From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 26987 invoked by alias); 18 Apr 2013 07:54:17 -0000 Mailing-List: contact libc-ports-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Post: List-Help: , Sender: libc-ports-owner@sourceware.org Received: (qmail 26973 invoked by uid 89); 18 Apr 2013 07:54:17 -0000 X-Spam-SWARE-Status: No, score=-3.3 required=5.0 tests=AWL,BAYES_00,KHOP_RCVD_UNTRUST,KHOP_THREADED,RCVD_IN_DNSWL_LOW,RCVD_IN_HOSTKARMA_YE,TW_CP,TW_HW autolearn=ham version=3.3.1 Received: from mail-ie0-f178.google.com (HELO mail-ie0-f178.google.com) (209.85.223.178) by sourceware.org (qpsmtpd/0.84/v0.84-167-ge50287c) with ESMTP; Thu, 18 Apr 2013 07:54:16 +0000 Received: by mail-ie0-f178.google.com with SMTP id aq17so1488034iec.23 for ; Thu, 18 Apr 2013 00:54:15 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=mime-version:x-received:in-reply-to:references:date:message-id :subject:from:to:cc:content-type:x-gm-message-state; bh=hXbYBbRYSRxYYurYLev8APZMQxJi16u7lO1N8WKYdC0=; b=hvVr9KzVwKh0LAavETvmsOBnALN8yeM89c7BwDNrH59UGv8IEhjnTqm8r3x3lfc1ma 4JM5F/Pjjo8C0cDS4m5+GRrPxMIpRvWj8WaSjFtFVrIwdwLDTAgNqQWC7VzDAUB++kSq EpqkadPbJ0+lXpmyQjwlhMZWPjdmw58rvuXT6Smz/mvH7D9RwbZxtypQ0xkjqCLBIjO2 aMCKuu2KTL2ndez+oSvZ+TX8/R/A2F6I+4SH5QN0GHvY4KZLDS0Ykw6nE3nEWpJtqhLi 8bXsgwyCjK9hV2/3saHUYCUhYf/lphGCvrCZuaEC2pAbsdnvjXi0EaScFHgl4/d/VbpU 2GfQ== MIME-Version: 1.0 X-Received: by 10.50.77.48 with SMTP id p16mr12666602igw.53.1366271655158; Thu, 18 Apr 2013 00:54:15 -0700 (PDT) Received: by 10.64.100.174 with HTTP; Thu, 18 Apr 2013 00:54:15 -0700 (PDT) In-Reply-To: References: <516D18F0.4060009@linaro.org> <516EC27E.8080502@twiddle.net> <516FA3CF.8080200@twiddle.net> Date: Thu, 18 Apr 2013 07:54:00 -0000 Message-ID: Subject: Re: [PATCH v2] ARM: Add Cortex-A15 optimized NEON and VFP memcpy routines, with IFUNC. From: Will Newton To: Siddhesh Poyarekar Cc: Richard Henderson , libc-ports@sourceware.org, Patch Tracking Content-Type: text/plain; charset=ISO-8859-1 X-Gm-Message-State: ALoCoQlz34kW5E4OlS1t6D0IVz1Ydxoe8su1zbc0CVrLWPG7Y8+NZg5UeQlxPtEji8DYdpHZBFKA X-SW-Source: 2013-04/txt/msg00086.txt.bz2 On 18 April 2013 08:47, Siddhesh Poyarekar wrote: > On 18 April 2013 13:12, Richard Henderson wrote: >> On 2013-04-17 17:53, Will Newton wrote: >>>> >>>> From reading the comments of the code v7 is required for NEON, v6 is >>> >>> required for VFP and unaligned access is required. The unaligned >>> access requirement may be a problem on v5 I'm not sure. NB: I did not >>> write the memcpy code so I have not looked at it in great detail. >> >> >> Neon won't exist before v7, true. But then it ought not be advertized >> by the hwcap either, so it won't be selected. >> >> Anyone got a raspberry pi to benchmark the current and the new here? > > I can do the benchmark on my raspberry pi. Hi Siddesh, Thanks for the offer, I don't have any armv6 hardware to test with. The easiest way to test is probably using the cortex-strings package (as the patch I posted is for armv7). https://launchpad.net/cortex-strings "try-this -t memcpy" will benchmark the code I submitted, "try-glibc -t memcpy" will benchmark the current glibc code. -- Will Newton Toolchain Working Group, Linaro