Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6

* 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: (44 commits)
  [NETLINK]: Fix unicast timeouts
  [INET]: Remove per bucket rwlock in tcp/dccp ehash table.
  [IPVS]: Synchronize closing of Connections
  [IPVS]: Bind connections on stanby if the destination exists
  [NET]: Remove Documentation/networking/pt.txt
  [NET]: Remove Documentation/networking/routing.txt
  [NET]: Remove Documentation/networking/ncsa-telnet
  [NET]: Remove comx driver docs.
  [NET]: Remove Documentation/networking/Configurable
  [NET]: Clean proto_(un)register from in-code ifdefs
  [IPSEC]: Fix crypto_alloc_comp error checking
  [VLAN]: Fix SET_VLAN_INGRESS_PRIORITY_CMD ioctl
  [NETNS]: Fix compiler error in net_namespace.c
  [TTY]: Use tty_mode_ioctl() in network drivers.
  [TTY]: Fix network driver interactions with TCGET/SET calls.
  [PKT_SCHED] CLS_U32: Fix endianness problem with u32 classifier hash masks.
  [NET]: Removing duplicit #includes
  [NET]: Let USB_USBNET always select MII.
  [RRUNNER]: Do not muck with sysctl_{r,w}mem_max
  [DLM] lowcomms: Do not muck with sysctl_rmem_max.
  ...

4 files changed, 136 insertions, 21 deletions

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 4427dcd1e53a..8461cda37490 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -37,7 +37,6 @@
  * I'll experiment with dynamic table growth later.
  */
 struct inet_ehash_bucket {
-        rwlock_t          lock;
         struct hlist_head chain;
         struct hlist_head twchain;
 };
@@ -100,6 +99,9 @@ struct inet_hashinfo {
          * TIME_WAIT sockets use a separate chain (twchain).
          */
         struct inet_ehash_bucket        *ehash;
+        rwlock_t                        *ehash_locks;
+        unsigned int                    ehash_size;
+        unsigned int                    ehash_locks_mask;
 
         /* Ok, let's try this, I give up, we do need a local binding
          * TCP hash as well as the others for fast bind/connect.
@@ -107,7 +109,7 @@ struct inet_hashinfo {
         struct inet_bind_hashbucket     *bhash;
 
         unsigned int                    bhash_size;
-        unsigned int                    ehash_size;
+        /* Note : 4 bytes padding on 64 bit arches */
 
         /* All sockets in TCP_LISTEN state will be in here.  This is the only
          * table where wildcard'd TCP sockets can exist.  Hash function here
@@ -134,6 +136,62 @@ static inline struct inet_ehash_bucket *inet_ehash_bucket(
         return &hashinfo->ehash[hash & (hashinfo->ehash_size - 1)];
 }
 
