Merge tag 'v5.3.10' into toradex_5.3.ytoradex_5.3.y

This is the 5.3.10 stable release

9 files changed, 44 insertions, 24 deletions

diff --git a/include/net/busy_poll.h b/include/net/busy_poll.h
index 127a5c4e3699..86e028388bad 100644
--- a/include/net/busy_poll.h
+++ b/include/net/busy_poll.h
@@ -122,7 +122,7 @@ static inline void skb_mark_napi_id(struct sk_buff *skb,
 static inline void sk_mark_napi_id(struct sock *sk, const struct sk_buff *skb)
 {
 #ifdef CONFIG_NET_RX_BUSY_POLL
-	sk->sk_napi_id = skb->napi_id;
+	WRITE_ONCE(sk->sk_napi_id, skb->napi_id);
 #endif
 	sk_rx_queue_set(sk, skb);
 }
@@ -132,8 +132,8 @@ static inline void sk_mark_napi_id_once(struct sock *sk,
 					const struct sk_buff *skb)
 {
 #ifdef CONFIG_NET_RX_BUSY_POLL
-	if (!sk->sk_napi_id)
-		sk->sk_napi_id = skb->napi_id;
+	if (!READ_ONCE(sk->sk_napi_id))
+		WRITE_ONCE(sk->sk_napi_id, skb->napi_id);
 #endif
 }
 
diff --git a/include/net/flow_dissector.h b/include/net/flow_dissector.h
index 90bd210be060..5cd12276ae21 100644
--- a/include/net/flow_dissector.h
+++ b/include/net/flow_dissector.h
@@ -4,6 +4,7 @@
 
 #include <linux/types.h>
 #include <linux/in6.h>
+#include <linux/siphash.h>
 #include <uapi/linux/if_ether.h>
 
 /**
@@ -276,7 +277,7 @@ struct flow_keys_basic {
 struct flow_keys {
 	struct flow_dissector_key_control control;
 #define FLOW_KEYS_HASH_START_FIELD basic
-	struct flow_dissector_key_basic basic;
+	struct flow_dissector_key_basic basic __aligned(SIPHASH_ALIGNMENT);
 	struct flow_dissector_key_tags tags;
 	struct flow_dissector_key_vlan vlan;
 	struct flow_dissector_key_vlan cvlan;
diff --git a/include/net/fq.h b/include/net/fq.h
index d126b5d20261..2ad85e683041 100644
--- a/include/net/fq.h
+++ b/include/net/fq.h
@@ -69,7 +69,7 @@ struct fq {
 	struct list_head backlogs;
 	spinlock_t lock;
 	u32 flows_cnt;
-	u32 perturbation;
+	siphash_key_t	perturbation;
 	u32 limit;
 	u32 memory_limit;
 	u32 memory_usage;
diff --git a/include/net/fq_impl.h b/include/net/fq_impl.h
index be40a4b327e3..107c0d700ed6 100644
--- a/include/net/fq_impl.h
+++ b/include/net/fq_impl.h
@@ -108,7 +108,7 @@ begin:
 
 static u32 fq_flow_idx(struct fq *fq, struct sk_buff *skb)
 {
-	u32 hash = skb_get_hash_perturb(skb, fq->perturbation);
+	u32 hash = skb_get_hash_perturb(skb, &fq->perturbation);
 
 	return reciprocal_scale(hash, fq->flows_cnt);
 }
@@ -308,7 +308,7 @@ static int fq_init(struct fq *fq, int flows_cnt)
 	INIT_LIST_HEAD(&fq->backlogs);
 	spin_lock_init(&fq->lock);
 	fq->flows_cnt = max_t(u32, flows_cnt, 1);
-	fq->perturbation = prandom_u32();
+	get_random_bytes(&fq->perturbation, sizeof(fq->perturbation));
 	fq->quantum = 300;
 	fq->limit = 8192;
 	fq->memory_limit = 16 << 20; /* 16 MBytes */
diff --git a/include/net/ip.h b/include/net/ip.h
index 29d89de39822..e6609ab69161 100644
--- a/include/net/ip.h
+++ b/include/net/ip.h
@@ -184,7 +184,7 @@ static inline struct sk_buff *ip_fraglist_next(struct ip_fraglist_iter *iter)
 }
 
 struct ip_frag_state {
-	struct iphdr	*iph;
+	bool		DF;
 	unsigned int	hlen;
 	unsigned int	ll_rs;
 	unsigned int	mtu;
@@ -195,7 +195,7 @@ struct ip_frag_state {
 };
 
 void ip_frag_init(struct sk_buff *skb, unsigned int hlen, unsigned int ll_rs,
-		  unsigned int mtu, struct ip_frag_state *state);
+		  unsigned int mtu, bool DF, struct ip_frag_state *state);
 struct sk_buff *ip_frag_next(struct sk_buff *skb,
 			     struct ip_frag_state *state);
 
diff --git a/include/net/llc_conn.h b/include/net/llc_conn.h
index df528a623548..ea985aa7a6c5 100644
--- a/include/net/llc_conn.h
+++ b/include/net/llc_conn.h
@@ -104,7 +104,7 @@ void llc_sk_reset(struct sock *sk);
 
