Date: Wed, 24 Oct 2012 11:12:32 +0300 From: Konstantin Belousov <kostikbel@gmail.com> To: Volodymyr Kostyrko <c.kworr@gmail.com> Cc: freebsd-stable@freebsd.org Subject: Re: wine, gcc and clang with CPUTYPE Message-ID: <20121024081232.GA35915@deviant.kiev.zoral.com.ua> In-Reply-To: <50879F62.2010004@gmail.com> References: <50879F62.2010004@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--OgaV/uj8Af8zlSb7 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, Oct 24, 2012 at 10:57:22AM +0300, Volodymyr Kostyrko wrote: > Hi all. >=20 > I just have taken some time to inspect CPUTYPE support for clang. It=20 > seems to me that clang generates incorrect code in some cases. >=20 > The first failure point I discovered was inability to build gcc from=20 > sources or compile something with gcc. Code produced by gcc seem to fail= =20 > whether this was gcc compiled from bootstrap or anything else: >=20 > http://lists.freebsd.org/pipermail/freebsd-multimedia/2012-October/013469= =2Ehtml >=20 > I started testing by commenting out CPUTYPE in make.conf. After first=20 > rebuild I also updated the ports and installed new version of=20 > wine-devel. And to my surprise it works like a charm. Rolling back to=20 > the world built with CPUTYPE=3Dnative makes wine break again. >=20 > To my surprise CPUTYPE was not the cause of wine failure per se. Wine=20 > continues to work for k6, k6-3, athlon and athlon-tbird. But it=20 > completely fails when the world was built with athlon-4 and athlon-xp. >=20 > Trying to recompile gcc I also found that everything works and yet again= =20 > up to the athlon-tbird. >=20 > My conclusion is: clang incorrectly produces code within one of core=20 > libraries (I haven't tested which one yet, but I suspect libgcc_s.so)=20 > when optimizing for athlon-4 or athlon-xp. I am not versed in the AMD marketing monikers. I guess that athlon-{4,xp} turns on SSE and might be SSE2, while previous selections turn it off. Can you confirm/deny this ? BTW, did you tested on i386 or amd64 ? --OgaV/uj8Af8zlSb7 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (FreeBSD) iEYEARECAAYFAlCHovAACgkQC3+MBN1Mb4hpBwCgmay2k/V/xNJeenBdNbV7mAlC pa8An2MqI9+nk1HAbKaTFr3ni/twtcDm =DfKs -----END PGP SIGNATURE----- --OgaV/uj8Af8zlSb7--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20121024081232.GA35915>