From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wg0-f47.google.com (mail-wg0-f47.google.com [74.125.82.47]) (using TLSv1 with cipher RC4-MD5 (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 57529202224 for ; Mon, 25 Jun 2012 08:30:48 -0700 (PDT) Received: by wgbfa7 with SMTP id fa7so3017540wgb.28 for ; Mon, 25 Jun 2012 08:30:46 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=qCCbiiM3V4MO3pV/rIwbdPuGnj/MCxxrBVLiwMQvUMI=; b=PdVFIg+Lvgn/xJl+f/IWxNH3tFCLhq5RpUi1rqmCkgWAxknqmno+mDElyzB8KQBurj tPI59V0nPpTRSflcToWFCnT0/8qHlipCFEICkNkUMEhXVYG6zCbzKhfqFrL8AbWHPIlO mbZgCqIfUU1R/WHBJrb33hHQEHcu9W4lzLHSOrSLrP9E1i//qpXyb4wThJ3/Cs73xzZk v6OQSASlLXuBTki8tDJb6ugAhGhmE1bQJcBUZmSqCgzWlrWtDt2pmA/bSBdO5K1Yi6EZ Bjcwcvx40Ql1IT/Rq7XuAL+WhK1gJw6llVhQGWOs8wWlem6+2Nlbg6Jn1KE+cNMNij6H wvNg== MIME-Version: 1.0 Received: by 10.180.81.65 with SMTP id y1mr25536078wix.1.1340638246123; Mon, 25 Jun 2012 08:30:46 -0700 (PDT) Received: by 10.223.103.199 with HTTP; Mon, 25 Jun 2012 08:30:46 -0700 (PDT) In-Reply-To: <1340637617.10893.56.camel@edumazet-glaptop> References: <1340600422-1806-1-git-send-email-dave.taht@bufferbloat.net> <1340600422-1806-2-git-send-email-dave.taht@bufferbloat.net> <1340601724.23933.16.camel@edumazet-glaptop> <1340637617.10893.56.camel@edumazet-glaptop> Date: Mon, 25 Jun 2012 08:30:46 -0700 Message-ID: From: Dave Taht To: Eric Dumazet Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Cc: codel@lists.bufferbloat.net, =?ISO-8859-1?Q?Dave_T=E4ht?= Subject: Re: [Codel] [RFCv2 PATCH] codel: add ecn_target for when to drop rather than mark ecn packets X-BeenThere: codel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: CoDel AQM discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 25 Jun 2012 15:30:48 -0000 On Mon, Jun 25, 2012 at 8:20 AM, Eric Dumazet wrot= e: > On Mon, 2012-06-25 at 07:50 -0700, Dave Taht wrote: >> On Sun, Jun 24, 2012 at 10:22 PM, Eric Dumazet = wrote: >> > On Sun, 2012-06-24 at 22:00 -0700, Dave T=E4ht wrote: >> >> From: Dave Taht > >> >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 codel_New= ton_step(vars); >> >> - =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (params-= >ecn && INET_ECN_set_ce(skb)) { >> >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (params-= >ecn && >> >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 par= ams->ecn_target > vars->ldelay && >> > >> > Wrong test ? >> > >> > We want ECN if delay < ecn_target, not if delay > ecn_target >> > >> > (unresponsive flows will make delay being above ecn_target, while >> > responsive ones should make delay more like target ( < en_target) >> > >> > if (params->ecn && >> > =A0 =A0vars->ldelay <=3D params->ecn_target && >> > =A0 =A0INET_ECN_set_ce(skb)) { >> >> The orig is the same test, as params->ecn_target > vars->ldelay >> is equivalent to vars->ldelay <=3D params->ecn_target > > Yes, I must say I dont like the testing of a variable against a LIMIT > using > > =A0if (LIMIT > variable) > > I prefer for readability > > =A0if (variable < LIMIT) > > If you take a look at linux code, your form is very seldom used. > >> My own mental debate was whether to switch ecn from a bool to being a >> codel_time_t, >> and use a value of 0 for noecn and whatever for the ecn target value. > > Well, why not. It was the way I was leaning until I observed me dropping ecn enabled mosh, ssh, and babel packets , which tend to be small, so I started thinking in terms of dropping ecn packets on a graduated packet size scale after exceeding target. ESPECIALLY in case of overload you want command and control packets to get through. ... So I figured getting another RFC out was a good idea. O brave new world that has such new drop strategies in it! > > --=20 Dave T=E4ht http://www.bufferbloat.net/projects/cerowrt/wiki - "3.3.8-6 is out with fq_codel!"