From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <libc-alpha-return-86930-listarch-libc-alpha=sources.redhat.com@sourceware.org>
Received: (qmail 99851 invoked by alias); 9 Nov 2017 05:46:31 -0000
Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm
Precedence: bulk
List-Id: <libc-alpha.sourceware.org>
List-Subscribe: <mailto:libc-alpha-subscribe@sourceware.org>
List-Archive: <http://sourceware.org/ml/libc-alpha/>
List-Post: <mailto:libc-alpha@sourceware.org>
List-Help: <mailto:libc-alpha-help@sourceware.org>, <http://sourceware.org/ml/#faqs>
Sender: libc-alpha-owner@sourceware.org
Received: (qmail 62882 invoked by uid 89); 9 Nov 2017 05:45:55 -0000
Authentication-Results: sourceware.org; auth=none
X-Virus-Found: No
X-Spam-SWARE-Status: No, score=-2.2 required=5.0 tests=AWL,BAYES_00,FREEMAIL_FROM,RCVD_IN_DNSWL_NONE,RCVD_IN_SORBS_SPAM,SPF_PASS autolearn=no version=3.3.2 spammy=mrs, H*F:U*pinskia, HX-Envelope-From:sk:pinskia
X-Spam-User: qpsmtpd, 2 recipients
X-HELO: mail-oi0-f67.google.com
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net; s=20161025;
        h=x-gm-message-state:mime-version:in-reply-to:references:from:date
         :message-id:subject:to:cc;
        bh=zZ2vCkIdYpyImR1ZjfMrYUPM+pDZeoBkAfo3unQaDPw=;
        b=Lb07YLJ8JoqKwYdahjU9rSBONIGLhtKZgc1WU42piNfm9Har9pdTApIxY0MtJsmrpD
         uEGaMB7l7SdnEFLqPkZWcP+85KfAgYGpaQDSn7Gg4s7axjVF3/mcN74f8K6cOSLC+hIi
         qAnt/HtBp1IuTqID36+B4k7iEgkxWXvxoZB+eHCGOoZRkVWcy2ZY/EdJJILLu3eNbCfI
         +faIUHmWaidIlIoU8k4f7Z5OW5gfA5bLI5F5tBb/CUyEPNIRfpfeMAFS5XqwGTewEo92
         yh4F2GUEzCZ57rqG9YJeUjNZ4laDl5F/xW9/XMHyegtMadbB0UUmOppeTk/o/xb8Drsw
         QcEQ==
X-Gm-Message-State: AJaThX6gjELfFBIrqcnIJfu4oqa1LyWTblS3NydNT7gqvTirEV6GAlW3
	iNvqERkKY6n3UjqLuAHDjmu1ezfRztIju3Vj03U=
X-Google-Smtp-Source: ABhQp+Rpjl120DKmaylIK6yl5uncMwtg4su9MY1PMUCLxoCYLv7B6E2QY7Mr4vPy4pvVr3FCBVHWFr+a6dJ+cShPUns=
X-Received: by 10.202.168.140 with SMTP id r134mr1748971oie.150.1510206346150;
 Wed, 08 Nov 2017 21:45:46 -0800 (PST)
MIME-Version: 1.0
In-Reply-To: <0bb0752d-cfa7-5b4a-ad97-d2580e103e14@sourceware.org>
References: <1510204408-1739-1-git-send-email-siddhesh@sourceware.org>
 <1510204408-1739-4-git-send-email-siddhesh@sourceware.org>
 <CA+=Sn1=MCd=JsNZwaHkRJ2j-kgKWKkxHdSskuSfi4dLd=2N96Q@mail.gmail.com> <0bb0752d-cfa7-5b4a-ad97-d2580e103e14@sourceware.org>
From: Andrew Pinski <pinskia@gmail.com>
Date: Thu, 09 Nov 2017 05:46:00 -0000
Message-ID: <CA+=Sn1nQfPQk65hjRe+rH=kNEO=_4_kUxf8BH85uRwL86P7+_w@mail.gmail.com>
Subject: Re: [PATCH 3/3] aarch64: Hoist ZVA check out of the memset function
To: Siddhesh Poyarekar <siddhesh@sourceware.org>
Cc: GNU C Library <libc-alpha@sourceware.org>, Wilco Dijkstra <Wilco.Dijkstra@arm.com>, 
	Szabolcs Nagy <szabolcs.nagy@arm.com>
Content-Type: text/plain; charset="UTF-8"
X-SW-Source: 2017-11/txt/msg00305.txt.bz2

On Wed, Nov 8, 2017 at 9:44 PM, Siddhesh Poyarekar
<siddhesh@sourceware.org> wrote:
> On Thursday 09 November 2017 10:59 AM, Andrew Pinski wrote:
>> I don't like this at all for the increase file size for the not so
>> significant gain on real platforms.  I think we should declare falkor
>> micro-arch is broken and move on.
>
> Are you sure this doesn't give any gains for thunderx for the sizes I
> mentioned (256B - ~1K)?  I see significant gains on mustang too and it
> is obvious to see why; it is 3 instructions and a branch less in a hot
> path and that should be significant regardless of the MRS cost.

How can it, ThunderX has 128 byte cache lines.

Thanks,
Andrew

>
> If you don't see any gain then I don't mind posting this as a
> falkor-specific change.  If you change your mind in future we can always
> change the IFUNC condition.
>
> Siddhesh