 /* Access to a connection */
 int llc_conn_state_process(struct sock *sk, struct sk_buff *skb);
-int llc_conn_send_pdu(struct sock *sk, struct sk_buff *skb);
+void llc_conn_send_pdu(struct sock *sk, struct sk_buff *skb);
 void llc_conn_rtn_pdu(struct sock *sk, struct sk_buff *skb);
 void llc_conn_resend_i_pdu_as_cmd(struct sock *sk, u8 nr, u8 first_p_bit);
 void llc_conn_resend_i_pdu_as_rsp(struct sock *sk, u8 nr, u8 first_f_bit);
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index ab40d7afdc54..8f8b37198f9b 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -52,6 +52,9 @@ struct bpf_prog;
 #define NETDEV_HASHENTRIES (1 << NETDEV_HASHBITS)
 
 struct net {
+	/* First cache line can be often dirtied.
+	 * Do not place here read-mostly fields.
+	 */
 	refcount_t		passive;	/* To decide when the network
 						 * namespace should be freed.
 						 */
@@ -60,7 +63,13 @@ struct net {
 						 */
 	spinlock_t		rules_mod_lock;
 
-	u32			hash_mix;
+	unsigned int		dev_unreg_count;
+
+	unsigned int		dev_base_seq;	/* protected by rtnl_mutex */
+	int			ifindex;
+
+	spinlock_t		nsid_lock;
+	atomic_t		fnhe_genid;
 
 	struct list_head	list;		/* list of network namespaces */
 	struct list_head	exit_list;	/* To linked to call pernet exit
@@ -76,11 +85,11 @@ struct net {
 #endif
 	struct user_namespace   *user_ns;	/* Owning user namespace */
 	struct ucounts		*ucounts;
-	spinlock_t		nsid_lock;
 	struct idr		netns_ids;
 
 	struct ns_common	ns;
 
+	struct list_head 	dev_base_head;
 	struct proc_dir_entry 	*proc_net;
 	struct proc_dir_entry 	*proc_net_stat;
 
@@ -93,12 +102,14 @@ struct net {
 
 	struct uevent_sock	*uevent_sock;		/* uevent socket */
 
-	struct list_head 	dev_base_head;
 	struct hlist_head 	*dev_name_head;
 	struct hlist_head	*dev_index_head;
-	unsigned int		dev_base_seq;	/* protected by rtnl_mutex */
-	int			ifindex;
-	unsigned int		dev_unreg_count;
+	/* Note that @hash_mix can be read millions times per second,
+	 * it is critical that it is on a read_mostly cache line.
+	 */
+	u32			hash_mix;
+
+	struct net_device       *loopback_dev;          /* The loopback */
 
 	/* core fib_rules */
 	struct list_head	rules_ops;
@@ -106,7 +117,6 @@ struct net {
 	struct list_head	fib_notifier_ops;  /* Populated by
 						    * register_pernet_subsys()
 						    */
-	struct net_device       *loopback_dev;          /* The loopback */
 	struct netns_core	core;
 	struct netns_mib	mib;
 	struct netns_packet	packet;
@@ -171,7 +181,6 @@ struct net {
 	struct netns_xdp	xdp;
 #endif
 	struct sock		*diag_nlsk;
-	atomic_t		fnhe_genid;
 } __randomize_layout;
 
 #include <linux/seq_file_net.h>
@@ -333,7 +342,7 @@ static inline struct net *read_pnet(const possible_net_t *pnet)
 #define __net_initconst	__initconst
 #endif
 
-int peernet2id_alloc(struct net *net, struct net *peer);
+int peernet2id_alloc(struct net *net, struct net *peer, gfp_t gfp);
 int peernet2id(struct net *net, struct net *peer);
 bool peernet_has_id(struct net *net, struct net *peer);
 struct net *get_net_ns_by_id(struct net *net, int id);
diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h
index 6b6b01234dd9..58b1fbc884a7 100644
--- a/include/net/sch_generic.h
+++ b/include/net/sch_generic.h
@@ -520,6 +520,11 @@ static inline struct Qdisc *qdisc_root(const struct Qdisc *qdisc)
 	return q;
 }
 
+static inline struct Qdisc *qdisc_root_bh(const struct Qdisc *qdisc)
+{
+	return rcu_dereference_bh(qdisc->dev_queue->qdisc);
+}
+
 static inline struct Qdisc *qdisc_root_sleeping(const struct Qdisc *qdisc)
 {
 	return qdisc->dev_queue->qdisc_sleeping;
diff --git a/include/net/sock.h b/include/net/sock.h
index 2c53f1a1d905..b03f96370f8e 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -949,8 +949,8 @@ static inline void sk_incoming_cpu_update(struct sock *sk)
 {
 	int cpu = raw_smp_processor_id();
 
-	if (unlikely(sk->sk_incoming_cpu != cpu))
-		sk->sk_incoming_cpu = cpu;
+	if (unlikely(READ_ONCE(sk->sk_incoming_cpu) != cpu))
+		WRITE_ONCE(sk->sk_incoming_cpu, cpu);
 }
 
 static inline void sock_rps_record_flow_hash(__u32 hash)
@@ -2233,12 +2233,17 @@ struct sk_buff *sk_stream_alloc_skb(struct sock *sk, int size, gfp_t gfp,
  * sk_page_frag - return an appropriate page_frag
  * @sk: socket
  *
- * If socket allocation mode allows current thread to sleep, it means its
- * safe to use the per task page_frag instead of the per socket one.
+ * Use the per task page_frag instead of the per socket one for
+ * optimization when we know that we're in the normal context and owns
+ * everything that's associated with %current.
+ *
+ * gfpflags_allow_blocking() isn't enough here as direct reclaim may nest
+ * inside other socket operations and end up recursing into sk_page_frag()
+ * while it's already in use.
  */
 static inline struct page_frag *sk_page_frag(struct sock *sk)
 {
-	if (gfpflags_allow_blocking(sk->sk_allocation))
+	if (gfpflags_normal_context(sk->sk_allocation))
 		return &current->task_frag;
 
 	return &sk->sk_frag;