From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-patches-return-281876-listarch-gcc-patches=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 9657 invoked by alias); 30 Dec 2010 11:19:12 -0000
Received: (qmail 9632 invoked by uid 22791); 30 Dec 2010 11:19:11 -0000
X-SWARE-Spam-Status: No, hits=-2.1 required=5.0	tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,RFC_ABUSE_POST,TW_AV,TW_MV,TW_VZ
X-Spam-Check-By: sourceware.org
Received: from mail-px0-f175.google.com (HELO mail-px0-f175.google.com) (209.85.212.175)    by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Thu, 30 Dec 2010 11:19:06 +0000
Received: by pxi17 with SMTP id 17so2548409pxi.20        for <gcc-patches@gcc.gnu.org>; Thu, 30 Dec 2010 03:19:05 -0800 (PST)
MIME-Version: 1.0
Received: by 10.142.60.13 with SMTP id i13mr13123283wfa.322.1293707944982; Thu, 30 Dec 2010 03:19:04 -0800 (PST)
Received: by 10.143.4.16 with HTTP; Thu, 30 Dec 2010 03:19:04 -0800 (PST)
In-Reply-To: <AANLkTi=0tNb1FX7XweRCUQBh6OUjNW7f6+vyO0YuYk=z@mail.gmail.com>
References: <20101217190347.GA2856@intel.com>	<AANLkTinempMG4qbZsZ_tHMnyHqpXd+2JDZEYzqdASEWo@mail.gmail.com>	<AANLkTi=ky0A+oH9e23zX=yWrwjvFKzkNG1qVrvQ1bcJO@mail.gmail.com>	<AANLkTi=K3bPqru+joTk1fsjthVnSe0EMkoEF=tGr7SxF@mail.gmail.com>	<AANLkTi=0tNb1FX7XweRCUQBh6OUjNW7f6+vyO0YuYk=z@mail.gmail.com>
Date: Thu, 30 Dec 2010 12:42:00 -0000
Message-ID: <AANLkTi=e2ZemY8AaWx6sGOBS_RMiS3=LG9z9pZ8SMbsq@mail.gmail.com>
Subject: Re: PATCH: PR target/46519: Missing vzeroupper
From: Uros Bizjak <ubizjak@gmail.com>
To: "H.J. Lu" <hjl.tools@gmail.com>
Cc: gcc-patches@gcc.gnu.org, Richard Guenther <rguenther@suse.de>, 	Jakub Jelinek <jakub@redhat.com>, Mark Mitchell <mark@codesourcery.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-patches.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-patches/>
List-Post: <mailto:gcc-patches@gcc.gnu.org>
List-Help: <mailto:gcc-patches-help@gcc.gnu.org>
Sender: gcc-patches-owner@gcc.gnu.org
X-SW-Source: 2010-12/txt/msg01974.txt.bz2

On Wed, Dec 29, 2010 at 4:32 PM, H.J. Lu <hjl.tools@gmail.com> wrote:

>> I think that due to these reasons, the patch should be committed to
>> SVN even in this development stage. Even if the algorithm is not
>> optimal, the patch demonstrably produces substantially better code.
>> This feature has no impact on generic code without -mvzeroupper /
>> -mavx switch, and since there are currently very few AVX users,
>> negligible overall impact.
>>
>>> gcc/
>>>
>>> 2010-12-18 =A0H.J. Lu =A0<hongjiu.lu@intel.com>
>>>
>>> =A0 =A0 =A0 =A0PR target/46519
>>> =A0 =A0 =A0 =A0* config/i386/i386.c (block_info_def): Remove referenced=
, count
>>> =A0 =A0 =A0 =A0and rescanned.
>>> =A0 =A0 =A0 =A0(move_or_delete_vzeroupper_2): Updated.
>>> =A0 =A0 =A0 =A0(move_or_delete_vzeroupper_1): Rewritten to avoid recurs=
ive call.
>>> =A0 =A0 =A0 =A0(rescan_move_or_delete_vzeroupper): Removed.
>>> =A0 =A0 =A0 =A0(move_or_delete_vzeroupper): Repeat processing all basic=
 blocks
>>> =A0 =A0 =A0 =A0until no basic block state is changed to used at exit.
>>>
>>> gcc/testsuite/
>>>
>>> 2010-12-18 =A0H.J. Lu =A0<hongjiu.lu@intel.com>
>>>
>>> =A0 =A0 =A0 =A0PR target/46519
>>> =A0 =A0 =A0 =A0* gfortran.dg/pr46519-2.f90: New.
>>>
>>
>> The patch is OK, but please allow a day or two for RMs (CC'd) to
>> eventually comment.
>
> We will investigate LCM for 4.7. =A0In the meantime, here is =A0a small p=
atch
> on top of the current one. If the upper 128bits are never changed in a ba=
sic
> block, we can skip it in the later passes. =A0OK for trunk together with =
the
> current patch?
>
> 2010-12-29  H.J. Lu  <hongjiu.lu@intel.com>
>
>        * config/i386/i386.c (upper_128bits_state): Update comments.
>        (block_info_def): Add unchanged.
>        (move_or_delete_vzeroupper_2): Short circuit if upper 128bits
>        are unchanged in the block.
>

OK, but please remove now redundant coments in

@@ -60,14 +60,17 @@ along with GCC; see the file COPYING3.  If not see
 enum upper_128bits_state
 {
   unknown =3D 0,		/* Unknown.  */
-  unused,		/* Not used or not referenced.  */
-  used			/* Used or referenced.  */
+  unused,		/* Not used.  */
+  used			/* Used.  */
 };

Thanks,
Uros.