help.  If
> your advice is just "don't do that, it's bad", then doesn't help much.  -O2 has
> always worked fine for our production servers.  What flag specifically are you
> nervous about?

-funroll-loops is kind of stupid to use globally.  Chances are it's
hurting your performance more than it's helping.  -O2 was also unsafe
prior to the 5.x branch, but it should be OK now post-5.x (where it's
the default).

Kris

--OwLcNYc0lM97+oe1
Content-Type: application/pgp-signature
Content-Disposition: inline

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

iD8DBQFDSKZKWry0BWjoQKURApo+AKCFQTYNDtxcCyCpjnG1Mz+gxStsPACfcuop
zib9ZSEZCMEpQXXxm2AjJj0=
=7K4X
-----END PGP SIGNATURE-----

--OwLcNYc0lM97+oe1--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051009051034.GB93028>