From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from lists.gentoo.org (pigeon.gentoo.org [208.92.234.80]) by finch.gentoo.org (Postfix) with ESMTP id 204BE13800E for ; Fri, 10 Aug 2012 15:08:36 +0000 (UTC) Received: from pigeon.gentoo.org (localhost [127.0.0.1]) by pigeon.gentoo.org (Postfix) with SMTP id DF36EE062B for ; Fri, 10 Aug 2012 15:08:34 +0000 (UTC) Received: from mail-lpp01m010-f53.google.com (mail-lpp01m010-f53.google.com [209.85.215.53]) by pigeon.gentoo.org (Postfix) with ESMTP id 0DE43E0710 for ; Fri, 10 Aug 2012 14:36:51 +0000 (UTC) Received: by lahc1 with SMTP id c1so862974lah.40 for ; Fri, 10 Aug 2012 07:36:51 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=D4vIQc8aigrLQCZtNQTuM6VNr75AhzbdEaJiO7iCtmI=; b=Iq1POtjdvX1iRmS7NaUA0y+yQuvnQ82zEaX9liENihpkxADmh+eaIP6b96pRV7Znkx fH8ZBrsLCOSZlfCXPFEXCAh4M7qiHSPPCRjdM3/qBxZKa98NC/ek6+iqkS4df5ngy9ZA DiCqppMDUXLDLtXAs0AQhfdSfz3xa95U7MJ/iBmZPphLx8P2lPceaPKG6Dn68uXBtwM7 RBXaRgef+P6+2+70VNd/0BYwwhbxpprmdlYazn7xVKAhPaFKVvEh7AAiS5dEnR0XzzzT GsGnCvPHQ3tf3FafMfU6gYiJc+Ano+Q+wIeov46cGktYm3vzu+BZprpoJFBASHXLDzLs h8RQ== Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Gentoo Linux mail X-BeenThere: gentoo-embedded@lists.gentoo.org Reply-to: gentoo-embedded@lists.gentoo.org MIME-Version: 1.0 Received: by 10.152.144.9 with SMTP id si9mr3276534lab.4.1344609411016; Fri, 10 Aug 2012 07:36:51 -0700 (PDT) Received: by 10.112.0.165 with HTTP; Fri, 10 Aug 2012 07:36:50 -0700 (PDT) Received: by 10.112.0.165 with HTTP; Fri, 10 Aug 2012 07:36:50 -0700 (PDT) In-Reply-To: <201205180110.47753.vapier@gentoo.org> References: <201205180110.47753.vapier@gentoo.org> Date: Fri, 10 Aug 2012 10:36:50 -0400 Message-ID: Subject: Re: [gentoo-embedded] default fp selection with armv6/armv7 targets From: Christopher Friedt To: gentoo-embedded@lists.gentoo.org Cc: "Ryan C. Gordon" Content-Type: multipart/alternative; boundary=e89a8f23470f4e26eb04c6ea464c X-Archives-Salt: 053e25ce-d0b2-435f-86bb-81641ab00fd9 X-Archives-Hash: 5380e93fe3a83e5ddb1dc3371943737c --e89a8f23470f4e26eb04c6ea464c Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Using the least common denominator is another food reason to consider link time optimization... I wonder if e.g. fatelf would make it easier to implement that. Multiple arm binaries glued together based on variant and hwcaps (e.g. v6+vfp-d16, v7+vfp-d32, v7+neon)? Heck, if fatelf were used, only one stage3 tar ball would be necessary for all arch's combined... although downloading would be significantly slower :-/ =A9 Sent from my Android On May 18, 2012 1:10 AM, "Mike Frysinger" wrote: > currently, we have the system: > - if chost matches *-softfloat-*, you get softfloat > - if chost matches *-hardfloat-*, you get hardfloat > - for everything else, you get the gcc default > > with the standardization work going on with armv7+ and hardfp, i've made > the > following change: > - if chost matches armv7*-softfloat-*, you got softfloat > - for all other armv7* targets, you get hardfloat vfp3-d16 > > along those lines, i've also slipped in: > - if chost matches armv6*-softfloat-*, you got softfloat > - for all other armv6* targets, you get hardfloat vfp > > considering vfp is required baseline in these cores now, it doesn't make > sense > to not use it if the user has explicitly stated they're targeting these > arches. > > if you really want to use a different default, you can still use > EXTRA_ECONF to > set whatever you want. > -mike > --e89a8f23470f4e26eb04c6ea464c Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable

Using the least common denominator is another food reason to= consider link time optimization... I wonder if e.g. fatelf would make it e= asier to implement that. Multiple arm binaries glued together based on vari= ant and hwcaps (e.g. v6+vfp-d16, v7+vfp-d32, v7+neon)? Heck, if fatelf were= used, only one stage3 tar ball would be necessary for all arch's combi= ned... although downloading would be significantly slower :-/

=A9

Sent from my Android

On May 18, 2012 1:10 AM, "Mike Frysinger&qu= ot; <vapier@gentoo.org> wrot= e:
currently, we have the system:
=A0 =A0 =A0 =A0 - if chost matches *-softfloat-*, you get softfloat
=A0 =A0 =A0 =A0 - if chost matches *-hardfloat-*, you get hardfloat
=A0 =A0 =A0 =A0 - for everything else, you get the gcc default

with the standardization work going on with armv7+ and hardfp, i've mad= e the
following change:
=A0 =A0 =A0 =A0 - if chost matches armv7*-softfloat-*, you got softfloat =A0 =A0 =A0 =A0 - for all other armv7* targets, you get hardfloat vfp3-d16<= br>
along those lines, i've also slipped in:
=A0 =A0 =A0 =A0 - if chost matches armv6*-softfloat-*, you got softfloat =A0 =A0 =A0 =A0 - for all other armv6* targets, you get hardfloat vfp

considering vfp is required baseline in these cores now, it doesn't mak= e sense
to not use it if the user has explicitly stated they're targeting these=
arches.

if you really want to use a different default, you can still use EXTRA_ECON= F to
set whatever you want.
-mike
--e89a8f23470f4e26eb04c6ea464c--