From patchwork Wed Nov 8 21:09:57 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Dave Taht X-Patchwork-Id: 835994 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="uerCiDJq"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 3yXJpw2vc3z9rxj for ; Thu, 9 Nov 2017 08:10:16 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753646AbdKHVKP (ORCPT ); Wed, 8 Nov 2017 16:10:15 -0500 Received: from mail-pg0-f65.google.com ([74.125.83.65]:45040 "EHLO mail-pg0-f65.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753058AbdKHVKK (ORCPT ); Wed, 8 Nov 2017 16:10:10 -0500 Received: by mail-pg0-f65.google.com with SMTP id j3so2903318pga.1 for ; Wed, 08 Nov 2017 13:10:10 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=qhFgC4xqORzUyKCa0L5GL+tThK3+xx/0kYw8p2YCCjw=; b=uerCiDJq+DzvJscAVM5iTpOCPQPr/ifwpQqrWxNSnfFoFvK3XjbxFJtW7tvKe7Gpcl zJIuLE2PASxGGxbnDCkuY6i6riom8S4eqK+Fe+yJl+h0fbEC+hodgeDkVS8YcX+E9bv5 Mez85Fv0iJ+h4zMxI5FcWpTWluq9drNBzmBhbCNnRnQM8Dp1gvT2GsWsXmzwkvl6h2Cu wPAC64tMvGG6VMWzU5tj1GpWM3QI7Tbx34lqQOiddJOfUSvMhx/AHHuMoSi/90LwbuiK TpCqbFqo4LvLwkF9ltpx4inNPUIPfbyOY9CGcuLD1x27cLPE2yGRd9P2ke1/htiqe+KN EaAA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=qhFgC4xqORzUyKCa0L5GL+tThK3+xx/0kYw8p2YCCjw=; b=XbEDWXdm/+ni7xNfII/PnwB3FWaZL9cG4hEo95lIHDb/4OR6SUj9rgDqgybJGKfDG9 zfKXHZZkT2aItpvtzjCMO+aKUWnq/V74r9J2tX28wMHicO0O1T9aEbBkbUFuyH3483fL 8kiGcfSR7VHfvpHdnIXGVlZxlg55eup2Ex6NnTdfIV+JQJzvNWa6NzOKGwGc7cTmaV8s DZumpUzpC56l6g+1M8yEDYN+zQJ4ORWTDcvvuuhwYVP87IvVQk1FsAcXZbf9gj0F8y7t PXNFEI9QXJAhYjHzZH/QD1Yc998CRdqydVZ7JBKwHM2MzAgP+Ir/d+2OfG8RHFekeLJK dS6Q== X-Gm-Message-State: AJaThX6b7tgJEMRJkcDOMlv/iVC5NoWyiB6TxuIC9LFUicGUuCq5qoJ3 Z7IRTw56Axvy9mOAxHlokLWYdw== X-Google-Smtp-Source: ABhQp+S8hXIEXA8pfHMdp0OyMHHXvY39aNVS3BdALr8/Vdss3ckYAY6JebVEZe5Mad4M7a77L3Am2g== X-Received: by 10.101.87.139 with SMTP id b11mr1724604pgr.314.1510175409571; Wed, 08 Nov 2017 13:10:09 -0800 (PST) Received: from nemesis.lab.teklibre.com ([2603:3024:1536:86f0:2e0:4cff:fec1:1206]) by smtp.gmail.com with ESMTPSA id s88sm10395079pfe.36.2017.11.08.13.10.08 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Wed, 08 Nov 2017 13:10:08 -0800 (PST) From: Dave Taht To: netdev@vger.kernel.org Cc: stephen@networkplumber.org, Dave Taht Subject: [PATCH v2 net-next 1/3] netem: convert to qdisc_watchdog_schedule_ns Date: Wed, 8 Nov 2017 13:09:57 -0800 Message-Id: <1510175399-7404-2-git-send-email-dave.taht@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> References: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Upgrade the internal netem scheduler to use nanoseconds rather than ticks throughout. Convert to and from the std "ticks" userspace api automatically, while allowing for finer grained scheduling to take place. Signed-off-by: Dave Taht --- net/sched/sch_netem.c | 56 +++++++++++++++++++++++++-------------------------- 1 file changed, 28 insertions(+), 28 deletions(-) diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index db0228a..5559ad1 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -77,8 +77,8 @@ struct