net: Convert GRO SKB handling to list_head.

Manage pending per-NAPI GRO packets via list_head. Return an SKB pointer from the GRO receive handlers. When GRO receive handlers return non-NULL, it means that this SKB needs to be completed at this time and removed from the NAPI queue. Several operations are greatly simplified by this transformation, especially timing out the oldest SKB in the list when gro_count exceeds MAX_GRO_SKBS, and napi_gro_flush() which walks the queue in reverse order. Signed-off-by: David S. Miller <[email protected]>
openbmc · Jun 26, 2018 · d4546c2 · d4546c2
1 parent 9ff3b40
commit d4546c2
Show file tree

Hide file tree

Showing 24 changed files with 133 additions and 141 deletions.
diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
@@ -418,11 +418,12 @@ static int geneve_hlen(struct genevehdr *gh)
 	return sizeof(*gh) + gh->opt_len * 4;
 }
 
-static struct sk_buff **geneve_gro_receive(struct sock *sk,
-					   struct sk_buff **head,
-					   struct sk_buff *skb)
+static struct sk_buff *geneve_gro_receive(struct sock *sk,
+					  struct list_head *head,
+					  struct sk_buff *skb)
 {
-	struct sk_buff *p, **pp = NULL;
+	struct sk_buff *pp = NULL;
+	struct sk_buff *p;
 	struct genevehdr *gh, *gh2;
 	unsigned int hlen, gh_len, off_gnv;
 	const struct packet_offload *ptype;
@@ -449,7 +450,7 @@ static struct sk_buff **geneve_gro_receive(struct sock *sk,
 			goto out;
 	}
 
-	for (p = *head; p; p = p->next) {
+	list_for_each_entry(p, head, list) {
 		if (!NAPI_GRO_CB(p)->same_flow)
 			continue;
 

diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
@@ -568,11 +568,12 @@ static struct vxlanhdr *vxlan_gro_remcsum(struct sk_buff *skb,
 	return vh;
 }
 
-static struct sk_buff **vxlan_gro_receive(struct sock *sk,
-					  struct sk_buff **head,
-					  struct sk_buff *skb)
+static struct sk_buff *vxlan_gro_receive(struct sock *sk,
+					 struct list_head *head,
+					 struct sk_buff *skb)
 {
-	struct sk_buff *p, **pp = NULL;
+	struct sk_buff *pp = NULL;
+	struct sk_buff *p;
 	struct vxlanhdr *vh, *vh2;
 	unsigned int hlen, off_vx;
 	int flush = 1;
@@ -607,7 +608,7 @@ static struct sk_buff **vxlan_gro_receive(struct sock *sk,
 
 	skb_gro_pull(skb, sizeof(struct vxlanhdr)); /* pull vxlan header */
 
-	for (p = *head; p; p = p->next) {
+	list_for_each_entry(p, head, list) {
 		if (!NAPI_GRO_CB(p)->same_flow)
 			continue;
 

diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h
@@ -59,8 +59,7 @@ struct net_device *devm_alloc_etherdev_mqs(struct device *dev, int sizeof_priv,
 					   unsigned int rxqs);
 #define devm_alloc_etherdev(dev, sizeof_priv) devm_alloc_etherdev_mqs(dev, sizeof_priv, 1, 1)
 
-struct sk_buff **eth_gro_receive(struct sk_buff **head,
-				 struct sk_buff *skb);
+struct sk_buff *eth_gro_receive(struct list_head *head, struct sk_buff *skb);
 int eth_gro_complete(struct sk_buff *skb, int nhoff);
 
 /* Reserved Ethernet Addresses per IEEE 802.1Q */

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
@@ -322,7 +322,7 @@ struct napi_struct {
 	int			poll_owner;
 #endif
 	struct net_device	*dev;
-	struct sk_buff		*gro_list;
+	struct list_head	gro_list;
 	struct sk_buff		*skb;
 	struct hrtimer		timer;
 	struct list_head	dev_list;
@@ -2255,10 +2255,10 @@ static inline int gro_recursion_inc_test(struct sk_buff *skb)
 	return ++NAPI_GRO_CB(skb)->recursion_counter == GRO_RECURSION_LIMIT;
 }
 
-typedef struct sk_buff **(*gro_receive_t)(struct sk_buff **, struct sk_buff *);
-static inline struct sk_buff **call_gro_receive(gro_receive_t cb,
-						struct sk_buff **head,
-						struct sk_buff *skb)
+typedef struct sk_buff *(*gro_receive_t)(struct list_head *, struct sk_buff *);
+static inline struct sk_buff *call_gro_receive(gro_receive_t cb,
+					       struct list_head *head,
+					       struct sk_buff *skb)
 {
 	if (unlikely(gro_recursion_inc_test(skb))) {
 		NAPI_GRO_CB(skb)->flush |= 1;
@@ -2268,12 +2268,12 @@ static inline struct sk_buff **call_gro_receive(gro_receive_t cb,
 	return cb(head, skb);
 }
 
