From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-yw0-f179.google.com (mail-yw0-f179.google.com [209.85.161.179]) by dpdk.org (Postfix) with ESMTP id 809F32B9F for ; Wed, 3 May 2017 07:24:40 +0200 (CEST) Received: by mail-yw0-f179.google.com with SMTP id u70so80023308ywe.2 for ; Tue, 02 May 2017 22:24:40 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc; bh=quwjM7ABEbuu/R0SnD0wmjzQn+6TkCJ3GdtlaIE5iek=; b=MYeELzgaLagzxb3jj/xDG+pf3cL8wPFlIzXAXG6L38cX45A0OLQ/E8kvNqEOFMPk83 sHpqJFzuyREEn31FtAY5R+Lr36TZa6PY5dR13BxcfpoM7RSFgNfpEC0QlsONn+oL5/+5 RIM+zewC2lhhwGEYgEXygWSykwD6qu3acfgMg= 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=quwjM7ABEbuu/R0SnD0wmjzQn+6TkCJ3GdtlaIE5iek=; b=Qm4Wm5NQoopK0jAmyWAc1CEjJ6p3IC8htJZ98OWbuMkWmeKshNGLBzTExZZVAP8SEj ouaaixl/Zr0i8UAAJpz11qU+q6WxSVNlNf5EEFUDPkGdoXY1cKMP9wxokRHRQ/0mK7Wk ZJtE2QTsuieyKS6V4Kcj72uerk6/Fx0X9nV3vVvRktrXTApILPvAzv+Xj3UHdYnHxrsb xnVhovD4uvckZmhznUIEifvlc9wxydF7/3GHy2H9pSEJ0O4g2vQgBBbjjyKXCD7U4KK0 ZUIRKLQdDKsC33cmDGVkLUQ2hExSBWzylyjZgeoSIi6OddYdfxOQzgxnHyZ8BRkVm3bW n1kw== X-Gm-Message-State: AN3rC/60qUw6yVc1i2sjwqhGcQWcGu1nuEIJtzWi4griAG48cbQXMKww K35UL2jh5lsFBfi3+tCjKAotA/M6iV1o X-Received: by 10.13.230.86 with SMTP id p83mr26670575ywe.207.1493789079672; Tue, 02 May 2017 22:24:39 -0700 (PDT) MIME-Version: 1.0 Received: by 10.37.4.16 with HTTP; Tue, 2 May 2017 22:24:39 -0700 (PDT) In-Reply-To: <1493725624.3602.34.camel@caviumnetworks.com> References: <1493709255-8887-1-git-send-email-jianbo.liu@linaro.org> <1493709255-8887-5-git-send-email-jianbo.liu@linaro.org> <1493725624.3602.34.camel@caviumnetworks.com> From: Jianbo Liu Date: Wed, 3 May 2017 13:24:39 +0800 Message-ID: To: "Sekhar, Ashwin" Cc: "tomasz.kantecki@intel.com" , "Jacob, Jerin" , "dev@dpdk.org" Content-Type: text/plain; charset=UTF-8 Subject: Re: [dpdk-dev] [PATCH 5/5] examples/l3fwd: add neon support for l3fwd X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 03 May 2017 05:24:40 -0000 Hi Ashwin, On 2 May 2017 at 19:47, Sekhar, Ashwin wrote: > Hi Jianbo, > > I tested your neon changes on thunderx. I am seeing a performance > regression of ~10% for LPM case and ~20% for EM case with your changes. > Did you see improvement on any arm64 platform with these changes. If > yes, how much was the improvement? Thanks for your reviewing and testing. For some reason, I have not done much with the performance testing. I'll send a new version later after tuning the performance. Thanks! Jianbo > > FYI, I had also tried vectorizing the l3fwd app with neon. Few of the > optimizations that I can suggest that helped in my case. > > * Packet data prefetch is missing in the x86 sse version compared to > the scalar version (l3fwd_lpm_send_packets vs > l3fwd_lpm_no_opt_send_packets) . I couldn't understand why this was not > done in x86. But adding the prefetch was improving performance for > thunderx. > > * Offsets to some packet elements like eth_hdr, ip header, packet type > etc. are recalculated in different functions. Calculating them once, > caching them and passing them directly to different functions was > improving performance. > > * There are 3 different loops in l3fwd_lpm_send_packets where we > iterate over the packets. One each for processx4_step1 and > processx4_step2 and one in send_packets_multi. Unifying these loops > were also helping. > > Thanks and Regards > Ashwin >