netem_sched_data { struct qdisc_watchdog watchdog; - psched_tdiff_t latency; - psched_tdiff_t jitter; + s64 latency; + s64 jitter; u32 loss; u32 ecn; @@ -145,7 +145,7 @@ struct netem_sched_data { * we save skb->tstamp value in skb->cb[] before destroying it. */ struct netem_skb_cb { - psched_time_t time_to_send; + u64 time_to_send; }; static inline struct netem_skb_cb *netem_skb_cb(struct sk_buff *skb) @@ -305,11 +305,11 @@ static bool loss_event(struct netem_sched_data *q) * std deviation sigma. Uses table lookup to approximate the desired * distribution, and a uniformly-distributed pseudo-random source. */ -static psched_tdiff_t tabledist(psched_tdiff_t mu, psched_tdiff_t sigma, - struct crndstate *state, - const struct disttable *dist) +static s64 tabledist(s64 mu, s64 sigma, + struct crndstate *state, + const struct disttable *dist) { - psched_tdiff_t x; + s64 x; long t; u32 rnd; @@ -332,10 +332,10 @@ static psched_tdiff_t tabledist(psched_tdiff_t mu, psched_tdiff_t sigma, return x / NETEM_DIST_SCALE + (sigma / NETEM_DIST_SCALE) * t + mu; } -static psched_time_t packet_len_2_sched_time(unsigned int len, struct netem_sched_data *q) +static s64 packet_len_2_sched_time(unsigned int len, + struct netem_sched_data *q) { - u64 ticks; - + s64 offset; len += q->packet_overhead; if (q->cell_size) { @@ -345,11 +345,9 @@ static psched_time_t packet_len_2_sched_time(unsigned int len, struct netem_sche cells++; len = cells * (q->cell_size + q->cell_overhead); } - - ticks = (u64)len * NSEC_PER_SEC; - - do_div(ticks, q->rate); - return PSCHED_NS2TICKS(ticks); + offset = (s64)len * NSEC_PER_SEC; + do_div(offset, q->rate); + return offset; } static void tfifo_reset(struct Qdisc *sch) @@ -369,7 +367,7 @@ static void tfifo_reset(struct Qdisc *sch) static void tfifo_enqueue(struct sk_buff *nskb, struct Qdisc *sch) { struct netem_sched_data *q = qdisc_priv(sch); - psched_time_t tnext = netem_skb_cb(nskb)->time_to_send; + u64 tnext = netem_skb_cb(nskb)->time_to_send; struct rb_node **p = &q->t_root.rb_node, *parent = NULL; while (*p) { @@ -515,13 +513,13 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch, if (q->gap == 0 || /* not doing reordering */ q->counter < q->gap - 1 || /* inside last reordering gap */ q->reorder < get_crandom(&q->reorder_cor)) { - psched_time_t now; - psched_tdiff_t delay; + u64 now; + s64 delay; delay = tabledist(q->latency, q->jitter, &q->delay_cor, q->delay_dist); - now = psched_get_time(); + now = ktime_get_ns(); if (q->rate) { struct netem_skb_cb *last = NULL; @@ -547,7 +545,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch, * from delay. */ delay -= last->time_to_send - now; - delay = max_t(psched_tdiff_t, 0, delay); + delay = max_t(s64, 0, delay); now = last->time_to_send; } @@ -562,7 +560,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch, * Do re-ordering by putting one out of N packets at the front * of the queue. */ - cb->time_to_send = psched_get_time(); + cb->time_to_send = ktime_get_ns(); q->counter = 0; netem_enqueue_skb_head(&sch->q, skb); @@ -609,13 +607,13 