+static inline rwlock_t *inet_ehash_lockp(
+        struct inet_hashinfo *hashinfo,
+        unsigned int hash)
+{
+        return &hashinfo->ehash_locks[hash & hashinfo->ehash_locks_mask];
+}
+
+static inline int inet_ehash_locks_alloc(struct inet_hashinfo *hashinfo)
+{
+        unsigned int i, size = 256;
+#if defined(CONFIG_PROVE_LOCKING)
+        unsigned int nr_pcpus = 2;
+#else
+        unsigned int nr_pcpus = num_possible_cpus();
+#endif
+        if (nr_pcpus >= 4)
+                size = 512;
+        if (nr_pcpus >= 8)
+                size = 1024;
+        if (nr_pcpus >= 16)
+                size = 2048;
+        if (nr_pcpus >= 32)
+                size = 4096;
+        if (sizeof(rwlock_t) != 0) {
+#ifdef CONFIG_NUMA
+                if (size * sizeof(rwlock_t) > PAGE_SIZE)
+                        hashinfo->ehash_locks = vmalloc(size * sizeof(rwlock_t));
+                else
+#endif
+                hashinfo->ehash_locks = kmalloc(size * sizeof(rwlock_t),
+                                                GFP_KERNEL);
+                if (!hashinfo->ehash_locks)
+                        return ENOMEM;
+                for (i = 0; i < size; i++)
+                        rwlock_init(&hashinfo->ehash_locks[i]);
+        }
+        hashinfo->ehash_locks_mask = size - 1;
+        return 0;
+}
+
+static inline void inet_ehash_locks_free(struct inet_hashinfo *hashinfo)
+{
+        if (hashinfo->ehash_locks) {
+#ifdef CONFIG_NUMA
+                unsigned int size = (hashinfo->ehash_locks_mask + 1) *
+                                                        sizeof(rwlock_t);
+                if (size > PAGE_SIZE)
+                        vfree(hashinfo->ehash_locks);
+                else
+#else
+                kfree(hashinfo->ehash_locks);
+#endif
+                hashinfo->ehash_locks = NULL;
+        }
+}
+
 extern struct inet_bind_bucket *
                     inet_bind_bucket_create(struct kmem_cache *cachep,
                                             struct inet_bind_hashbucket *head,
@@ -222,7 +280,7 @@ static inline void __inet_hash(struct inet_hashinfo *hashinfo,
                 sk->sk_hash = inet_sk_ehashfn(sk);
                 head = inet_ehash_bucket(hashinfo, sk->sk_hash);
                 list = &head->chain;
-                lock = &head->lock;
+                lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
                 write_lock(lock);
         }
         __sk_add_node(sk, list);
@@ -253,7 +311,7 @@ static inline void inet_unhash(struct inet_hashinfo *hashinfo, struct sock *sk)
                 inet_listen_wlock(hashinfo);
                 lock = &hashinfo->lhash_lock;
         } else {
-                lock = &inet_ehash_bucket(hashinfo, sk->sk_hash)->lock;
+                lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
                 write_lock_bh(lock);
         }
 
@@ -354,9 +412,10 @@ static inline struct sock *
          */
         unsigned int hash = inet_ehashfn(daddr, hnum, saddr, sport);
         struct inet_ehash_bucket *head = inet_ehash_bucket(hashinfo, hash);
+        rwlock_t *lock = inet_ehash_lockp(hashinfo, hash);
 
         prefetch(head->chain.first);
-        read_lock(&head->lock);
+        read_lock(lock);
         sk_for_each(sk, node, &head->chain) {
                 if (INET_MATCH(sk, hash, acookie, saddr, daddr, ports, dif))
                         goto hit; /* You sunk my battleship! */
@@ -369,7 +428,7 @@ static inline struct sock *
         }
         sk = NULL;
 out:
-        read_unlock(&head->lock);
+        read_unlock(lock);
         return sk;
 hit:
         sock_hold(sk);
diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 8cadc77c7df4..ed514bfb61ba 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -185,6 +185,12 @@ static inline void fib_select_default(const struct flowi *flp, struct fib_result
 }
 
 #else /* CONFIG_IP_MULTIPLE_TABLES */
+extern void __init fib4_rules_init(void);
+
+#ifdef CONFIG_NET_CLS_ROUTE
+extern u32 fib_rules_tclass(struct fib_result *res);
+#endif
+
 #define ip_fib_local_table fib_get_table(RT_TABLE_LOCAL)
 #define ip_fib_main_table fib_get_table(RT_TABLE_MAIN)
 
@@ -214,15 +220,6 @@ extern __be32  __fib_res_prefsrc(struct fib_result *res);
 /* Exported by fib_hash.c */
 extern struct fib_table *fib_hash_init(u32 id);
 
-#ifdef CONFIG_IP_MULTIPLE_TABLES
-extern void __init fib4_rules_init(void);
-
-#ifdef CONFIG_NET_CLS_ROUTE
-extern u32 fib_rules_tclass(struct fib_result *res);
-#endif
-
-#endif
-
 static inline void fib_combine_itag(u32 *itag, struct fib_result *res)
 {
 #ifdef CONFIG_NET_CLS_ROUTE
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 41870564df8e..67ea2c0c0ab7 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -520,6 +520,10 @@ struct ip_vs_conn {
         spinlock_t              lock;           /* lock for state transition */
         volatile __u16          flags;          /* status flags */
         volatile __u16          state;          /* state info */
+        volatile __u16          old_state;      /* old state, to be used for
+                                                 * state transition triggerd
+                                                 * synchronization
+                                                 */
 
