Revert: "net: sched: put back q.qlen into a single location"

This revert commit 46b1c18f ("net: sched: put back q.qlen into
a single location").
After the previous patch, when a NOLOCK qdisc is enslaved to a
locking qdisc it switches to global stats accounting. As a consequence,
when a classful qdisc accesses directly a child qdisc's qlen, such
qdisc is not doing per CPU accounting and qlen value is consistent.

In the control path nobody uses directly qlen since commit
e5f0e8f8 ("net: sched: introduce and use qdisc tree flush/purge
helpers"), so we can remove the contented atomic ops from the
datapath.

v1 -> v2:
 - complete the qdisc_qstats_atomic_qlen_dec() ->
   qdisc_qstats_cpu_qlen_dec() replacement, fix build issue
 - more descriptive commit message
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/sch_generic.h

@@ -52,10 +52,7 @@ struct qdisc_size_table {
 struct qdisc_skb_head {
 	struct sk_buff	*head;
 	struct sk_buff	*tail;
-	union {
-		u32		qlen;
-		atomic_t	atomic_qlen;
-	};
+	__u32		qlen;
 	spinlock_t	lock;
 };
 
@@ -486,19 +483,27 @@ static inline void qdisc_cb_private_validate(const struct sk_buff *skb, int sz)
 	BUILD_BUG_ON(sizeof(qcb->data) < sz);
 }
 
+static inline int qdisc_qlen_cpu(const struct Qdisc *q)
+{
+	return this_cpu_ptr(q->cpu_qstats)->qlen;
+}
+
 static inline int qdisc_qlen(const struct Qdisc *q)
 {
 	return q->q.qlen;
 }
 
-static inline u32 qdisc_qlen_sum(const struct Qdisc *q)
+static inline int qdisc_qlen_sum(const struct Qdisc *q)
 {
-	u32 qlen = q->qstats.qlen;
+	__u32 qlen = q->qstats.qlen;
+	int i;
 
-	if (qdisc_is_percpu_stats(q))
-		qlen += atomic_read(&q->q.atomic_qlen);
-	else
+	if (qdisc_is_percpu_stats(q)) {
+		for_each_possible_cpu(i)
+			qlen += per_cpu_ptr(q->cpu_qstats, i)->qlen;
+	} else {
 		qlen += q->q.qlen;
+	}
 
 	return qlen;
 }
@@ -889,14 +894,14 @@ static inline void qdisc_qstats_cpu_backlog_inc(struct Qdisc *sch,
 	this_cpu_add(sch->cpu_qstats->backlog, qdisc_pkt_len(skb));
 }
 
-static inline void qdisc_qstats_atomic_qlen_inc(struct Qdisc *sch)
+static inline void qdisc_qstats_cpu_qlen_inc(struct Qdisc *sch)
 {
-	atomic_inc(&sch->q.atomic_qlen);
+	this_cpu_inc(sch->cpu_qstats->qlen);
 }
 
-static inline void qdisc_qstats_atomic_qlen_dec(struct Qdisc *sch)
+static inline void qdisc_qstats_cpu_qlen_dec(struct Qdisc *sch)
 {
-	atomic_dec(&sch->q.atomic_qlen);
+	this_cpu_dec(sch->cpu_qstats->qlen);
 }
 
 static inline void qdisc_qstats_cpu_requeues_inc(struct Qdisc *sch)
@@ -1112,7 +1117,7 @@ static inline void qdisc_update_stats_at_dequeue(struct Qdisc *sch,
 	if (qdisc_is_percpu_stats(sch)) {
 		qdisc_qstats_cpu_backlog_dec(sch, skb);
 		qdisc_bstats_cpu_update(sch, skb);
-		qdisc_qstats_atomic_qlen_dec(sch);
+		qdisc_qstats_cpu_qlen_dec(sch);
 	} else {
 		qdisc_qstats_backlog_dec(sch, skb);
 		qdisc_bstats_update(sch, skb);
@@ -1124,7 +1129,7 @@ static inline void qdisc_update_stats_at_enqueue(struct Qdisc *sch,
 						 unsigned int pkt_len)
 {
 	if (qdisc_is_percpu_stats(sch)) {
-		qdisc_qstats_atomic_qlen_inc(sch);
+		qdisc_qstats_cpu_qlen_inc(sch);
 		this_cpu_add(sch->cpu_qstats->backlog, pkt_len);
 	} else {
 		sch->qstats.backlog += pkt_len;
@@ -1141,7 +1146,7 @@ static inline struct sk_buff *qdisc_dequeue_peeked(struct Qdisc *sch)
 		skb = __skb_dequeue(&sch->gso_skb);
 		if (qdisc_is_percpu_stats(sch)) {
 			qdisc_qstats_cpu_backlog_dec(sch, skb);
-			qdisc_qstats_atomic_qlen_dec(sch);
+			qdisc_qstats_cpu_qlen_dec(sch);
 		} else {
 			qdisc_qstats_backlog_dec(sch, skb);
 			sch->q.qlen--;