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch) } p = rb_first(&q->t_root); if (p) { - psched_time_t time_to_send; + u64 time_to_send; skb = rb_to_skb(p); /* if more time remaining? */ time_to_send = netem_skb_cb(skb)->time_to_send; - if (time_to_send <= psched_get_time()) { + if (time_to_send <= ktime_get_ns()) { rb_erase(p, &q->t_root); sch->q.qlen--; @@ -659,7 +657,7 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch) if (skb) goto deliver; } - qdisc_watchdog_schedule(&q->watchdog, time_to_send); + qdisc_watchdog_schedule_ns(&q->watchdog, time_to_send); } if (q->qdisc) { @@ -888,8 +886,8 @@ static int netem_change(struct Qdisc *sch, struct nlattr *opt) sch->limit = qopt->limit; - q->latency = qopt->latency; - q->jitter = qopt->jitter; + q->latency = PSCHED_TICKS2NS(qopt->latency); + q->jitter = PSCHED_TICKS2NS(qopt->jitter); q->limit = qopt->limit; q->gap = qopt->gap; q->counter = 0; @@ -1011,8 +1009,10 @@ static int netem_dump(struct Qdisc *sch, struct sk_buff *skb) struct tc_netem_corrupt corrupt; struct tc_netem_rate rate; - qopt.latency = q->latency; - qopt.jitter = q->jitter; + qopt.latency = min_t(psched_tdiff_t, PSCHED_NS2TICKS(q->latency), + UINT_MAX); + qopt.jitter = min_t(psched_tdiff_t, PSCHED_NS2TICKS(q->jitter), + UINT_MAX); qopt.limit = q->limit; qopt.loss = q->loss; qopt.gap = q->gap; From patchwork Wed Nov 8 21:09:58 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Dave Taht X-Patchwork-Id: 835997 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="qvZYx5u4"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 3yXJrz11C7z9s0Z for ; Thu, 9 Nov 2017 08:12:03 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753231AbdKHVMB (ORCPT ); Wed, 8 Nov 2017 16:12:01 -0500 Received: from mail-pf0-f194.google.com ([209.85.192.194]:54427 "EHLO mail-pf0-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753202AbdKHVKL (ORCPT ); Wed, 8 Nov 2017 16:10:11 -0500 Received: by mail-pf0-f194.google.com with SMTP id n89so2620670pfk.11 for ; Wed, 08 Nov 2017 13:10:11 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=H8USETst+KqaaFCQOYqfQsRb5FIjeJ6csJqygXEpAIE=; b=qvZYx5u4/qgnkDqtVqatNNuqAIN6nUz4H7hOpN0PokuLZj2KXtWzm2OLq/+pabD8Iu xpI+xrrPtpLGzClBmJnSq0Cw5pqD5jarRr/C+3vDfj4k3LVRXg+3D0rPpH2jfTVXk95Y SWz67ORItFqwN01oPRyr/p2zrMSSgdLawXmk80R2vDT4bBJuro8Rz4TGQx5E+PXtP6Q0 9Np1yI54wzxI2hcyl6aHNsNFoVTeYIAMrIjxaklLvILxS6tDsYHTv6p4WuhpP47Vy0yF CfDK9QEZn4WqPI37aW57GBm7rThURJYx2EUaEPzbBLiz7qXRuQn4IRyQ3FLV1Up6Iss0 4hkg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=H8USETst+KqaaFCQOYqfQsRb5FIjeJ6csJqygXEpAIE=; b=NuO025E33T7oGeESpOeYwS1bf2WtGF7gV/nceh4dGTsvU8WS6fBHWIkLxNK+cqMT95 s+SoRe82PRH1Hd/lPUi8zq0j0fpmW+vs9EYNVLvhcezR79ocRurROQkwbipZPjMtuxMr oU8Tj1VbDP40EgTiuNBcl3sBBRR/xbuVNz7zc7GMJgWvAvH0+/H//G3MOSnOYBZmeStn 8w9BU8POO8JtNZU0JhJoexOReoF2yvXPIU7LZYVsdzYsWq3tLCzhjvAOE68Blk1Wypob xTqhNnVA6LW2HdAvZLcXt/hGPwSO+UFdc57OpzcYjIeONQ4xC1Z/4wOBea/93pn48ZQs J+eA== X-Gm-Message-State: AJaThX6FQ3WfrUY5z4BDsv3r3Vd7OJoJUwS3eT1QRBat6FoSMAkrENOU shp2CoiHyEyAZFgEPNh9a0EJVw== X-Google-Smtp-Source: