From mboxrd@z Thu Jan  1 00:00:00 1970
From: Dave Love <fx@gnu.org>
Subject: Re: Packaging BLIS
Date: Fri, 08 Sep 2017 23:36:10 +0100
Message-ID: <87tw0com1x.fsf@albion.it.manchester.ac.uk>
References: <87inhhw1ms.fsf@elephly.net> <87a82s9cw3.fsf@gnu.org>
	<87wp5ufkqs.fsf@albion.it.manchester.ac.uk> <87k21nerqa.fsf@inria.fr>
	<87inh2zog7.fsf@albion.it.manchester.ac.uk>
	<87a826tslo.fsf_-_@inria.fr>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Return-path: <guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org>
Received: from eggs.gnu.org ([2001:4830:134:3::10]:45142)
	by lists.gnu.org with esmtp (Exim 4.71) (envelope-from <fx@gnu.org>)
	id 1dqRso-0001rd-QE
	for guix-devel@gnu.org; Fri, 08 Sep 2017 18:36:19 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <fx@gnu.org>) id 1dqRsj-0000rC-Th
	for guix-devel@gnu.org; Fri, 08 Sep 2017 18:36:18 -0400
Received: from serenity.mcc.ac.uk ([130.88.200.93]:48943)
	by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32)
	(Exim 4.71) (envelope-from <fx@gnu.org>) id 1dqRsj-0000ph-N2
	for guix-devel@gnu.org; Fri, 08 Sep 2017 18:36:13 -0400
In-Reply-To: <87a826tslo.fsf_-_@inria.fr> ("Ludovic
	\=\?iso-8859-1\?Q\?Court\?\=
	\=\?iso-8859-1\?Q\?\=E8s\=22's\?\= message of
	"Thu, 07 Sep 2017 17:51:15 +0200")
List-Id: "Development of GNU Guix and the GNU System distribution."
	<guix-devel.gnu.org>
List-Unsubscribe: <https://lists.gnu.org/mailman/options/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/archive/html/guix-devel/>
List-Post: <mailto:guix-devel@gnu.org>
List-Help: <mailto:guix-devel-request@gnu.org?subject=help>
List-Subscribe: <https://lists.gnu.org/mailman/listinfo/guix-devel>,
	<mailto:guix-devel-request@gnu.org?subject=subscribe>
Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org
Sender: "Guix-devel" <guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org>
To: Ludovic =?iso-8859-1?Q?Court=E8s?= <ludovic.courtes@inria.fr>
Cc: guix-devel <guix-devel@gnu.org>

ludovic.courtes@inria.fr (Ludovic Court=C3=A8s) writes:

> That caught my attention so I packaged BLIS:
>
>   https://git.savannah.gnu.org/cgit/guix.git/commit/?id=3D5a7deb117424ff4=
d430b771b50e534cf065c0ba1
>
> There are several =E2=80=9Cflavors=E2=80=9D of BLIS, so you can always re=
build your
> favorite program with:
>
>   --with-input=3Dopenblas=3Dblis-haswell
>
> and similar (or even =E2=80=98--with-graft=3Dblis=3Dblis-haswell=E2=80=99=
 where applicable).

I'm not aware of an advantage of BLIS over OpenBLAS except for KNL (and
maybe Skylake Xeon, if that's covered).  My interest was specifically
for avx512.  (BLIS is documented as basically a research effort, which I
think explains the way it's built.)

> Hopefully the issue you linked to above will be fixed in future versions
> of BLIS, at which point we can probably provide a single =E2=80=9Cblis=E2=
=80=9D package.
>
> Ludo=E2=80=99.

I still think it's important to be able to select the BLAS at run time
with dynamic linking.  I suppose the trick I used for rpm will work
similarly, though I'm not sure how ld.so behaves in Guix.