net/core/gen_stats.c

@@ -291,6 +291,7 @@ __gnet_stats_copy_queue_cpu(struct gnet_stats_queue *qstats,
 	for_each_possible_cpu(i) {
 		const struct gnet_stats_queue *qcpu = per_cpu_ptr(q, i);
 
+		qstats->qlen = 0;
 		qstats->backlog += qcpu->backlog;
 		qstats->drops += qcpu->drops;
 		qstats->requeues += qcpu->requeues;
@@ -306,6 +307,7 @@ void __gnet_stats_copy_queue(struct gnet_stats_queue *qstats,
 	if (cpu) {
 		__gnet_stats_copy_queue_cpu(qstats, cpu);
 	} else {
+		qstats->qlen = q->qlen;
 		qstats->backlog = q->backlog;
 		qstats->drops = q->drops;
 		qstats->requeues = q->requeues;

net/sched/sch_generic.c

@@ -68,7 +68,7 @@ static inline struct sk_buff *__skb_dequeue_bad_txq(struct Qdisc *q)
 			skb = __skb_dequeue(&q->skb_bad_txq);
 			if (qdisc_is_percpu_stats(q)) {
 				qdisc_qstats_cpu_backlog_dec(q, skb);
-				qdisc_qstats_atomic_qlen_dec(q);
+				qdisc_qstats_cpu_qlen_dec(q);
 			} else {
 				qdisc_qstats_backlog_dec(q, skb);
 				q->q.qlen--;
@@ -108,7 +108,7 @@ static inline void qdisc_enqueue_skb_bad_txq(struct Qdisc *q,
 
 	if (qdisc_is_percpu_stats(q)) {
 		qdisc_qstats_cpu_backlog_inc(q, skb);
-		qdisc_qstats_atomic_qlen_inc(q);
+		qdisc_qstats_cpu_qlen_inc(q);
 	} else {
 		qdisc_qstats_backlog_inc(q, skb);
 		q->q.qlen++;
@@ -136,7 +136,7 @@ static inline void dev_requeue_skb(struct sk_buff *skb, struct Qdisc *q)
 		if (qdisc_is_percpu_stats(q)) {
 			qdisc_qstats_cpu_requeues_inc(q);
 			qdisc_qstats_cpu_backlog_inc(q, skb);
-			qdisc_qstats_atomic_qlen_inc(q);
+			qdisc_qstats_cpu_qlen_inc(q);
 		} else {
 			q->qstats.requeues++;
 			qdisc_qstats_backlog_inc(q, skb);
@@ -236,7 +236,7 @@ static struct sk_buff *dequeue_skb(struct Qdisc *q, bool *validate,
 			skb = __skb_dequeue(&q->gso_skb);
 			if (qdisc_is_percpu_stats(q)) {
 				qdisc_qstats_cpu_backlog_dec(q, skb);
-				qdisc_qstats_atomic_qlen_dec(q);
+				qdisc_qstats_cpu_qlen_dec(q);
 			} else {
 				qdisc_qstats_backlog_dec(q, skb);
 				q->q.qlen--;
@@ -694,6 +694,7 @@ static void pfifo_fast_reset(struct Qdisc *qdisc)
 		struct gnet_stats_queue *q = per_cpu_ptr(qdisc->cpu_qstats, i);
 
 		q->backlog = 0;
+		q->qlen = 0;
 	}
 }