ABhQp+Tv/a+zxGqam5y+aCWXvMrYVEF/XOWtv/v8To0T1ZFhhF1aruV7cp1Q5CnJZLYYIkRb330Pug== X-Received: by 10.159.234.10 with SMTP id be10mr1624553plb.386.1510175410808; Wed, 08 Nov 2017 13:10:10 -0800 (PST) Received: from nemesis.lab.teklibre.com ([2603:3024:1536:86f0:2e0:4cff:fec1:1206]) by smtp.gmail.com with ESMTPSA id s88sm10395079pfe.36.2017.11.08.13.10.09 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Wed, 08 Nov 2017 13:10:09 -0800 (PST) From: Dave Taht To: netdev@vger.kernel.org Cc: stephen@networkplumber.org, Dave Taht Subject: [PATCH v2 net-next 2/3] netem: add uapi to express delay and jitter in nanoseconds Date: Wed, 8 Nov 2017 13:09:58 -0800 Message-Id: <1510175399-7404-3-git-send-email-dave.taht@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> References: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org netem userspace has long relied on a horrible /proc/net/psched hack to translate the current notion of "ticks" to nanoseconds. Expressing latency and jitter instead, in well defined nanoseconds, increases the dynamic range of emulated delays and jitter in netem. It will also ease a transition where reducing a tick to nsec equivalence would constrain the max delay in prior versions of netem to only 4.3 seconds. Signed-off-by: Dave Taht Reviewed-by: Eric Dumazet --- include/uapi/linux/pkt_sched.h | 2 ++ net/sched/sch_netem.c | 14 ++++++++++++++ 2 files changed, 16 insertions(+) diff --git a/include/uapi/linux/pkt_sched.h b/include/uapi/linux/pkt_sched.h index 6a2c5ea..8fe6d18 100644 --- a/include/uapi/linux/pkt_sched.h +++ b/include/uapi/linux/pkt_sched.h @@ -537,6 +537,8 @@ enum { TCA_NETEM_ECN, TCA_NETEM_RATE64, TCA_NETEM_PAD, + TCA_NETEM_LATENCY64, + TCA_NETEM_JITTER64, __TCA_NETEM_MAX, }; diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index 5559ad1..ef63ae4 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -819,6 +819,8 @@ static const struct nla_policy netem_policy[TCA_NETEM_MAX + 1] = { [TCA_NETEM_LOSS] = { .type = NLA_NESTED }, [TCA_NETEM_ECN] = { .type = NLA_U32 }, [TCA_NETEM_RATE64] = { .type = NLA_U64 }, + [TCA_NETEM_LATENCY64] = { .type = NLA_S64 }, + [TCA_NETEM_JITTER64] = { .type = NLA_S64 }, }; static int parse_attr(struct nlattr *tb[], int maxtype, struct nlattr *nla, @@ -916,6 +918,12 @@ static int netem_change(struct Qdisc *sch, struct nlattr *opt) q->rate = max_t(u64, q->rate, nla_get_u64(tb[TCA_NETEM_RATE64])); + if (tb[TCA_NETEM_LATENCY64]) + q->latency = nla_get_s64(tb[TCA_NETEM_LATENCY64]); + + if (tb[TCA_NETEM_JITTER64]) + q->jitter = nla_get_s64(tb[TCA_NETEM_JITTER64]); + if (tb[TCA_NETEM_ECN]) q->ecn = nla_get_u32(tb[TCA_NETEM_ECN]); @@ -1020,6 +1028,12 @@ static int netem_dump(struct Qdisc *sch, struct sk_buff *skb) if (nla_put(skb, TCA_OPTIONS, sizeof(qopt), &qopt)) goto nla_put_failure; + if (nla_put(skb, TCA_NETEM_LATENCY64, sizeof(q->latency), &q->latency)) + goto nla_put_failure; + + if (nla_put(skb, TCA_NETEM_JITTER64, sizeof(q->jitter), &q->jitter)) + goto nla_put_failure; + cor.delay_corr = q->delay_cor.rho; cor.loss_corr = q->loss_cor.rho; cor.dup_corr = q->dup_cor.rho; From patchwork Wed Nov 8 21:09:59 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Dave Taht X-Patchwork-Id: 835998 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b="XVuf6/5S"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 3yXJs250rVz9s0Z for ; Thu, 9 Nov 2017 08:12:06 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753136AbdKHVMA (ORCPT ); Wed, 8 Nov 2017 16:12:00 -0500 Received: from mail-pg0-f67.google.com ([74.125.83.67]:54137 "EHLO mail-pg0-f67.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753583AbdKHVKM (ORCPT ); Wed, 8 Nov 2017 16:10:12 -0500 Received: by mail-pg0-f67.google.com with SMTP id s2so2873643pge.10 for ; Wed, 08 Nov 2017 13:10:12 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=0rlF7TvnICYd9T9w5yMRAfI/EfpQx97fS3c+iJrpmIU=; b=XVuf6/5SPakhWWTrmkx8gZ9CnXi9J02uHMvBOmqAcHt2qC8JVuuU7N6ineeXvi2FMr lOgKYKdTmq3m9vtFarxR/eFRYM4IqhK1gD/q1pootuV4zIxSqqLuAa0a6FV+Oe8o95g6 h1FAQYlNKIWwpbGgLreKNEwgiuTBb3ORl0GhpDVRYFjpCMhB4Q8Rbd6vVNHdbN6lKV5d iRg4/y2N/pEvaz6AecDdRbRbDpEkaV8tyDFaex5YzkXFOBMi6RQIyQGFx+IekOfMVNbM BeVfiFPrtB81FzrnQghEwCdMWyo8GSk6nDqX7JntNk/tOPG71PCr/R+wltbqTzU5ByqQ Ft2g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=0rlF7TvnICYd9T9w5yMRAfI/EfpQx97fS3c+iJrpmIU=; b=FgkSywM2kPz0Mz2HdzLgMwgkOZDhhuDOgleCwAmXnYzGgQBi784pholc0Hfqc6JXwa SdM2ICj88mwbc+G+/RLj4zUL4fX6F2+vJ2AKLfs2x65tR7h3cXW49gWGLY/3qQJpVXHp 3mH3DAMMI/5O+DCUG2GGMv2UOL90zofIAaE1K+kvts37WytGMrmRgJrBvuZYxrpXipa3 MvdOxPP03GZLcrR7gJbrnhxIZQ6+M186SoC7nmBU4Y6iwXSPiyamHKw04w3r5haKp2HZ XQ5idB30F8UE9y5QQE/a7c0WiOXzmJ4Kg4xdr/j47nGlNur6dXKzhclmNZs0OLEBdu6B HgKQ== X-Gm-Message-State: AJaThX4G6+zWJLCTPzPZ0m8eYULxRNv2uruGmkrW/1PyhxXEvKzGNEMu MlQfh5VoGFBIEhAsfLNdnkf4SA== X-Google-Smtp-Source: ABhQp+Soty4WIdBA5DyeY4fL//G4N4/+olunZFC3xjAUy6RVutwQq0S+fK5SKcrU09w4TBs93F5SdA== X-Received: by 10.84.233.70 with SMTP id k6mr1648397plt.134.1510175411777; Wed, 08 Nov 2017 13:10:11 -0800 (PST) Received: from nemesis.lab.teklibre.com ([2603:3024:1536:86f0:2e0:4cff:fec1:1206]) by smtp.gmail.com with ESMTPSA id s88sm10395079pfe.36.2017.11.08.13.10.10 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Wed, 08 Nov 2017 13:10:11 -0800 (PST) From: Dave Taht To: netdev@vger.kernel.org Cc: stephen@networkplumber.org, Dave Taht Subject: [PATCH v2 net-next 3/3] netem: support delivering packets in delayed time slots Date: Wed, 8 Nov 2017 13:09:59 -0800 Message-Id: <1510175399-7404-4-git-send-email-dave.taht@gmail.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> References: <1510175399-7404-1-git-send-email-dave.taht@gmail.com> Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Slotting is a crude approximation of the behaviors of shared media such as cable, wifi, and LTE, which gather up a bunch of packets within a varying delay window and deliver them, relative to that, nearly all at once. It works within the existing loss, duplication, jitter and delay parameters of netem. Some amount of inherent latency must be specified, regardless. The new "slot" parameter specifies a minimum and maximum