From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qe0-x22e.google.com (mail-qe0-x22e.google.com [IPv6:2607:f8b0:400d:c02::22e]) by dpdk.org (Postfix) with ESMTP id 87AEF4C6E for ; Thu, 1 Aug 2013 22:22:37 +0200 (CEST) Received: by mail-qe0-f46.google.com with SMTP id i11so1392853qej.19 for ; Thu, 01 Aug 2013 13:22:57 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type; bh=oHuQxO+7h+ja4hBv7zioxlZo9iTk0yAahNN6YS/XYnY=; b=BOA6fpFoF44heXtJ8mIqG63wPFspEWvFQVoHVZ5sCFrXPLmuVjfxh0e+ayhfTbrD3g nn5m/RdSGZrdz5P1wm+V8c2uBTpbVPlOHwzhPMt6SFn9uzTBMD078xSi7s7NPBrRW3a0 xINK4u+W1bpINJ5jcB0r9LRI0B39KOvDBz1K//LVG29HP9l755TXUOunr/F+RXeShLQU z9wB3e8MRCF4p54xHbOSNMqh7mBmEQX/6Q7Ur5dKN8agbQrxfm/NpyXwyL/7JiYZTH8p cVGFeHiXh1TSFfnTYiIGDEftWr1mcgAiGN3n6gXGMzXg/VtSMBb0onMVnfTpdQDeAPeJ MsGg== MIME-Version: 1.0 X-Received: by 10.49.131.36 with SMTP id oj4mr4553496qeb.51.1375388576975; Thu, 01 Aug 2013 13:22:56 -0700 (PDT) Sender: silverrocker@gmail.com Received: by 10.49.12.136 with HTTP; Thu, 1 Aug 2013 13:22:56 -0700 (PDT) Received: by 10.49.12.136 with HTTP; Thu, 1 Aug 2013 13:22:56 -0700 (PDT) In-Reply-To: <20130801122309.5d8d100b@samsung-9> References: <51FA80BF.2020801@bisdn.de> <20130801100638.4c0f06a8@samsung-9> <51FAA647.7070709@bisdn.de> <20130801122309.5d8d100b@samsung-9> Date: Thu, 1 Aug 2013 22:22:56 +0200 X-Google-Sender-Auth: pywdygjiKJm-N5wsZXdeegMnKG8 Message-ID: From: Balazs Nemeth To: Stephen Hemminger Content-Type: multipart/alternative; boundary=047d7bb03d368f11fe04e2e89b72 Cc: dev@dpdk.org Subject: Re: [dpdk-dev] Non-argv dependant rte_eal_init() call X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 01 Aug 2013 20:22:37 -0000 --047d7bb03d368f11fe04e2e89b72 Content-Type: text/plain; charset=ISO-8859-1 Letting DPDK deduce de cpus and memchannels is just a move of feautures in your application. In some cases you might want to limit the number of cores. It's just simple a way to allow more control when initializing DPDK. Kind Regards Balazs Nemeth Hasselt University, Belgium 2nd Master in Computer Science On Aug 1, 2013 9:23 PM, "Stephen Hemminger" wrote: > On Thu, 01 Aug 2013 20:17:43 +0200 > Marc Sune wrote: > > > Thanks Stephen for the hack. > > > > Unfortunately, our main already has parameters, and are all > > platform(architecture) agnostic, so this would break the assumption that > > arguments should be platform agnostic. > > > > But anyway thanks ;) > > marc > > > > Also, our application is started by an init script and there is some shell > magic to attempt to deduce number of cpus (not hard) and number of memory > channels (kind of messy). I do think the DPDK should work without any > arguments > and do this itself. The CPU part is easy by looking at /sys, but the memory > channel information has to come from DMI which is a real ugly mess. > > > --047d7bb03d368f11fe04e2e89b72 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable

Letting DPDK deduce de cpus and memchannels is just a move o= f feautures in your application. In some cases you might want to limit the = number of cores. It's just simple a way to allow more control when init= ializing DPDK.

Kind Regards

Balazs Nemeth
Hasselt University, Belgium
2nd Master in Computer Science

On Aug 1, 2013 9:23 PM, "Stephen Hemminger&= quot; <stephen@networkplum= ber.org> wrote:
On Thu, 01 Aug 2013 20:17:43 +0200
Marc Sune <marc.sune@bisdn.de&= gt; wrote:

> Thanks Stephen for the hack.
>
> Unfortunately, our main already has parameters, and are all
> platform(architecture) agnostic, so this would break the assumption th= at
> arguments should be platform agnostic.
>
> But anyway thanks ;)
> marc
>

Also, our application is started by an init script and there is some shell<= br> magic to attempt to deduce number of cpus (not hard) and number of memory channels (kind of messy). =A0I do think the DPDK should work without any ar= guments
and do this itself. The CPU part is easy by looking at /sys, but the memory=
channel information has to come from DMI which is a real ugly mess.


--047d7bb03d368f11fe04e2e89b72--