net: sched: Merge Qdisc::bstats and Qdisc::cpu_bstats data types

The only factor differentiating per-CPU bstats data type (struct
gnet_stats_basic_cpu) from the packed non-per-CPU one (struct
gnet_stats_basic_packed) was a u64_stats sync point inside the former.
The two data types are now equivalent: earlier commits added a u64_stats
sync point to the latter.

Combine both data types into "struct gnet_stats_basic_sync". This
eliminates redundancy and simplifies the bstats read/write APIs.

Use u64_stats_t for bstats "packets" and "bytes" data types. On 64-bit
architectures, u64_stats sync points do not use sequence counter
protection.

Signed-off-by: Ahmed S. Darwish <a.darwish@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

24 files changed, 114 insertions, 105 deletions

diff --git a/net/core/gen_estimator.c b/net/core/gen_estimator.c
index 64978e77368f..a73ad0bf324c 100644
--- a/net/core/gen_estimator.c
+++ b/net/core/gen_estimator.c
@@ -40,10 +40,10 @@
  */
 
 struct net_rate_estimator {
-	struct gnet_stats_basic_packed	*bstats;
+	struct gnet_stats_basic_sync	*bstats;
 	spinlock_t		*stats_lock;
 	seqcount_t		*running;
-	struct gnet_stats_basic_cpu __percpu *cpu_bstats;
+	struct gnet_stats_basic_sync __percpu *cpu_bstats;
 	u8			ewma_log;
 	u8			intvl_log; /* period : (250ms << intvl_log) */
 
@@ -60,9 +60,9 @@ struct net_rate_estimator {
 };
 
 static void est_fetch_counters(struct net_rate_estimator *e,
-			       struct gnet_stats_basic_packed *b)
+			       struct gnet_stats_basic_sync *b)
 {
-	gnet_stats_basic_packed_init(b);
+	gnet_stats_basic_sync_init(b);
 	if (e->stats_lock)
 		spin_lock(e->stats_lock);
 
@@ -76,14 +76,18 @@ static void est_fetch_counters(struct net_rate_estimator *e,
 static void est_timer(struct timer_list *t)
 {
 	struct net_rate_estimator *est = from_timer(est, t, timer);
-	struct gnet_stats_basic_packed b;
+	struct gnet_stats_basic_sync b;
+	u64 b_bytes, b_packets;
 	u64 rate, brate;
 
 	est_fetch_counters(est, &b);
-	brate = (b.bytes - est->last_bytes) << (10 - est->intvl_log);
+	b_bytes = u64_stats_read(&b.bytes);
+	b_packets = u64_stats_read(&b.packets);
+
+	brate = (b_bytes - est->last_bytes) << (10 - est->intvl_log);
 	brate = (brate >> est->ewma_log) - (est->avbps >> est->ewma_log);
 
-	rate = (b.packets - est->last_packets) << (10 - est->intvl_log);
+	rate = (b_packets - est->last_packets) << (10 - est->intvl_log);
 	rate = (rate >> est->ewma_log) - (est->avpps >> est->ewma_log);
 
 	write_seqcount_begin(&est->seq);
@@ -91,8 +95,8 @@ static void est_timer(struct timer_list *t)
 	est->avpps += rate;
 	write_seqcount_end(&est->seq);
 
-	est->last_bytes = b.bytes;
-	est->last_packets = b.packets;
+	est->last_bytes = b_bytes;
+	est->last_packets = b_packets;
 
 	est->next_jiffies += ((HZ/4) << est->intvl_log);
 
@@ -121,8 +125,8 @@ static void est_timer(struct timer_list *t)
  * Returns 0 on success or a negative error code.
  *
  */
-int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
-		      struct gnet_stats_basic_cpu __percpu *cpu_bstats,
+int gen_new_estimator(struct gnet_stats_basic_sync *bstats,
+		      struct gnet_stats_basic_sync __percpu *cpu_bstats,
 		      struct net_rate_estimator __rcu **rate_est,
 		      spinlock_t *lock,
 		      seqcount_t *running,
@@ -130,7 +134,7 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
 {
 	struct gnet_estimator *parm = nla_data(opt);
 	struct net_rate_estimator *old, *est;
-	struct gnet_stats_basic_packed b;
+	struct gnet_stats_basic_sync b;
 	int intvl_log;
 
 	if (nla_len(opt) < sizeof(*parm))
@@ -164,8 +168,8 @@ int gen_new_estimator(struct gnet_stats_basic_packed *bstats,
 	est_fetch_counters(est, &b);
 	if (lock)
 		local_bh_enable();
-	est->last_bytes = b.bytes;
-	est->last_packets = b.packets;
+	est->last_bytes = u64_stats_read(&b.bytes);
+	est->last_packets = u64_stats_read(&b.packets);
 