delay between transmission attempts. The "bytes" and "packets" parameters can be used to limit the amount of information transferred per slot. Examples of use: tc qdisc add dev eth0 root netem delay 200us \ slot 800us 10ms bytes 64k packets 42 A more correct example, using stacked netem instances and a packet limit to emulate a tail drop wifi queue with slots and variable packet delivery, with a 200Mbit isochronous underlying rate, and 20ms path delay: tc qdisc add dev eth0 root handle 1: netem delay 20ms rate 200mbit \ limit 10000 tc qdisc add dev eth0 parent 1:1 handle 10:1 netem delay 200us \ slot 800us 10ms bytes 64k packets 42 limit 512 Signed-off-by: Dave Taht --- include/uapi/linux/pkt_sched.h | 8 +++++ net/sched/sch_netem.c | 76 ++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 81 insertions(+), 3 deletions(-) diff --git a/include/uapi/linux/pkt_sched.h b/include/uapi/linux/pkt_sched.h index 8fe6d18..af3cc2f 100644 --- a/include/uapi/linux/pkt_sched.h +++ b/include/uapi/linux/pkt_sched.h @@ -539,6 +539,7 @@ enum { TCA_NETEM_PAD, TCA_NETEM_LATENCY64, TCA_NETEM_JITTER64, + TCA_NETEM_SLOT, __TCA_NETEM_MAX, }; @@ -576,6 +577,13 @@ struct tc_netem_rate { __s32 cell_overhead; }; +struct tc_netem_slot { + __s64 min_delay; /* nsec */ + __s64 max_delay; + __s32 max_packets; + __s32 max_bytes; +}; + enum { NETEM_LOSS_UNSPEC, NETEM_LOSS_GI, /* General Intuitive - 4 state model */ diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index ef63ae4..b697f89 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -135,6 +135,13 @@ struct netem_sched_data { u32 a5; /* p23 used only in 4-states */ } clg; + struct tc_netem_slot slot_config; + struct slotstate { + u64 slot_next; + s32 packets_left; + s32 bytes_left; + } slot; + }; /* Time stamp put into socket buffer control block @@ -591,6 +598,20 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch, return NET_XMIT_SUCCESS; } +/* Delay the next round with a new future slot with a + * correct number of bytes and packets. + */ + +static void get_slot_next(struct netem_sched_data *q, u64 now) +{ + q->slot.slot_next = now + q->slot_config.min_delay + + (prandom_u32() * + (q->slot_config.max_delay - + q->slot_config.min_delay) >> 32); + q->slot.packets_left = q->slot_config.max_packets; + q->slot.bytes_left = q->slot_config.max_bytes; +} + static struct sk_buff *netem_dequeue(struct Qdisc *sch) { struct netem_sched_data *q = qdisc_priv(sch); @@ -608,14 +629,17 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch) p = rb_first(&q->t_root); if (p) { u64 time_to_send; + u64 now = ktime_get_ns(); skb = rb_to_skb(p); /* if more time remaining? */ time_to_send = netem_skb_cb(skb)->time_to_send; - if (time_to_send <= ktime_get_ns()) { - rb_erase(p, &q->t_root); + if (q->slot.slot_next && q->slot.slot_next < time_to_send) + get_slot_next(q, now); + if (time_to_send <= now && q->slot.slot_next <= now) { + rb_erase(p, &q->t_root); sch->q.qlen--; qdisc_qstats_backlog_dec(sch, skb); skb->next = NULL; @@ -634,6 +658,14 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch) skb->tstamp = 0; #endif + if (q->slot.slot_next) { + q->slot.packets_left--; + q->slot.bytes_left -= qdisc_pkt_len(skb); + if (q->slot.packets_left <= 0 || + q->slot.bytes_left <= 0) + get_slot_next(q, now); + } + if (q->qdisc) { unsigned int pkt_len = qdisc_pkt_len(skb); struct sk_buff *to_free = NULL; @@ -657,7 +689,12 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch) if (skb) goto deliver; } - qdisc_watchdog_schedule_ns(&q->watchdog, time_to_send); + + if (q->slot.slot_next > now) + qdisc_watchdog_schedule_ns(&q->watchdog, + q->slot.slot_next); + else + qdisc_watchdog_schedule_ns(&q->watchdog, time_to_send); } if (q->qdisc) { @@ -688,6 +725,7 @@ static void dist_free(struct disttable *d) * Distribution data is a variable size payload containing * signed 16 bit values. */ + static int get_dist_table(struct Qdisc *sch, const struct nlattr *attr) { struct netem_sched_data *q = qdisc_priv(sch); @@ -718,6 +756,23 @@ static int get_dist_table(struct Qdisc *sch, const struct nlattr *attr) return 0; } +static void get_slot(struct netem_sched_data *q, const struct nlattr *attr) +{ + const struct tc_netem_slot *c = nla_data(attr); + + q->slot_config = *c; + if (q->slot_config.max_packets == 0) + q->slot_config.max_packets = INT_MAX; + if (q->slot_config.max_bytes == 0) + q->slot_config.max_bytes = INT_MAX; + q->slot.packets_left = q->slot_config.max_packets; + q->slot.bytes_left = q->slot_config.max_bytes; + if (q->slot_config.min_delay | q->slot_config.max_delay) + q->slot.slot_next = ktime_get_ns(); + else + q->slot.slot_next = 0; +} + static void get_correlation(struct netem_sched_data *q, const struct nlattr *attr) { const struct tc_netem_corr *c = nla_data(attr); @@ -821,6 +876,7 @@ static const struct nla_policy netem_policy[TCA_NETEM_MAX + 1] = { [TCA_NETEM_RATE64] = { .type = NLA_U64 }, [TCA_NETEM_LATENCY64] = { .type = NLA_S64 }, [TCA_NETEM_JITTER64] = { .type = NLA_S64 }, + [TCA_NETEM_SLOT] = { .len = sizeof(struct tc_netem_slot) }, }; static int parse_attr(struct nlattr *tb[], int maxtype, struct nlattr *nla, @@ -927,6 +983,9 @@ static int netem_change(struct Qdisc *sch, struct nlattr *opt) if (tb[TCA_NETEM_ECN]) q->ecn = nla_get_u32(tb[TCA_NETEM_ECN]); + if (tb[TCA_NETEM_SLOT]) + get_slot(q, tb[TCA_NETEM_SLOT]); + return ret; } @@ -1016,6 +1075,7 @@ static int netem_dump(struct Qdisc *sch, struct sk_buff *skb) struct tc_netem_reorder reorder; struct tc_netem_corrupt corrupt; struct tc_netem_rate rate; + struct tc_netem_slot slot; qopt.latency = min_t(psched_tdiff_t, PSCHED_NS2TICKS(q->latency), UINT_MAX); @@ -1070,6 +1130,16 @@ static int netem_dump(struct Qdisc *sch, struct sk_buff *skb) if (dump_loss_model(q, skb) != 0) goto nla_put_failure; + if (q->slot_config.min_delay | q->slot_config.max_delay) { + slot = q->slot_config; + if (slot.max_packets == INT_MAX) + slot.max_packets = 0; + if (slot.max_bytes == INT_MAX) + slot.max_bytes = 0; + if (nla_put(skb, TCA_NETEM_SLOT, sizeof(slot), &slot)) + goto nla_put_failure; + } + return nla_nest_end(skb, nla); nla_put_failure: