Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 3 Nov 2014 21:53:26 +0100
From:      "O. Hartmann" <ohartman@zedat.fu-berlin.de>
To:        Dimitry Andric <dim@FreeBSD.org>
Cc:        Ed Maste <emaste@freebsd.org>, Andriy Gapon <avg@freebsd.org>, Harald Schmalzbauer <h.schmalzbauer@omnilan.de>, FreeBSD Current <freebsd-current@freebsd.org>, Nathan Whitehorn <nwhitehorn@freebsd.org>, Allan Jude <allanjude@freebsd.org>
Subject:   Re: CURRENT: EFI boot failure
Message-ID:  <20141103215326.4a2a0e91.ohartman@zedat.fu-berlin.de>
In-Reply-To: <9F1F287F-751C-4456-BB5E-540824E1B4E8@FreeBSD.org>
References:  <20140916020541.03c18d04.ohartman@zedat.fu-berlin.de> <54178607.1060305@freebsd.org> <541786BE.6010105@freebsd.org> <20140916075121.29989a53.ohartman@zedat.fu-berlin.de> <5417E20D.8070607@freebsd.org> <20140916230348.189e80cd.ohartman@zedat.fu-berlin.de> <CAPyFy2AicKf-PWGDnq80Zmm8ukwTkURZm8RRUFxdfHK=Xd7XAQ@mail.gmail.com> <5418B8C3.7040406@FreeBSD.org> <20140919152207.0473e213.ohartman@zedat.fu-berlin.de> <542183A6.7060802@omnilan.de> <54218B0F.20707@freebsd.org> <9F1F287F-751C-4456-BB5E-540824E1B4E8@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
--Sig_/IL3yE1G9y_cZe8hsVz_9mrW
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

Am Tue, 23 Sep 2014 17:14:46 +0200
Dimitry Andric <dim@FreeBSD.org> schrieb:

> On 23 Sep 2014, at 17:00, Nathan Whitehorn <nwhitehorn@freebsd.org> wrote:
> > On 09/23/14 07:28, Harald Schmalzbauer wrote:
> >>  Bez=FCglich O. Hartmann's Nachricht vom 19.09.2014 15:22 (localtime):
> >>> =85
> >>> The problem I reported about in the first place is triggered by a fau=
lty loader.efi
> >>> that arises, when optimisation level is -O3. -O2 works fine.
> >> I can confirm that this problem also shows up when using
> >> 'CPUTYPE?=3Dcore-avx2'
> >> Setting CPUTYPE to core-avx-i doesnt ehibit the problem.
> >>=20
> >> I could narrow down the cause to libefi.a (sys/boot/efi).
> >> But I don't understand the things going on there, so no clue how to fix
> >> besides maybe
> >>=20
> >> --- sys/boot/efi/Makefile.inc.orig 2014-09-23 16:22:46.000000000 +0200
> >> +++ sys/boot/efi/Makefile.inc 2014-09-23 16:25:16.000000000 +0200
> >> @@ -2,6 +2,10 @@
> >>=20
> >> BINDIR?=3D /boot
> >>=20
> >> +.ifdef CPUTYPE
> >> +.undef CPUTYPE
> >> +.endif
> >> +
> >> .if ${MACHINE_CPUARCH} =3D=3D "i386"
> >> CFLAGS+=3D -march=3Di386
> >> .endif
> > Could you try adding -mno-avx2 to /sys/boot/amd64/Makefile.inc line 9?
> > -Nathan
>=20
> IMHO CPUTYPE should be ignored for any boot loader program, and the
> lowest common denominator should be used instead (i486 for 32-bit, plain
> x86_64 for 64-bit).  It makes no sense to optimize boot loaders for e.g.
> core-avx2. :-)
>=20
> But indeed, it appears that we need to add more -mno-foo magic flags...
>=20
> -Dimitry
>=20

I repoted a bug at
Bug 194641 - [EFI] boot/loader.efi: miscompilation on Intel Haswell with AV=
X2=20


Please feel free to comment and replenish my superficial observation.

Hopefullz, this doesn't get lost. This nasty bug on Haswell CPU bothers me =
all the days I
update world.

--Sig_/IL3yE1G9y_cZe8hsVz_9mrW
Content-Type: application/pgp-signature
Content-Description: OpenPGP digital signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQEcBAEBAgAGBQJUV+tJAAoJEOgBcD7A/5N8/qYH/3BrCIInWFJb56OWcItBnEtx
1J44YlezR+3b8XP0Fve0MwcVNPHcloA8t7l1/19dI+jlmTZDX668RLdmOI4fKhyL
5OT5hDhcq4Bq+PW0CBN7yxU9/dID1O8jR5JI3M88TD+ixqUSfuHx1CFy1brncJmj
gbtqmffHaAxw7Vs/h2Hbo8qKmOPHDFdNXZm0gyq1VNeI/W2/CbZiym/iz9CRSbRT
MtZf+63kxO55oDl++Lk6kOqHaTBXS3ZEdbVHVA6/ksL3/zLOTZNh+7An05jbx81l
jUP2Ic5XCSKwKbeskz5HysBnOyUCo2OlPvxTKB2u8f0G0378UjqmtjP9Ke9fmv8=
=UWZg
-----END PGP SIGNATURE-----

--Sig_/IL3yE1G9y_cZe8hsVz_9mrW--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20141103215326.4a2a0e91.ohartman>