-typedef struct sk_buff **(*gro_receive_sk_t)(struct sock *, struct sk_buff **,
-					     struct sk_buff *);
-static inline struct sk_buff **call_gro_receive_sk(gro_receive_sk_t cb,
-						   struct sock *sk,
-						   struct sk_buff **head,
-						   struct sk_buff *skb)
+typedef struct sk_buff *(*gro_receive_sk_t)(struct sock *, struct list_head *,
+					    struct sk_buff *);
+static inline struct sk_buff *call_gro_receive_sk(gro_receive_sk_t cb,
+						  struct sock *sk,
+						  struct list_head *head,
+						  struct sk_buff *skb)
 {
 	if (unlikely(gro_recursion_inc_test(skb))) {
 		NAPI_GRO_CB(skb)->flush |= 1;
@@ -2299,8 +2299,8 @@ struct packet_type {
 struct offload_callbacks {
 	struct sk_buff		*(*gso_segment)(struct sk_buff *skb,
 						netdev_features_t features);
-	struct sk_buff		**(*gro_receive)(struct sk_buff **head,
-						 struct sk_buff *skb);
+	struct sk_buff		*(*gro_receive)(struct list_head *head,
+						struct sk_buff *skb);
 	int			(*gro_complete)(struct sk_buff *skb, int nhoff);
 };
 
@@ -2568,7 +2568,7 @@ struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex);
 struct net_device *dev_get_by_napi_id(unsigned int napi_id);
 int netdev_get_name(struct net *net, char *name, int ifindex);
 int dev_restart(struct net_device *dev);
-int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb);
+int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb);
 
 static inline unsigned int skb_gro_offset(const struct sk_buff *skb)
 {
@@ -2784,13 +2784,13 @@ static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb,
 }
 
 #ifdef CONFIG_XFRM_OFFLOAD
-static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, int flush)
+static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush)
 {
 	if (PTR_ERR(pp) != -EINPROGRESS)
 		NAPI_GRO_CB(skb)->flush |= flush;
 }
 #else
-static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, int flush)
+static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush)
 {
 	NAPI_GRO_CB(skb)->flush |= flush;
 }

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
@@ -677,7 +677,8 @@ struct sk_buff {
 				int			ip_defrag_offset;
 			};
 		};
-		struct rb_node	rbnode; /* used in netem & tcp stack */
+		struct rb_node		rbnode; /* used in netem & tcp stack */
+		struct list_head	list;
 	};
 	struct sock		*sk;
 

diff --git a/include/linux/udp.h b/include/linux/udp.h
@@ -74,8 +74,8 @@ struct udp_sock {
 	void (*encap_destroy)(struct sock *sk);
 
 	/* GRO functions for UDP socket */
-	struct sk_buff **	(*gro_receive)(struct sock *sk,
-					       struct sk_buff **head,
+	struct sk_buff *	(*gro_receive)(struct sock *sk,
+					       struct list_head *head,
 					       struct sk_buff *skb);
 	int			(*gro_complete)(struct sock *sk,
 						struct sk_buff *skb,

diff --git a/include/net/inet_common.h b/include/net/inet_common.h
@@ -43,7 +43,7 @@ int inet_ctl_sock_create(struct sock **sk, unsigned short family,
 int inet_recv_error(struct sock *sk, struct msghdr *msg, int len,
 		    int *addr_len);
 
-struct sk_buff **inet_gro_receive(struct sk_buff **head, struct sk_buff *skb);
+struct sk_buff *inet_gro_receive(struct list_head *head, struct sk_buff *skb);
 int inet_gro_complete(struct sk_buff *skb, int nhoff);
 struct sk_buff *inet_gso_segment(struct sk_buff *skb,
 				 netdev_features_t features);

diff --git a/include/net/tcp.h b/include/net/tcp.h
@@ -1788,7 +1788,7 @@ void tcp_v4_destroy_sock(struct sock *sk);
 
 struct sk_buff *tcp_gso_segment(struct sk_buff *skb,
 				netdev_features_t features);
-struct sk_buff **tcp_gro_receive(struct sk_buff **head, struct sk_buff *skb);
+struct sk_buff *tcp_gro_receive(struct list_head *head, struct sk_buff *skb);
 int tcp_gro_complete(struct sk_buff *skb);
 
 void __tcp_v4_send_check(struct sk_buff *skb, __be32 saddr, __be32 daddr);

diff --git a/include/net/udp.h b/include/net/udp.h
@@ -170,8 +170,8 @@ static inline void udp_csum_pull_header(struct sk_buff *skb)
 typedef struct sock *(*udp_lookup_t)(struct sk_buff *skb, __be16 sport,
 				     __be16 dport);
 
-struct sk_buff **udp_gro_receive(struct sk_buff **head, struct sk_buff *skb,
-				 struct udphdr *uh, udp_lookup_t lookup);
+struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb,
+				struct udphdr *uh, udp_lookup_t lookup);
 int udp_gro_complete(struct sk_buff *skb, int nhoff, udp_lookup_t lookup);
 
 struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,

diff --git a/include/net/udp_tunnel.h b/include/net/udp_tunnel.h
@@ -65,9 +65,9 @@ static inline int udp_sock_create(struct net *net,
 
 typedef int (*udp_tunnel_encap_rcv_t)(struct sock *sk, struct sk_buff *skb);
 typedef void (*udp_tunnel_encap_destroy_t)(struct sock *sk);
-typedef struct sk_buff **(*udp_tunnel_gro_receive_t)(struct sock *sk,
-						     struct sk_buff **head,
-						     struct sk_buff *skb);
+typedef struct sk_buff *(*udp_tunnel_gro_receive_t)(struct sock *sk,
+						    struct list_head *head,
+						    struct sk_buff *skb);
 typedef int (*udp_tunnel_gro_complete_t)(struct sock *sk, struct sk_buff *skb,
 					 int nhoff);
 

diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c
@@ -647,13 +647,14 @@ static int vlan_ioctl_handler(struct net *net, void __user *arg)
 	return err;
 }
 
-static struct sk_buff **vlan_gro_receive(struct sk_buff **head,
-					 struct sk_buff *skb)
+static struct sk_buff *vlan_gro_receive(struct list_head *head,
+					struct sk_buff *skb)
 {
-	struct sk_buff *p, **pp = NULL;
-	struct vlan_hdr *vhdr;
-	unsigned int hlen, off_vlan;
 	const struct packet_offload *ptype;
+	unsigned int hlen, off_vlan;
+	struct sk_buff *pp = NULL;
+	struct vlan_hdr *vhdr;
+	struct sk_buff *p;
 	__be16 type;
 	int flush = 1;
 
@@ -675,7 +676,7 @@ static struct sk_buff **vlan_gro_receive(struct sk_buff **head,
 
 	flush = 0;
 
-	for (p = *head; p; p = p->next) {
+	list_for_each_entry(p, head, list) {
 		struct vlan_hdr *vhdr2;
 
 		if (!NAPI_GRO_CB(p)->same_flow)

diff --git a/net/core/dev.c b/net/core/dev.c
@@ -4881,36 +4881,25 @@ static int napi_gro_complete(struct sk_buff *skb)
  */
 void napi_gro_flush(struct napi_struct *napi, bool flush_old)
 {
-	struct sk_buff *skb, *prev = NULL;
-
-	/* scan list and build reverse chain */
-	for (skb = napi->gro_list; skb != NULL; skb = skb->next) {
-		skb->prev = prev;
-		prev = skb;
-	}
-
-	for (skb = prev; skb; skb = prev) {
-		skb->next = NULL;
+	struct sk_buff *skb, *p;
 
+	list_for_each_entry_safe_reverse(skb, p, &napi->gro_list, list) {
 		if (flush_old && NAPI_GRO_CB(skb)->age == jiffies)
 			return;
-
-		prev = skb->prev;
+		list_del_init(&skb->list);
 		napi_gro_complete(skb);
 		napi->gro_count--;
 	}
-
-	napi->gro_list = NULL;
 }
 EXPORT_SYMBOL(napi_gro_flush);
 
 static void gro_list_prepare(struct napi_struct *napi, struct sk_buff *skb)
 {
-	struct sk_buff *p;
 	unsigned int maclen = skb->dev->hard_header_len;
 	u32 hash = skb_get_hash_raw(skb);
+	struct sk_buff *p;
 
-	for (p = napi->gro_list; p; p = p->next) {
+	list_for_each_entry(p, &napi->gro_list, list) {
 		unsigned long diffs;
 
