Date: Fri, 25 May 2007 19:26:09 +1000 From: Peter Jeremy <peterjeremy@optushome.com.au> To: "Conrad J. Sabatier" <conrads@cox.net> Cc: freebsd-amd64@freebsd.org Subject: Re: CPUTYPE Message-ID: <20070525092609.GD1159@turion.vk2pj.dyndns.org> In-Reply-To: <200705250445.l4P4jRp5076893@serene.no-ip.org> References: <465454CF.3060601@sun-fish.com> <200705250445.l4P4jRp5076893@serene.no-ip.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--J/dobhs11T7y2rNN Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 2007-May-24 23:45:27 -0500, "Conrad J. Sabatier" <conrads@cox.net> wrote: >Actually, section 3.17.14 in the gcc 4.2.0 docs (Intel 386 and AMD >x86-64 Options) states that there is now also available a new "native" >CPU type, which will cause gcc to determine the CPU type of the build >machine and generate code for the same. Sounds good. But does gcc 4.2 generate optimal code for the CPU? ISTR bde@ pointing out a number of cases where the "obvious" CPU type generates code that is slower than code compiled for a different architecture (for gcc 3.4). --=20 Peter Jeremy --J/dobhs11T7y2rNN Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (FreeBSD) iD8DBQFGVqux/opHv/APuIcRAhS/AKC473SGMj/sWKc0GBJRdjBwW/TEygCffW8Q IQReEfPtx/ZUj5rPfX+TZdo= =DSTR -----END PGP SIGNATURE----- --J/dobhs11T7y2rNN--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070525092609.GD1159>