Received: by 10.58.229.232 with SMTP id st8mr3324146vec.34.1350898097311;
        Mon, 22 Oct 2012 02:28:17 -0700 (PDT)
X-BeenThere: v8-dev@googlegroups.com
Received: by 10.220.21.82 with SMTP id i18ls326938vcb.7.gmail; Mon, 22 Oct
 2012 02:28:16 -0700 (PDT)
Received: by 10.58.94.103 with SMTP id db7mr3248611veb.0.1350898096738;
        Mon, 22 Oct 2012 02:28:16 -0700 (PDT)
Received: by 10.58.94.103 with SMTP id db7mr3248610veb.0.1350898096725;
        Mon, 22 Oct 2012 02:28:16 -0700 (PDT)
Return-Path: <3sBGFUBUJAHERWgdbXjbRdSTgTkXTl-WgVbPXa....@2uix4h7xygsz66weerlq.apphosting.bounces.google.com>
Received: from mail-vc0-f199.google.com (mail-vc0-f199.google.com [209.85.220.199])
        by gmr-mx.google.com with ESMTPS id q19si2194687vds.3.2012.10.22.02.28.16
        (version=TLSv1/SSLv3 cipher=OTHER);
        Mon, 22 Oct 2012 02:28:16 -0700 (PDT)
Received-SPF: pass (google.com: domain of 3sBGFUBUJAHERWgdbXjbRdSTgTkXTl-WgVbPXa....@2uix4h7xygsz66weerlq.apphosting.bounces.google.com designates 209.85.220.199 as permitted sender) client-ip=209.85.220.199;
Authentication-Results: gmr-mx.google.com; spf=pass (google.com: domain of 3sBGFUBUJAHERWgdbXjbRdSTgTkXTl-WgVbPXa....@2uix4h7xygsz66weerlq.apphosting.bounces.google.com designates 209.85.220.199 as permitted sender) smtp.mail=3sBGFUBUJAHERWgdbXjbRdSTgTkXTl-WgVbPXa....@2uix4h7xygsz66weerlq.apphosting.bounces.google.com; dkim=pass header...@chromium.org
Received: by mail-vc0-f199.google.com with SMTP id gb23so4643397vcb.6
        for <v8-dev@googlegroups.com>; Mon, 22 Oct 2012 02:28:16 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=chromium.org; s=google;
        h=mime-version:reply-to:x-google-appengine-app-id:message-id:date
         :subject:from:to:cc:content-type;
        bh=jeq6NRWEA9eADrKRTgOULZNEUmylZ7S8uqXch1obKmg=;
        b=UUg/MtCCqKm2fssg0pSssffMJfzuOJK2MZBVUBYX0m9bZjZK3apAzAXAwYS0R6ZZIh
         W1tG65rQmNLcw7sQaZpsvdTJIhg5mwLbJicIXAwXIefXcgQLWqPdv6oMiPDuHehDZp/V
         7b82VeNsXQkudlrtfWjQGLDa02ChPKytLJyj8=
MIME-Version: 1.0
Received: by 10.52.72.100 with SMTP id c4mr2804281vdv.5.1350898096622; Mon, 22
 Oct 2012 02:28:16 -0700 (PDT)
Reply-To: rkrit...@codeaurora.org, yang...@chromium.org, da...@chromium.org, 
	v8-dev@googlegroups.com
Message-ID: <20cf307f31342d4d4c04cca27...@google.com>
Date: Mon, 22 Oct 2012 09:28:16 +0000
Subject: Re: Improves performance for calls from Crankshaft code to certain
 string and... (issue 10996065)
From: yang...@chromium.org
To: rkrit...@codeaurora.org, da...@chromium.org
Cc: v8-dev@googlegroups.com
Content-Type: text/plain; charset=ISO-8859-1; format=flowed; delsp=yes

On 2012/10/12 15:51:15, danno wrote:
> Whoops, hit the wrong button. This doesn't look good to me on this patch  
> quite
> yet,

A quick test shows that this patch causes the V8 benchmark bundled in the V8
repository (benchmark/run.js) to crash with segfaults (RayTrace and RegExp).

This patch seems to attempt to improve V8's performance for string and  
regexp
operations when the result is not used. I expect this to be very rare in  
actual
javascript applications (why would you run an expensive string/regexp  
operation
when you don't use its result?).
However, V8's RegExp benchmark contains a lot of those operations with  
unused
results. This has been a known fact and there are plans (see 10913269) to  
change
that so that this benchmark tests what it's intended to test
(https://chromiumcodereview.appspot.com/10913269/). That's why I think this
patch is taking the wrong direction.

http://codereview.chromium.org/10996065/