Received: by 2002:ac0:a594:0:0:0:0:0 with SMTP id m20-v6csp4432049imm; Mon, 14 May 2018 07:30:18 -0700 (PDT) X-Google-Smtp-Source: AB8JxZqGN0/CgB0KO2UC8DlSHM/3Q7JYCWtYS6W92X8hZjQKNCoj0kZHF8hrAM7KNGzzQOjAXjxm X-Received: by 2002:a17:902:8217:: with SMTP id x23-v6mr10068844pln.380.1526308218841; Mon, 14 May 2018 07:30:18 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1526308218; cv=none; d=google.com; s=arc-20160816; b=HHnEDEhY0vooQZ0gnCYu2+PZDxLSJQJK1y7M0LKXDKw/OsHMqYltYLJ/H7T27Bm0Tv RFUYG41y0aNiRP6ow5vThMewALUSalGCxZaRb1OFV96vfFRJ3HvOScLZXbyJSBaUGm4n cHxjTMf8rEQ/Nwn+D/IFfpyqlmnsNeckcjRlF7kYOBwCHHGvQouCDr72C1wqDVqEgQLC muwEuRaeFODn9kL22RQUS8zqmp8kjP6h0unrOFUDr57YJI2Lr+P13VJZcD6Q7VWt9cX4 22NIUxHC6+cCNRIFMa4HfUSAxVVpZsj7uUqsC3pfiRPhH2BUYFZjTPIMWRoDHCm3cJCC w/ow== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:references:in-reply-to:message-id:date :subject:cc:to:from:arc-authentication-results; bh=ZVWjk6VznB3Jl515ZD1+2qsFEPceqDBHWS5XJyGIHB4=; b=XbUOJY02XDc5wMKyBJEzM4OZrEI1p87hRP2b9gZFGiSUyvqc1Cau3pKUBBg0qPYtgo gxRGU4YQZEEvb0xCC21/gnYnazVnuxx6v28jjhsmL2ADiHexaY8pGNoH7/jSWXubIb8/ 7TlabFKLo7cJehJOvRctObwTyfvh0TOvwZoJ+Xk7eSQJFjVZ0s3L74A0IgccgDSK3Me2 rQuNnY9qz5TIfSPnobvyR3QlHrNjJlX0B2yX5fdVHZlYdx/dawqWi8ZzL7IdKH5viqa+ WO7KQYOdBm6yUnuUjx1Hm1REnsccnTKsB0SUaSAUoxL6/ej2VzYUqO05eiPjAg0tJQ5K pcbg== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=mellanox.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id f9-v6si7583176pgr.123.2018.05.14.07.30.04; Mon, 14 May 2018 07:30:18 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=mellanox.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932498AbeENO2e (ORCPT + 99 others); Mon, 14 May 2018 10:28:34 -0400 Received: from mail-il-dmz.mellanox.com ([193.47.165.129]:35171 "EHLO mellanox.co.il" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753983AbeENO1m (ORCPT ); Mon, 14 May 2018 10:27:42 -0400 Received: from Internal Mail-Server by MTLPINE1 (envelope-from vladbu@mellanox.com) with ESMTPS (AES256-SHA encrypted); 14 May 2018 17:29:23 +0300 Received: from reg-r-vrt-018-180.mtr.labs.mlnx (reg-r-vrt-018-180.mtr.labs.mlnx [10.213.18.180]) by labmailer.mlnx (8.13.8/8.13.8) with ESMTP id w4EERXLd002750; Mon, 14 May 2018 17:27:35 +0300 From: Vlad Buslov To: netdev@vger.kernel.org Cc: davem@davemloft.net, jhs@mojatatu.com, xiyou.wangcong@gmail.com, jiri@resnulli.us, pablo@netfilter.org, kadlec@blackhole.kfki.hu, fw@strlen.de, ast@kernel.org, daniel@iogearbox.net, edumazet@google.com, vladbu@mellanox.com, keescook@chromium.org, linux-kernel@vger.kernel.org, netfilter-devel@vger.kernel.org, coreteam@netfilter.org, kliteyn@mellanox.com Subject: [PATCH 11/14] net: core: add new/replace rate estimator lock parameter Date: Mon, 14 May 2018 17:27:12 +0300 Message-Id: <1526308035-12484-12-git-send-email-vladbu@mellanox.com> X-Mailer: git-send-email 2.7.5 In-Reply-To: <1526308035-12484-1-git-send-email-vladbu@mellanox.com> References: <1526308035-12484-1-git-send-email-vladbu@mellanox.com> Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Extend rate estimator new and replace APIs with additional spinlock parameter used by lockless actions to protect rate_est pointer from concurrent modification. Signed-off-by: Vlad Buslov --- include/net/gen_stats.h | 2 ++ net/core/gen_estimator.c | 58 +++++++++++++++++++++++++++++++++++----------- net/netfilter/xt_RATEEST.c | 2 +- net/sched/act_api.c | 2 +- net/sched/act_police.c | 2 +- net/sched/sch_api.c | 2 ++ net/sched/sch_cbq.c | 4 ++-- net/sched/sch_drr.c | 4 ++-- net/sched/sch_hfsc.c | 4 ++-- net/sched/sch_htb.c | 4 ++-- net/sched/sch_qfq.c | 4 ++-- 11 files changed, 61 insertions(+), 27 deletions(-) diff --git a/include/net/gen_stats.h b/include/net/gen_stats.h index 0304ba2..d1ef63d 100644 --- a/include/net/gen_stats.h +++ b/include/net/gen_stats.h @@ -59,12 +59,14 @@ int gnet_stats_finish_copy(struct gnet_dump *d); int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct net_rate_estimator __rcu **rate_est, + spinlock_t *rate_est_lock, spinlock_t *stats_lock, seqcount_t *running, struct nlattr *opt); void gen_kill_estimator(struct net_rate_estimator __rcu **ptr); int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct