Merge branch 'sch_act_lockless'

Eric Dumazet says:

====================
net_sched: act: lockless operation

As mentioned by Alexei last week in Budapest, it is a bit weird
to take a spinlock in order to drop a packet in a tc filter...

Lets add percpu infra for tc actions and use it for gact & mirred.

Before changes, my host with 8 RX queues was handling 5 Mpps with gact,
and more than 11 Mpps after.

Mirred change is not yet visible if ifb+qdisc is used, as ifb is
not yet multi queue enabled, but is a step forward.
====================

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

4 files changed, 40 insertions, 15 deletions

diff --git a/include/net/act_api.h b/include/net/act_api.h
index 3ee4c92afd1b..8d2a707a9e87 100644
--- a/include/net/act_api.h
+++ b/include/net/act_api.h
@@ -21,6 +21,8 @@ struct tcf_common {
 	struct gnet_stats_rate_est64	tcfc_rate_est;
 	spinlock_t			tcfc_lock;
 	struct rcu_head			tcfc_rcu;
+	struct gnet_stats_basic_cpu __percpu *cpu_bstats;
+	struct gnet_stats_queue __percpu *cpu_qstats;
 };
 #define tcf_head	common.tcfc_head
 #define tcf_index	common.tcfc_index
@@ -68,6 +70,17 @@ static inline void tcf_hashinfo_destroy(struct tcf_hashinfo *hf)
 	kfree(hf->htab);
 }
 
+/* Update lastuse only if needed, to avoid dirtying a cache line.
+ * We use a temp variable to avoid fetching jiffies twice.
+ */
+static inline void tcf_lastuse_update(struct tcf_t *tm)
+{
+	unsigned long now = jiffies;
+
+	if (tm->lastuse != now)
+		tm->lastuse = now;
+}
+
 #ifdef CONFIG_NET_CLS_ACT
 
 #define ACT_P_CREATED 1
@@ -103,7 +116,7 @@ int tcf_hash_release(struct tc_action *a, int bind);
 u32 tcf_hash_new_index(struct tcf_hashinfo *hinfo);
 int tcf_hash_check(u32 index, struct tc_action *a, int bind);
 int tcf_hash_create(u32 index, struct nlattr *est, struct tc_action *a,
-		    int size, int bind);
+		    int size, int bind, bool cpustats);
 void tcf_hash_cleanup(struct tc_action *a, struct nlattr *est);
 void tcf_hash_insert(struct tc_action *a);
 
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 2738f6f87908..2eab08c38e32 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -513,17 +513,20 @@ static inline void bstats_update(struct gnet_stats_basic_packed *bstats,
 	bstats->packets += skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1;
 }
 
-static inline void qdisc_bstats_update_cpu(struct Qdisc *sch,
-					   const struct sk_buff *skb)
+static inline void bstats_cpu_update(struct gnet_stats_basic_cpu *bstats,
+				     const struct sk_buff *skb)
 {
-	struct gnet_stats_basic_cpu *bstats =
-				this_cpu_ptr(sch->cpu_bstats);
-
 	u64_stats_update_begin(&bstats->syncp);
 	bstats_update(&bstats->bstats, skb);
 	u64_stats_update_end(&bstats->syncp);
 }
 
+static inline void qdisc_bstats_cpu_update(struct Qdisc *sch,
+					   const struct sk_buff *skb)
+{
+	bstats_cpu_update(this_cpu_ptr(sch->cpu_bstats), skb);
+}
+
 static inline void qdisc_bstats_update(struct Qdisc *sch,
 				       const struct sk_buff *skb)
 {
@@ -547,16 +550,24 @@ static inline void __qdisc_qstats_drop(struct Qdisc *sch, int count)
 	sch->qstats.drops += count;
 }
 
-static inline void qdisc_qstats_drop(struct Qdisc *sch)
+static inline void qstats_drop_inc(struct gnet_stats_queue *qstats)
 {
-	sch->qstats.drops++;
+	qstats->drops++;
 }
 
-static inline void qdisc_qstats_drop_cpu(struct Qdisc *sch)
+static inline void qstats_overlimit_inc(struct gnet_stats_queue *qstats)
 {
-	struct gnet_stats_queue *qstats = this_cpu_ptr(sch->cpu_qstats);
+	qstats->overlimits++;
+}
 
-	qstats->drops++;
+static inline void qdisc_qstats_drop(struct Qdisc *sch)
+{
+	qstats_drop_inc(&sch->qstats);
+}
+
+static inline void qdisc_qstats_cpu_drop(struct Qdisc *sch)
+{
+	qstats_drop_inc(this_cpu_ptr(sch->cpu_qstats));
 }
 
 static inline void qdisc_qstats_overlimit(struct Qdisc *sch)
diff --git a/include/net/tc_act/tc_gact.h b/include/net/tc_act/tc_gact.h
index 9fc9b578908a..592a6bc02b0b 100644
--- a/include/net/tc_act/tc_gact.h
+++ b/include/net/tc_act/tc_gact.h
@@ -6,9 +6,10 @@
 struct tcf_gact {
 	struct tcf_common	common;
 #ifdef CONFIG_GACT_PROB
-        u16			tcfg_ptype;
-        u16			tcfg_pval;
-        int			tcfg_paction;
+	u16			tcfg_ptype;
+	u16			tcfg_pval;
+	int			tcfg_paction;
+	atomic_t		packets;
 #endif
 };
 #define to_gact(a) \
diff --git a/include/net/tc_act/tc_mirred.h b/include/net/tc_act/tc_mirred.h
index 4dd77a1c106b..dae96bae1c19 100644
--- a/include/net/tc_act/tc_mirred.h
+++ b/include/net/tc_act/tc_mirred.h
@@ -8,7 +8,7 @@ struct tcf_mirred {
 	int			tcfm_eaction;
 	int			tcfm_ifindex;
 	int			tcfm_ok_push;
-	struct net_device	*tcfm_dev;
+	struct net_device __rcu	*tcfm_dev;
 	struct list_head	tcfm_list;
 };
 #define to_mirred(a) \