 	if (lock)
 		spin_lock_bh(lock);
@@ -222,8 +226,8 @@ EXPORT_SYMBOL(gen_kill_estimator);
  *
  * Returns 0 on success or a negative error code.
  */
-int gen_replace_estimator(struct gnet_stats_basic_packed *bstats,
-			  struct gnet_stats_basic_cpu __percpu *cpu_bstats,
+int gen_replace_estimator(struct gnet_stats_basic_sync *bstats,
+			  struct gnet_stats_basic_sync __percpu *cpu_bstats,
 			  struct net_rate_estimator __rcu **rate_est,
 			  spinlock_t *lock,
 			  seqcount_t *running, struct nlattr *opt)
diff --git a/net/core/gen_stats.c b/net/core/gen_stats.c
index 69576972a25f..5f57f761def6 100644
--- a/net/core/gen_stats.c
+++ b/net/core/gen_stats.c
@@ -115,29 +115,29 @@ gnet_stats_start_copy(struct sk_buff *skb, int type, spinlock_t *lock,
 EXPORT_SYMBOL(gnet_stats_start_copy);
 
 /* Must not be inlined, due to u64_stats seqcount_t lockdep key */
-void gnet_stats_basic_packed_init(struct gnet_stats_basic_packed *b)
+void gnet_stats_basic_sync_init(struct gnet_stats_basic_sync *b)
 {
-	b->bytes = 0;
-	b->packets = 0;
+	u64_stats_set(&b->bytes, 0);
+	u64_stats_set(&b->packets, 0);
 	u64_stats_init(&b->syncp);
 }
-EXPORT_SYMBOL(gnet_stats_basic_packed_init);
+EXPORT_SYMBOL(gnet_stats_basic_sync_init);
 
-static void gnet_stats_add_basic_cpu(struct gnet_stats_basic_packed *bstats,
-				     struct gnet_stats_basic_cpu __percpu *cpu)
+static void gnet_stats_add_basic_cpu(struct gnet_stats_basic_sync *bstats,
+				     struct gnet_stats_basic_sync __percpu *cpu)
 {
 	u64 t_bytes = 0, t_packets = 0;
 	int i;
 
 	for_each_possible_cpu(i) {
-		struct gnet_stats_basic_cpu *bcpu = per_cpu_ptr(cpu, i);
+		struct gnet_stats_basic_sync *bcpu = per_cpu_ptr(cpu, i);
 		unsigned int start;
 		u64 bytes, packets;
 
 		do {
 			start = u64_stats_fetch_begin_irq(&bcpu->syncp);
-			bytes = bcpu->bstats.bytes;
-			packets = bcpu->bstats.packets;
+			bytes = u64_stats_read(&bcpu->bytes);
+			packets = u64_stats_read(&bcpu->packets);
 		} while (u64_stats_fetch_retry_irq(&bcpu->syncp, start));
 
 		t_bytes += bytes;
@@ -147,9 +147,9 @@ static void gnet_stats_add_basic_cpu(struct gnet_stats_basic_packed *bstats,
 }
 
 void gnet_stats_add_basic(const seqcount_t *running,
-			  struct gnet_stats_basic_packed *bstats,
-			  struct gnet_stats_basic_cpu __percpu *cpu,
-			  struct gnet_stats_basic_packed *b)
+			  struct gnet_stats_basic_sync *bstats,
+			  struct gnet_stats_basic_sync __percpu *cpu,
+			  struct gnet_stats_basic_sync *b)
 {
 	unsigned int seq;
 	u64 bytes = 0;
@@ -162,8 +162,8 @@ void gnet_stats_add_basic(const seqcount_t *running,
 	do {
 		if (running)
 			seq = read_seqcount_begin(running);
-		bytes = b->bytes;
-		packets = b->packets;
+		bytes = u64_stats_read(&b->bytes);
+		packets = u64_stats_read(&b->packets);
 	} while (running && read_seqcount_retry(running, seq));
 
 	_bstats_update(bstats, bytes, packets);
@@ -173,18 +173,22 @@ EXPORT_SYMBOL(gnet_stats_add_basic);
 static int
 ___gnet_stats_copy_basic(const seqcount_t *running,
 			 struct gnet_dump *d,
-			 struct gnet_stats_basic_cpu __percpu *cpu,
-			 struct gnet_stats_basic_packed *b,
+			 struct gnet_stats_basic_sync __percpu *cpu,
+			 struct gnet_stats_basic_sync *b,
 			 int type)
 {
-	struct gnet_stats_basic_packed bstats;
+	struct gnet_stats_basic_sync bstats;
+	u64 bstats_bytes, bstats_packets;
 
-	gnet_stats_basic_packed_init(&bstats);
+	gnet_stats_basic_sync_init(&bstats);
 	gnet_stats_add_basic(running, &bstats, cpu, b);
 
+	bstats_bytes = u64_stats_read(&bstats.bytes);
+	bstats_packets = u64_stats_read(&bstats.packets);
+
 	if (d->compat_tc_stats && type == TCA_STATS_BASIC) {
-		d->tc_stats.bytes = bstats.bytes;
-		d->tc_stats.packets = bstats.packets;
+		d->tc_stats.bytes = bstats_bytes;
+		d->tc_stats.packets = bstats_packets;
 	}
 
 	if (d->tail) {
@@ -192,14 +196,14 @@ ___gnet_stats_copy_basic(const seqcount_t *running,
 		int res;
 
 		memset(&sb, 0, sizeof(sb));
-		sb.bytes = bstats.bytes;
-		sb.packets = bstats.packets;
+		sb.bytes = bstats_bytes;
+		sb.packets = bstats_packets;
 		res = gnet_stats_copy(d, type, &sb, sizeof(sb), TCA_STATS_PAD);
-		if (res < 0 || sb.packets == bstats.packets)
+		if (res < 0 || sb.packets == bstats_packets)
 			return res;
 		/* emit 64bit stats only if needed */
-		return gnet_stats_copy(d, TCA_STATS_PKT64, &bstats.packets,
-				       sizeof(bstats.packets), TCA_STATS_PAD);
+		return gnet_stats_copy(d, TCA_STATS_PKT64, &bstats_packets,
+				       sizeof(bstats_packets), TCA_STATS_PAD);
 	}
 	return 0;
 }
@@ -220,8 +224,8 @@ ___gnet_stats_copy_basic(const seqcount_t *running,
 int
 gnet_stats_copy_basic(const seqcount_t *running,
 		      struct gnet_dump *d,
-		      struct gnet_stats_basic_cpu __percpu *cpu,
-		      struct gnet_stats_basic_packed *b)
+		      struct gnet_stats_basic_sync __percpu *cpu,
+		      struct gnet_stats_basic_sync *b)
 {
 	return ___gnet_stats_copy_basic(running, d, cpu, b,
 					TCA_STATS_BASIC);
@@ -244,8 +248,8 @@ EXPORT_SYMBOL(gnet_stats_copy_basic);
 int
 gnet_stats_copy_basic_hw(const seqcount_t *running,
 			 struct gnet_dump *d,
-			 struct gnet_stats_basic_cpu __percpu *cpu,
-			 struct gnet_stats_basic_packed *b)
+			 struct gnet_stats_basic_sync __percpu *cpu,
+			 struct gnet_stats_basic_sync *b)
 {
 	return ___gnet_stats_copy_basic(running, d, cpu, b,
 					TCA_STATS_BASIC_HW);
diff --git a/net/netfilter/xt_RATEEST.c b/net/netfilter/xt_RATEEST.c
index d5200725ca62..8aec1b529364 100644
--- a/net/netfilter/xt_RATEEST.c
+++ b/net/netfilter/xt_RATEEST.c
@@ -94,11 +94,11 @@ static unsigned int
 xt_rateest_tg(struct sk_buff *skb, const struct xt_action_param *par)
 {
 	const struct xt_rateest_target_info *info = par->targinfo;
-	struct gnet_stats_basic_packed *stats = &info->est->bstats;
+	struct gnet_stats_basic_sync *stats = &info->est->bstats;
 
 	spin_lock_bh(&info->est->lock);
-	stats->bytes += skb->len;
-	stats->packets++;
+	u64_stats_add(&stats->bytes, skb->len);
+	u64_stats_inc(&stats->packets);
 	spin_unlock_bh(&info->est->lock);
 
 	return XT_CONTINUE;
@@ -143,7 +143,7 @@ static int xt_rateest_tg_checkentry(const struct xt_tgchk_param *par)
 	if (!est)
 		goto err1;
 
-	gnet_stats_basic_packed_init(&est->bstats);
+	gnet_stats_basic_sync_init(&est->bstats);
 	strlcpy(est->name, info->name, sizeof(est->name));
 	spin_lock_init(&est->lock);
 	est->refcnt		= 1;
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index 0302dad42df1..585829ffa0c4 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -480,18 +480,18 @@ int tcf_idr_create(struct tc_action_net *tn, u32 index, struct nlattr *est,
 		atomic_set(&p->tcfa_bindcnt, 1);
 
 	if (cpustats) {
-		p->cpu_bstats = netdev_alloc_pcpu_stats(struct gnet_stats_basic_cpu);
+		p->cpu_bstats = netdev_alloc_pcpu_stats(struct gnet_stats_basic_sync);
 		if (!p->cpu_bstats)
 			goto err1;
-		p->cpu_bstats_hw = netdev_alloc_pcpu_stats(struct gnet_stats_basic_cpu);
+		p->cpu_bstats_hw = netdev_alloc_pcpu_stats(struct gnet_stats_basic_sync);
 		if (!p->cpu_bstats_hw)
 			goto err2;
 		p->cpu_qstats = alloc_percpu(struct gnet_stats_queue);
 		if (!p->cpu_qstats)
 			goto err3;
 	}
-	gnet_stats_basic_packed_init(&p->tcfa_bstats);
-	gnet_stats_basic_packed_init(&p->tcfa_bstats_hw);
+	gnet_stats_basic_sync_init(&p->tcfa_bstats);
+	gnet_stats_basic_sync_init(&p->tcfa_bstats_hw);
 	spin_lock_init(&p->tcfa_lock);
 	p->tcfa_index = index;
 	p->tcfa_tm.install = jiffies;
@@ -1128,13 +1128,13 @@ void tcf_action_update_stats(struct tc_action *a, u64 bytes, u64 packets,
 			     u64 drops, bool hw)
 {
 	if (a->cpu_bstats) {
-		_bstats_cpu_update(this_cpu_ptr(a->cpu_bstats), bytes, packets);
+		_bstats_update(this_cpu_ptr(a->cpu_bstats), bytes, packets);
 
 		this_cpu_ptr(a->cpu_qstats)->drops += drops;
 
 		if (hw)
-			_bstats_cpu_update(this_cpu_ptr(a->cpu_bstats_hw),
-					   bytes, packets);
+			_bstats_update(this_cpu_ptr(a->cpu_bstats_hw),
+				       bytes, packets);
 		return;
 	}
 
diff --git a/net/sched/act_bpf.c b/net/sched/act_bpf.c
index 5c36013339e1..f2bf896331a5 100644
--- a/net/sched/act_bpf.c
+++ b/net/sched/act_bpf.c
@@ -41,7 +41,7 @@ static int tcf_bpf_act(struct sk_buff *skb, const struct tc_action *act,
 	int action, filter_res;
 
 	tcf_lastuse_update(&prog->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(prog->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(prog->common.cpu_bstats), skb);
 
 	filter = rcu_dereference(prog->filter);
 	if (at_ingress) {
diff --git a/net/sched/act_ife.c b/net/sched/act_ife.c
index 7064a365a1a9..b757f90a2d58 100644
--- a/net/sched/act_ife.c
+++ b/net/sched/act_ife.c
@@ -718,7 +718,7 @@ static int tcf_ife_decode(struct sk_buff *skb, const struct tc_action *a,
 	u8 *tlv_data;
 	u16 metalen;
 
-	bstats_cpu_update(this_cpu_ptr(ife->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(ife->common.cpu_bstats), skb);
 	tcf_lastuse_update(&ife->tcf_tm);
 
 	if (skb_at_tc_ingress(skb))
@@ -806,7 +806,7 @@ static int tcf_ife_encode(struct sk_buff *skb, const struct tc_action *a,
 			exceed_mtu = true;
 	}
 
-	bstats_cpu_update(this_cpu_ptr(ife->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(ife->common.cpu_bstats), skb);
 	tcf_lastuse_update(&ife->tcf_tm);
 
 	if (!metalen) {		/* no metadata to send */
diff --git a/net/sched/act_mpls.c b/net/sched/act_mpls.c
index e4529b428cf4..8faa4c58305e 100644
--- a/net/sched/act_mpls.c
+++ b/net/sched/act_mpls.c
@@ -59,7 +59,7 @@ static int tcf_mpls_act(struct sk_buff *skb, const struct tc_action *a,
 	int ret, mac_len;
 
 	tcf_lastuse_update(&m->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(m->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(m->common.cpu_bstats), skb);
 
 	/* Ensure 'data' points at mac_header prior calling mpls manipulating
 	 * functions.
diff --git a/net/sched/act_police.c b/net/sched/act_police.c
index 832157a840fc..c9383805222d 100644
--- a/net/sched/act_police.c
+++ b/net/sched/act_police.c
@@ -248,7 +248,7 @@ static int tcf_police_act(struct sk_buff *skb, const struct tc_action *a,
 	int ret;
 
 	tcf_lastuse_update(&police->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(police->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(police->common.cpu_bstats), skb);
 
 	ret = READ_ONCE(police->tcf_action);
 	p = rcu_dereference_bh(police->params);
diff --git a/net/sched/act_sample.c b/net/sched/act_sample.c
index 230501eb9e06..ce859b0e0deb 100644
--- a/net/sched/act_sample.c
+++ b/net/sched/act_sample.c
@@ -163,7 +163,7 @@ static int tcf_sample_act(struct sk_buff *skb, const struct tc_action *a,
 	int retval;
 
 	tcf_lastuse_update(&s->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(s->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(s->common.cpu_bstats), skb);
 	retval = READ_ONCE(s->tcf_action);
 
 	psample_group = rcu_dereference_bh(s->psample_group);
diff --git a/net/sched/act_simple.c b/net/sched/act_simple.c
index cbbe1861d3a2..e617ab4505ca 100644
--- a/net/sched/act_simple.c
+++ b/net/sched/act_simple.c
@@ -36,7 +36,8 @@ static int tcf_simp_act(struct sk_buff *skb, const struct tc_action *a,
 	 * then it would look like "hello_3" (without quotes)
 	 */
 	pr_info("simple: %s_%llu\n",
-	       (char *)d->tcfd_defdata, d->tcf_bstats.packets);
+		(char *)d->tcfd_defdata,
+		u64_stats_read(&d->tcf_bstats.packets));
 	spin_unlock(&d->tcf_lock);
 	return d->tcf_action;
 }
diff --git a/net/sched/act_skbedit.c b/net/sched/act_skbedit.c
index 605418538347..d30ecbfc8f84 100644
--- a/net/sched/act_skbedit.c
+++ b/net/sched/act_skbedit.c
@@ -31,7 +31,7 @@ static int tcf_skbedit_act(struct sk_buff *skb, const struct tc_action *a,
 	int action;
 
 	tcf_lastuse_update(&d->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(d->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(d->common.cpu_bstats), skb);
 
 	params = rcu_dereference_bh(d->params);
 	action = READ_ONCE(d->tcf_action);
diff --git a/net/sched/act_skbmod.c b/net/sched/act_skbmod.c
index ecb9ee666095..9b6b52c5e24e 100644
--- a/net/sched/act_skbmod.c
+++ b/net/sched/act_skbmod.c
@@ -31,7 +31,7 @@ static int tcf_skbmod_act(struct sk_buff *skb, const struct tc_action *a,
 	u64 flags;
 
 	tcf_lastuse_update(&d->tcf_tm);
-	bstats_cpu_update(this_cpu_ptr(d->common.cpu_bstats), skb);
+	bstats_update(this_cpu_ptr(d->common.cpu_bstats), skb);
 
 	action = READ_ONCE(d->tcf_action);
 	if (unlikely(action == TC_ACT_SHOT))
diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c
index 91820f67275c..70f006cbf212 100644
--- a/net/sched/sch_api.c
+++ b/net/sched/sch_api.c
@@ -885,7 +885,7 @@ static void qdisc_offload_graft_root(struct net_device *dev,
 static int tc_fill_qdisc(struct sk_buff *skb, struct Qdisc *q, u32 clid,
 			 u32 portid, u32 seq, u16 flags, int event)
 {
-	struct gnet_stats_basic_cpu __percpu *cpu_bstats = NULL;
+	struct gnet_stats_basic_sync __percpu *cpu_bstats = NULL;
 	struct gnet_stats_queue __percpu *cpu_qstats = NULL;
 	struct tcmsg *tcm;
 	struct nlmsghdr  *nlh;
diff --git a/net/sched/sch_atm.c b/net/sched/sch_atm.c
index c8e1771383f9..fbfe4ce9497b 100644
--- a/net/sched/sch_atm.c
+++ b/net/sched/sch_atm.c
@@ -52,7 +52,7 @@ struct atm_flow_data {
 	struct atm_qdisc_data	*parent;	/* parent qdisc */
 	struct socket		*sock;		/* for closing */
 	int			ref;		/* reference count */
-	struct gnet_stats_basic_packed	bstats;
+	struct gnet_stats_basic_sync	bstats;
 	struct gnet_stats_queue	qstats;
 	struct list_head	list;
 	struct atm_flow_data	*excess;	/* flow for excess traffic;
@@ -548,7 +548,7 @@ static int atm_tc_init(struct Qdisc *sch, struct nlattr *opt,
 	pr_debug("atm_tc_init(sch %p,[qdisc %p],opt %p)\n", sch, p, opt);
 	INIT_LIST_HEAD(&p->flows);
 	INIT_LIST_HEAD(&p->link.list);
-	gnet_stats_basic_packed_init(&p->link.bstats);
+	gnet_stats_basic_sync_init(&p->link.bstats);
 	list_add(&p->link.list, &p->flows);
 	p->link.q = qdisc_create_dflt(sch->dev_queue,
 				      &pfifo_qdisc_ops, sch->handle, extack);
diff --git a/net/sched/sch_cbq.c b/net/sched/sch_cbq.c
index ef9e87175d35..f0b1282fae11 100644
--- a/net/sched/sch_cbq.c
+++ b/net/sched/sch_cbq.c
@@ -116,7 +116,7 @@ struct cbq_class {
 	long			avgidle;
 	long			deficit;	/* Saved deficit for WRR */
 	psched_time_t		penalized;
-	struct gnet_stats_basic_packed bstats;
+	struct gnet_stats_basic_sync bstats;
 	struct gnet_stats_queue qstats;
 	struct net_rate_estimator __rcu *rate_est;
 	struct tc_cbq_xstats	xstats;
@@ -1610,7 +1610,7 @@ cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **t
 	if (cl == NULL)
 		goto failure;
 
-	gnet_stats_basic_packed_init(&cl->bstats);
+	gnet_stats_basic_sync_init(&cl->bstats);
 	err = tcf_block_get(&cl->block, &cl->filter_list, sch, extack);
 	if (err) {
 		kfree(cl);
diff --git a/net/sched/sch_drr.c b/net/sched/sch_drr.c
index 319906e19a6b..7243617a3595 100644
--- a/net/sched/sch_drr.c
+++ b/net/sched/sch_drr.c
@@ -19,7 +19,7 @@ struct drr_class {
 	struct Qdisc_class_common	common;
 	unsigned int			filter_cnt;
 
-	struct gnet_stats_basic_packed		bstats;
+	struct gnet_stats_basic_sync		bstats;
 	struct gnet_stats_queue		qstats;
 	struct net_rate_estimator __rcu *rate_est;
 	struct list_head		alist;
@@ -106,7 +106,7 @@ static int drr_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
 	if (cl == NULL)
 		return -ENOBUFS;
 
-	gnet_stats_basic_packed_init(&cl->bstats);
+	gnet_stats_basic_sync_init(&cl->bstats);
 	cl->common.classid = classid;
 	cl->quantum	   = quantum;
 	cl->qdisc	   = qdisc_create_dflt(sch->dev_queue,
diff --git a/net/sched/sch_ets.c b/net/sched/sch_ets.c
index 83693107371f..af56d155e7fc 100644
--- a/net/sched/sch_ets.c
+++ b/net/sched/sch_ets.c
@@ -41,7 +41,7 @@ struct ets_class {
 	struct Qdisc *qdisc;
 	u32 quantum;
 	u32 deficit;
-	struct gnet_stats_basic_packed bstats;
+	struct gnet_stats_basic_sync bstats;
 	struct gnet_stats_queue qstats;
 };
 
@@ -689,7 +689,7 @@ static int ets_qdisc_change(struct Qdisc *sch, struct nlattr *opt,
 		q->classes[i].qdisc = NULL;
 		q->classes[i].quantum = 0;
 		q->classes[i].deficit = 0;
-		gnet_stats_basic_packed_init(&q->classes[i].bstats);
+		gnet_stats_basic_sync_init(&q->classes[i].bstats);
 		memset(&q->classes[i].qstats, 0, sizeof(q->classes[i].qstats));
 	}
 	return 0;
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index ef27ff3ddee4..989186e7f1a0 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -892,12 +892,12 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue,
 	__skb_queue_head_init(&sch->gso_skb);
 	__skb_queue_head_init(&sch->skb_bad_txq);
 	qdisc_skb_head_init(&sch->q);
-	gnet_stats_basic_packed_init(&sch->bstats);
+	gnet_stats_basic_sync_init(&sch->bstats);
 	spin_lock_init(&sch->q.lock);
 
 	if (ops->static_flags & TCQ_F_CPUSTATS) {
 		sch->cpu_bstats =
-			netdev_alloc_pcpu_stats(struct gnet_stats_basic_cpu);
+			netdev_alloc_pcpu_stats(struct gnet_stats_basic_sync);
 		if (!sch->cpu_bstats)
 			goto errout1;
 
diff --git a/net/sched/sch_gred.c b/net/sched/sch_gred.c
index 02b03d6d24ea..72de08ef8335 100644
--- a/net/sched/sch_gred.c
+++ b/net/sched/sch_gred.c
@@ -366,7 +366,7 @@ static int gred_offload_dump_stats(struct Qdisc *sch)
 	hw_stats->parent = sch->parent;
 
 	for (i = 0; i < MAX_DPs; i++) {
-		gnet_stats_basic_packed_init(&hw_stats->stats.bstats[i]);
+		gnet_stats_basic_sync_init(&hw_stats->stats.bstats[i]);
 		if (table->tab[i])
 			hw_stats->stats.xstats[i] = &table->tab[i]->stats;
 	}
@@ -378,12 +378,12 @@ static int gred_offload_dump_stats(struct Qdisc *sch)
 	for (i = 0; i < MAX_DPs; i++) {
 		if (!table->tab[i])
 			continue;
-		table->tab[i]->packetsin += hw_stats->stats.bstats[i].packets;
-		table->tab[i]->bytesin += hw_stats->stats.bstats[i].bytes;
+		table->tab[i]->packetsin += u64_stats_read(&hw_stats->stats.bstats[i].packets);
+		table->tab[i]->bytesin += u64_stats_read(&hw_stats->stats.bstats[i].bytes);
 		table->tab[i]->backlog += hw_stats->stats.qstats[i].backlog;
 