net_rate_estimator __rcu **ptr, + spinlock_t *rate_est_lock, spinlock_t *stats_lock, seqcount_t *running, struct nlattr *opt); bool gen_estimator_active(struct net_rate_estimator __rcu **ptr); diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c index 98fd127..3512720 100644 --- a/net/core/gen_estimator.c +++ b/net/core/gen_estimator.c @@ -107,11 +107,43 @@ static void est_timer(struct timer_list *t) mod_timer(&est->timer, est->next_jiffies); } +static void __replace_estimator(struct net_rate_estimator __rcu **rate_est, + struct net_rate_estimator *new) +{ + struct net_rate_estimator *old = rcu_dereference_protected(*rate_est, + 1); + + if (old) { + del_timer_sync(&old->timer); + new->avbps = old->avbps; + new->avpps = old->avpps; + } + + new->next_jiffies = jiffies + ((HZ/4) << new->intvl_log); + timer_setup(&new->timer, est_timer, 0); + mod_timer(&new->timer, new->next_jiffies); + + rcu_assign_pointer(*rate_est, new); + + if (old) + kfree_rcu(old, rcu); +} + +static void replace_estimator(struct net_rate_estimator __rcu **rate_est, + struct net_rate_estimator *new, + spinlock_t *rate_est_lock) +{ + spin_lock(rate_est_lock); + __replace_estimator(rate_est, new); + spin_unlock(rate_est_lock); +} + /** * gen_new_estimator - create a new rate estimator * @bstats: basic statistics * @cpu_bstats: bstats per cpu * @rate_est: rate estimator statistics + * @rate_est_lock: rate_est lock (might be NULL) * @stats_lock: statistics lock * @running: qdisc running seqcount * @opt: rate estimator configuration TLV @@ -128,12 +160,13 @@ static void est_timer(struct timer_list *t) int gen_new_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct net_rate_estimator __rcu **rate_est, + spinlock_t *rate_est_lock, spinlock_t *stats_lock, seqcount_t *running, struct nlattr *opt) { struct gnet_estimator *parm = nla_data(opt); - struct net_rate_estimator *old, *est; + struct net_rate_estimator *est; struct gnet_stats_basic_packed b; int intvl_log; @@ -167,20 +200,15 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats, local_bh_enable(); est->last_bytes = b.bytes; est->last_packets = b.packets; - old = rcu_dereference_protected(*rate_est, 1); - if (old) { - del_timer_sync(&old->timer); - est->avbps = old->avbps; - est->avpps = old->avpps; - } - est->next_jiffies = jiffies + ((HZ/4) << intvl_log); - timer_setup(&est->timer, est_timer, 0); - mod_timer(&est->timer, est->next_jiffies); + if (rate_est_lock) + replace_estimator(rate_est, est, rate_est_lock); + else + /* If no spinlock argument provided, + * then assume that caller is already synchronized. + */ + __replace_estimator(rate_est, est); - rcu_assign_pointer(*rate_est, est); - if (old) - kfree_rcu(old, rcu); return 0; } EXPORT_SYMBOL(gen_new_estimator); @@ -209,6 +237,7 @@ EXPORT_SYMBOL(gen_kill_estimator); * @bstats: basic statistics * @cpu_bstats: bstats per cpu * @rate_est: rate estimator statistics + * @rate_est_lock: rate_est lock (might be NULL) * @stats_lock: statistics lock * @running: qdisc running seqcount (might be NULL) * @opt: rate estimator configuration TLV @@ -221,10 +250,11 @@ EXPORT_SYMBOL(gen_kill_estimator); int gen_replace_estimator(struct gnet_stats_basic_packed *bstats, struct gnet_stats_basic_cpu __percpu *cpu_bstats, struct net_rate_estimator __rcu **rate_est, + spinlock_t *rate_est_lock, spinlock_t *stats_lock, seqcount_t *running, struct nlattr *opt) { - return gen_new_estimator(bstats, cpu_bstats, rate_est, + return gen_new_estimator(bstats, cpu_bstats, rate_est, rate_est_lock, stats_lock, running, opt); } EXPORT_SYMBOL(gen_replace_estimator); diff --git a/net/netfilter/xt_RATEEST.c b/net/netfilter/xt_RATEEST.c index dec843c..8e79bd5 100644 --- a/net/netfilter/xt_RATEEST.c +++ b/net/netfilter/xt_RATEEST.c @@ -154,7 +154,7 @@ static int xt_rateest_tg_checkentry(const struct xt_tgchk_param *par) cfg.est.interval = info->interval; cfg.est.ewma_log = info->ewma_log; - ret = gen_new_estimator(&est->bstats, NULL, &est->rate_est, + ret = gen_new_estimator(&est->bstats, NULL, &est->rate_est, NULL, &est->lock, NULL, &cfg.opt); if (ret < 0) goto err2; diff --git a/net/sched/act_api.c