Merge branch 'net-generic-busy-polling'

Eric Dumazet says:

====================
net: extend busy polling support

This patch series extends busy polling range to tunnels devices,
and adds busy polling generic support to all NAPI drivers.

No need to provide ndo_busy_poll() method and extra synchronization
between ndo_busy_poll() and normal napi->poll() method.
This was proven very difficult and bug prone.

mlx5 driver is changed to support busy polling using this new method,
and a second mlx5 patch adds napi_complete_done() support and proper
SNMP accounting.

bnx2x and mlx4 drivers are converted to new infrastructure,
reducing kernel bloat and improving performance.

Latest patch, adding generic support, adds a new requirement :

 -free_netdev() and netif_napi_del() must be called from process context.

Since this might not be the case in some drivers, we might have to
either : fix the non conformant drivers (by disabling busy polling on them)
or revert this last patch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

3 files changed, 36 insertions, 15 deletions

diff --git a/include/linux/hashtable.h b/include/linux/hashtable.h
index 519b6e2d769e..661e5c2a8e2a 100644
--- a/include/linux/hashtable.h
+++ b/include/linux/hashtable.h
@@ -16,6 +16,10 @@
 	struct hlist_head name[1 << (bits)] =					\
 			{ [0 ... ((1 << (bits)) - 1)] = HLIST_HEAD_INIT }
 
+#define DEFINE_READ_MOSTLY_HASHTABLE(name, bits)				\
+	struct hlist_head name[1 << (bits)] __read_mostly =			\
+			{ [0 ... ((1 << (bits)) - 1)] = HLIST_HEAD_INIT }
+
 #define DECLARE_HASHTABLE(name, bits)                                   	\
 	struct hlist_head name[1 << (bits)]
 
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 67bfac1abfc1..7d2d1d7aaec7 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -326,7 +326,8 @@ enum {
 	NAPI_STATE_SCHED,	/* Poll is scheduled */
 	NAPI_STATE_DISABLE,	/* Disable pending */
 	NAPI_STATE_NPSVC,	/* Netpoll - don't dequeue from poll_list */
-	NAPI_STATE_HASHED,	/* In NAPI hash */
+	NAPI_STATE_HASHED,	/* In NAPI hash (busy polling possible) */
+	NAPI_STATE_NO_BUSY_POLL,/* Do not add in napi_hash, no busy polling */
 };
 
 enum gro_result {
@@ -461,19 +462,13 @@ static inline void napi_complete(struct napi_struct *n)
 }
 
 /**
- *	napi_by_id - lookup a NAPI by napi_id
- *	@napi_id: hashed napi_id
- *
- * lookup @napi_id in napi_hash table
- * must be called under rcu_read_lock()
- */
-struct napi_struct *napi_by_id(unsigned int napi_id);
-
-/**
  *	napi_hash_add - add a NAPI to global hashtable
  *	@napi: napi context
  *
  * generate a new napi_id and store a @napi under it in napi_hash
+ * Used for busy polling (CONFIG_NET_RX_BUSY_POLL)
+ * Note: This is normally automatically done from netif_napi_add(),
+ * so might disappear in a future linux version.
  */
 void napi_hash_add(struct napi_struct *napi);
 
@@ -482,9 +477,14 @@ void napi_hash_add(struct napi_struct *napi);
  *	@napi: napi context
  *
  * Warning: caller must observe rcu grace period
- * before freeing memory containing @napi
+ * before freeing memory containing @napi, if
+ * this function returns true.
+ * Note: core networking stack automatically calls it
+ * from netif_napi_del()
+ * Drivers might want to call this helper to combine all
+ * the needed rcu grace periods into a single one.
  */
-void napi_hash_del(struct napi_struct *napi);
+bool napi_hash_del(struct napi_struct *napi);
 
 /**
  *	napi_disable - prevent NAPI from scheduling
@@ -1948,6 +1948,26 @@ void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
 		    int (*poll)(struct napi_struct *, int), int weight);
 
 /**
+ *	netif_tx_napi_add - initialize a napi context
+ *	@dev:  network device
+ *	@napi: napi context
+ *	@poll: polling function
+ *	@weight: default weight
+ *
+ * This variant of netif_napi_add() should be used from drivers using NAPI
+ * to exclusively poll a TX queue.
+ * This will avoid we add it into napi_hash[], thus polluting this hash table.
+ */
+static inline void netif_tx_napi_add(struct net_device *dev,
+				     struct napi_struct *napi,
+				     int (*poll)(struct napi_struct *, int),
+				     int weight)
+{
+	set_bit(NAPI_STATE_NO_BUSY_POLL, &napi->state);
+	netif_napi_add(dev, napi, poll, weight);
+}
+
+/**
  *  netif_napi_del - remove a napi context
  *  @napi: napi context
  *
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 4355129fff91..c9c394bf0771 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1082,9 +1082,6 @@ static inline void skb_copy_hash(struct sk_buff *to, const struct sk_buff *from)
 
 static inline void skb_sender_cpu_clear(struct sk_buff *skb)
 {
-#ifdef CONFIG_XPS
-	skb->sender_cpu = 0;
-#endif
 }
 
 #ifdef NET_SKBUFF_DATA_USES_OFFSET