-		bytes += hw_stats->stats.bstats[i].bytes;
-		packets += hw_stats->stats.bstats[i].packets;
+		bytes += u64_stats_read(&hw_stats->stats.bstats[i].bytes);
+		packets += u64_stats_read(&hw_stats->stats.bstats[i].packets);
 		sch->qstats.qlen += hw_stats->stats.qstats[i].qlen;
 		sch->qstats.backlog += hw_stats->stats.qstats[i].backlog;
 		sch->qstats.drops += hw_stats->stats.qstats[i].drops;
diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c
index ff6ff54806fc..181c2905ff98 100644
--- a/net/sched/sch_hfsc.c
+++ b/net/sched/sch_hfsc.c
@@ -111,7 +111,7 @@ enum hfsc_class_flags {
 struct hfsc_class {
 	struct Qdisc_class_common cl_common;
 
-	struct gnet_stats_basic_packed bstats;
+	struct gnet_stats_basic_sync bstats;
 	struct gnet_stats_queue qstats;
 	struct net_rate_estimator __rcu *rate_est;
 	struct tcf_proto __rcu *filter_list; /* filter list */
@@ -1406,7 +1406,7 @@ hfsc_init_qdisc(struct Qdisc *sch, struct nlattr *opt,
 	if (err)
 		return err;
 
-	gnet_stats_basic_packed_init(&q->root.bstats);
+	gnet_stats_basic_sync_init(&q->root.bstats);
 	q->root.cl_common.classid = sch->handle;
 	q->root.sched   = q;
 	q->root.qdisc = qdisc_create_dflt(sch->dev_queue, &pfifo_qdisc_ops,
diff --git a/net/sched/sch_htb.c b/net/sched/sch_htb.c
index 324ecfdf842a..adceb9e210f6 100644
--- a/net/sched/sch_htb.c
+++ b/net/sched/sch_htb.c
@@ -113,8 +113,8 @@ struct htb_class {
 	/*
 	 * Written often fields
 	 */
-	struct gnet_stats_basic_packed bstats;
-	struct gnet_stats_basic_packed bstats_bias;
+	struct gnet_stats_basic_sync bstats;
+	struct gnet_stats_basic_sync bstats_bias;
 	struct tc_htb_xstats	xstats;	/* our special stats */
 
 	/* token bucket parameters */
@@ -1312,7 +1312,7 @@ static void htb_offload_aggregate_stats(struct htb_sched *q,
 	struct htb_class *c;
 	unsigned int i;
 
-	gnet_stats_basic_packed_init(&cl->bstats);
+	gnet_stats_basic_sync_init(&cl->bstats);
 
 	for (i = 0; i < q->clhash.hashsize; i++) {
 		hlist_for_each_entry(c, &q->clhash.hash[i], common.hnode) {
@@ -1324,11 +1324,11 @@ static void htb_offload_aggregate_stats(struct htb_sched *q,
 			if (p != cl)
 				continue;
 
-			bytes += c->bstats_bias.bytes;
-			packets += c->bstats_bias.packets;
+			bytes += u64_stats_read(&c->bstats_bias.bytes);
+			packets += u64_stats_read(&c->bstats_bias.packets);
 			if (c->level == 0) {
-				bytes += c->leaf.q->bstats.bytes;
-				packets += c->leaf.q->bstats.packets;
+				bytes += u64_stats_read(&c->leaf.q->bstats.bytes);
+				packets += u64_stats_read(&c->leaf.q->bstats.packets);
 			}
 		}
 	}
@@ -1359,10 +1359,10 @@ htb_dump_class_stats(struct Qdisc *sch, unsigned long arg, struct gnet_dump *d)
 			if (cl->leaf.q)
 				cl->bstats = cl->leaf.q->bstats;
 			else
-				gnet_stats_basic_packed_init(&cl->bstats);
+				gnet_stats_basic_sync_init(&cl->bstats);
 			_bstats_update(&cl->bstats,
-				       cl->bstats_bias.bytes,
-				       cl->bstats_bias.packets);
+				       u64_stats_read(&cl->bstats_bias.bytes),
+				       u64_stats_read(&cl->bstats_bias.packets));
 		} else {
 			htb_offload_aggregate_stats(q, cl);
 		}
@@ -1582,8 +1582,8 @@ static int htb_destroy_class_offload(struct Qdisc *sch, struct htb_class *cl,
 
 	if (cl->parent) {
 		_bstats_update(&cl->parent->bstats_bias,
-			       q->bstats.bytes,
-			       q->bstats.packets);
+			       u64_stats_read(&q->bstats.bytes),
+			       u64_stats_read(&q->bstats.packets));
 	}
 
 	offload_opt = (struct tc_htb_qopt_offload) {
@@ -1853,8 +1853,8 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
 		if (!cl)
 			goto failure;
 