b/net/sched/act_api.c index a5193dc..1dc092e 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -409,7 +409,7 @@ int tcf_idr_create(struct tc_action_net *tn, u32 index, struct nlattr *est, p->tcfa_tm.firstuse = 0; if (est) { err = gen_new_estimator(&p->tcfa_bstats, p->cpu_bstats, - &p->tcfa_rate_est, + &p->tcfa_rate_est, NULL, &p->tcfa_lock, NULL, est); if (err) goto err4; diff --git a/net/sched/act_police.c b/net/sched/act_police.c index 86d9417..c480d68 100644 --- a/net/sched/act_police.c +++ b/net/sched/act_police.c @@ -133,7 +133,7 @@ static int tcf_act_police_init(struct net *net, struct nlattr *nla, if (est) { err = gen_replace_estimator(&police->tcf_bstats, NULL, - &police->tcf_rate_est, + &police->tcf_rate_est, NULL, &police->tcf_lock, NULL, est); if (err) diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 106dae7e..de6a297 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -1178,6 +1178,7 @@ static struct Qdisc *qdisc_create(struct net_device *dev, sch->cpu_bstats, &sch->rate_est, NULL, + NULL, running, tca[TCA_RATE]); if (err) { @@ -1253,6 +1254,7 @@ static int qdisc_change(struct Qdisc *sch, struct nlattr **tca, sch->cpu_bstats, &sch->rate_est, NULL, + NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); } diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c index f42025d..2a7ff53 100644 --- a/net/sched/sch_cbq.c +++ b/net/sched/sch_cbq.c @@ -1503,7 +1503,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, - &cl->rate_est, + &cl->rate_est, NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); @@ -1605,7 +1605,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t if (tca[TCA_RATE]) { err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est, - NULL, + NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); if (err) { diff --git a/net/sched/sch_drr.c b/net/sched/sch_drr.c index e0b0cf8..0896e23 100644 --- a/net/sched/sch_drr.c +++ b/net/sched/sch_drr.c @@ -95,7 +95,7 @@ static int drr_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (cl != NULL) { if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, - &cl->rate_est, + &cl->rate_est, NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); @@ -129,7 +129,7 @@ static int drr_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, &cl->rate_est, - NULL, + NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); if (err) { diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index 3ae9877..f341324 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -972,7 +972,7 @@ hfsc_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, - &cl->rate_est, + &cl->rate_est, NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); @@ -1042,7 +1042,7 @@ hfsc_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (tca[TCA_RATE]) { err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est, - NULL, + NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); if (err) { diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c index 2a4ab7c..acc0355 100644 --- a/net/sched/sch_htb.c +++ b/net/sched/sch_htb.c @@ -1407,7 +1407,7 @@ static int htb_change_class(struct Qdisc *sch, u32 classid, if (htb_rate_est || tca[TCA_RATE]) { err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est, - NULL, + NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE] ? : &est.nla); if (err) { @@ -1473,7 +1473,7 @@ static int htb_change_class(struct Qdisc *sch, u32 classid, } else { if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, - &cl->rate_est, + &cl->rate_est, NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index bb1a9c1..8026c1e 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -461,7 +461,7 @@ static int qfq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (cl != NULL) { /* modify existing class */ if (tca[TCA_RATE]) { err = gen_replace_estimator(&cl->bstats, NULL, - &cl->rate_est, + &cl->rate_est, NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); @@ -488,7 +488,7 @@ static int qfq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, if (tca[TCA_RATE]) { err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est, - NULL, + NULL, NULL, qdisc_root_sleeping_running(sch), tca[TCA_RATE]); if (err) -- 2.7.5