 		NAPI_GRO_CB(p)->flush = 0;
@@ -4977,12 +4966,12 @@ static void gro_pull_from_frag0(struct sk_buff *skb, int grow)
 
 static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
 {
-	struct sk_buff **pp = NULL;
+	struct list_head *head = &offload_base;
 	struct packet_offload *ptype;
 	__be16 type = skb->protocol;
-	struct list_head *head = &offload_base;
-	int same_flow;
+	struct sk_buff *pp = NULL;
 	enum gro_result ret;
+	int same_flow;
 	int grow;
 
 	if (netif_elide_gro(skb->dev))
@@ -5039,11 +5028,8 @@ static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff
 	ret = NAPI_GRO_CB(skb)->free ? GRO_MERGED_FREE : GRO_MERGED;
 
 	if (pp) {
-		struct sk_buff *nskb = *pp;
-
-		*pp = nskb->next;
-		nskb->next = NULL;
-		napi_gro_complete(nskb);
+		list_del_init(&pp->list);
+		napi_gro_complete(pp);
 		napi->gro_count--;
 	}
 
@@ -5054,15 +5040,10 @@ static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff
 		goto normal;
 
 	if (unlikely(napi->gro_count >= MAX_GRO_SKBS)) {
-		struct sk_buff *nskb = napi->gro_list;
+		struct sk_buff *nskb;
 
-		/* locate the end of the list to select the 'oldest' flow */
-		while (nskb->next) {
-			pp = &nskb->next;
-			nskb = *pp;
-		}
-		*pp = NULL;
-		nskb->next = NULL;
+		nskb = list_last_entry(&napi->gro_list, struct sk_buff, list);
+		list_del(&nskb->list);
 		napi_gro_complete(nskb);
 	} else {
 		napi->gro_count++;
@@ -5071,8 +5052,7 @@ static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff
 	NAPI_GRO_CB(skb)->age = jiffies;
 	NAPI_GRO_CB(skb)->last = skb;
 	skb_shinfo(skb)->gso_size = skb_gro_len(skb);
-	skb->next = napi->gro_list;
-	napi->gro_list = skb;
+	list_add(&skb->list, &napi->gro_list);
 	ret = GRO_HELD;
 
 pull:
@@ -5478,7 +5458,7 @@ bool napi_complete_done(struct napi_struct *n, int work_done)
 				 NAPIF_STATE_IN_BUSY_POLL)))
 		return false;
 
-	if (n->gro_list) {
+	if (!list_empty(&n->gro_list)) {
 		unsigned long timeout = 0;
 
 		if (work_done)
@@ -5687,7 +5667,7 @@ static enum hrtimer_restart napi_watchdog(struct hrtimer *timer)
 	/* Note : we use a relaxed variant of napi_schedule_prep() not setting
 	 * NAPI_STATE_MISSED, since we do not react to a device IRQ.
 	 */
-	if (napi->gro_list && !napi_disable_pending(napi) &&
+	if (!list_empty(&napi->gro_list) && !napi_disable_pending(napi) &&
 	    !test_and_set_bit(NAPI_STATE_SCHED, &napi->state))
 		__napi_schedule_irqoff(napi);
 
@@ -5701,7 +5681,7 @@ void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
 	hrtimer_init(&napi->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED);
 	napi->timer.function = napi_watchdog;
 	napi->gro_count = 0;
-	napi->gro_list = NULL;
+	INIT_LIST_HEAD(&napi->gro_list);
 	napi->skb = NULL;
 	napi->poll = poll;
 	if (weight > NAPI_POLL_WEIGHT)
@@ -5734,6 +5714,14 @@ void napi_disable(struct napi_struct *n)
 }
 EXPORT_SYMBOL(napi_disable);
 
+static void gro_list_free(struct list_head *head)
+{
+	struct sk_buff *skb, *p;
+
+	list_for_each_entry_safe(skb, p, head, list)
+		kfree_skb(skb);
+}
+
 /* Must be called in process context */
 void netif_napi_del(struct napi_struct *napi)
 {
@@ -5743,8 +5731,8 @@ void netif_napi_del(struct napi_struct *napi)
 	list_del_init(&napi->dev_list);
 	napi_free_frags(napi);
 
-	kfree_skb_list(napi->gro_list);
-	napi->gro_list = NULL;
+	gro_list_free(&napi->gro_list);
+	INIT_LIST_HEAD(&napi->gro_list);
 	napi->gro_count = 0;
 }
 EXPORT_SYMBOL(netif_napi_del);
@@ -5787,7 +5775,7 @@ static int napi_poll(struct napi_struct *n, struct list_head *repoll)
 		goto out_unlock;
 	}
 
-	if (n->gro_list) {
+	if (!list_empty(&n->gro_list)) {
 		/* flush too old packets
 		 * If HZ < 1000, flush all packets.
 		 */