-		gnet_stats_basic_packed_init(&cl->bstats);
-		gnet_stats_basic_packed_init(&cl->bstats_bias);
+		gnet_stats_basic_sync_init(&cl->bstats);
+		gnet_stats_basic_sync_init(&cl->bstats_bias);
 
 		err = tcf_block_get(&cl->block, &cl->filter_list, sch, extack);
 		if (err) {
@@ -1930,8 +1930,8 @@ static int htb_change_class(struct Qdisc *sch, u32 classid,
 				goto err_kill_estimator;
 			}
 			_bstats_update(&parent->bstats_bias,
-				       old_q->bstats.bytes,
-				       old_q->bstats.packets);
+				       u64_stats_read(&old_q->bstats.bytes),
+				       u64_stats_read(&old_q->bstats.packets));
 			qdisc_put(old_q);
 		}
 		new_q = qdisc_create_dflt(dev_queue, &pfifo_qdisc_ops,
diff --git a/net/sched/sch_mq.c b/net/sched/sch_mq.c
index 704e14a58f09..cedd0b3ef9cf 100644
--- a/net/sched/sch_mq.c
+++ b/net/sched/sch_mq.c
@@ -132,7 +132,7 @@ static int mq_dump(struct Qdisc *sch, struct sk_buff *skb)
 	unsigned int ntx;
 
 	sch->q.qlen = 0;
-	gnet_stats_basic_packed_init(&sch->bstats);
+	gnet_stats_basic_sync_init(&sch->bstats);
 	memset(&sch->qstats, 0, sizeof(sch->qstats));
 
 	/* MQ supports lockless qdiscs. However, statistics accounting needs
diff --git a/net/sched/sch_mqprio.c b/net/sched/sch_mqprio.c
index fe6b4a178fc9..3f7f756f92ca 100644
--- a/net/sched/sch_mqprio.c
+++ b/net/sched/sch_mqprio.c
@@ -390,7 +390,7 @@ static int mqprio_dump(struct Qdisc *sch, struct sk_buff *skb)
 	unsigned int ntx, tc;
 
 	sch->q.qlen = 0;
-	gnet_stats_basic_packed_init(&sch->bstats);
+	gnet_stats_basic_sync_init(&sch->bstats);
 	memset(&sch->qstats, 0, sizeof(sch->qstats));
 
 	/* MQ supports lockless qdiscs. However, statistics accounting needs
@@ -500,11 +500,11 @@ static int mqprio_dump_class_stats(struct Qdisc *sch, unsigned long cl,
 		int i;
 		__u32 qlen;
 		struct gnet_stats_queue qstats = {0};
-		struct gnet_stats_basic_packed bstats;
+		struct gnet_stats_basic_sync bstats;
 		struct net_device *dev = qdisc_dev(sch);
 		struct netdev_tc_txq tc = dev->tc_to_txq[cl & TC_BITMASK];
 
-		gnet_stats_basic_packed_init(&bstats);
+		gnet_stats_basic_sync_init(&bstats);
 		/* Drop lock here it will be reclaimed before touching
 		 * statistics this is required because the d->lock we
 		 * hold here is the look on dev_queue->qdisc_sleeping
diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c
index bea68c91027a..a35200f591a2 100644
--- a/net/sched/sch_qfq.c
+++ b/net/sched/sch_qfq.c
@@ -131,7 +131,7 @@ struct qfq_class {
 
 	unsigned int filter_cnt;
 
-	struct gnet_stats_basic_packed bstats;
+	struct gnet_stats_basic_sync bstats;
 	struct gnet_stats_queue qstats;
 	struct net_rate_estimator __rcu *rate_est;
 	struct Qdisc *qdisc;
@@ -465,7 +465,7 @@ static int qfq_change_class(struct Qdisc *sch, u32 classid, u32 parentid,
 	if (cl == NULL)
 		return -ENOBUFS;
 
-	gnet_stats_basic_packed_init(&cl->bstats);
+	gnet_stats_basic_sync_init(&cl->bstats);
 	cl->common.classid = classid;
 	cl->deficit = lmax;