         /* Control members */
         struct ip_vs_conn       *control;       /* Master control connection */
@@ -901,6 +905,10 @@ extern int ip_vs_use_count_inc(void);
 extern void ip_vs_use_count_dec(void);
 extern int ip_vs_control_init(void);
 extern void ip_vs_control_cleanup(void);
+extern struct ip_vs_dest *
+ip_vs_find_dest(__be32 daddr, __be16 dport,
+                 __be32 vaddr, __be16 vport, __u16 protocol);
+extern struct ip_vs_dest *ip_vs_try_bind_dest(struct ip_vs_conn *cp);
 
 
 /*
diff --git a/include/net/sock.h b/include/net/sock.h
index 20de3fa7ae40..5504fb9fa88a 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -560,6 +560,14 @@ struct proto {
         void                    (*unhash)(struct sock *sk);
         int                     (*get_port)(struct sock *sk, unsigned short snum);
 
+#ifdef CONFIG_SMP
+        /* Keeping track of sockets in use */
+        void                    (*inuse_add)(struct proto *prot, int inc);
+        int                     (*inuse_getval)(const struct proto *prot);
+        int                     *inuse_ptr;
+#else
+        int                     inuse;
+#endif
         /* Memory pressure */
         void                    (*enter_memory_pressure)(void);
         atomic_t                *memory_allocated;      /* Current allocated memory. */
@@ -592,12 +600,38 @@ struct proto {
 #ifdef SOCK_REFCNT_DEBUG
         atomic_t                socks;
 #endif
-        struct {
-                int inuse;
-                u8  __pad[SMP_CACHE_BYTES - sizeof(int)];
-        } stats[NR_CPUS];
 };
 
+/*
+ * Special macros to let protos use a fast version of inuse{get|add}
+ * using a static percpu variable per proto instead of an allocated one,
+ * saving one dereference.
+ * This might be changed if/when dynamic percpu vars become fast.
+ */
+#ifdef CONFIG_SMP
+# define DEFINE_PROTO_INUSE(NAME)                       \
+static DEFINE_PER_CPU(int, NAME##_inuse);               \
+static void NAME##_inuse_add(struct proto *prot, int inc)       \
+{                                                       \
+        __get_cpu_var(NAME##_inuse) += inc;             \
+}                                                       \
+                                                        \
+static int NAME##_inuse_getval(const struct proto *prot)\
+{                                                       \
+        int res = 0, cpu;                               \
+                                                        \
+        for_each_possible_cpu(cpu)                      \
+                res += per_cpu(NAME##_inuse, cpu);      \
+        return res;                                     \
+}
+# define REF_PROTO_INUSE(NAME)                          \
+        .inuse_add = NAME##_inuse_add,                  \
+        .inuse_getval = NAME##_inuse_getval,
+#else
+# define DEFINE_PROTO_INUSE(NAME)
+# define REF_PROTO_INUSE(NAME)
+#endif
+
 extern int proto_register(struct proto *prot, int alloc_slab);
 extern void proto_unregister(struct proto *prot);
 
@@ -629,12 +663,29 @@ static inline void sk_refcnt_debug_release(const struct sock *sk)
 /* Called with local bh disabled */
 static __inline__ void sock_prot_inc_use(struct proto *prot)
 {
-        prot->stats[smp_processor_id()].inuse++;
+#ifdef CONFIG_SMP
+        prot->inuse_add(prot, 1);
+#else
+        prot->inuse++;
+#endif
 }
 
 static __inline__ void sock_prot_dec_use(struct proto *prot)
 {
-        prot->stats[smp_processor_id()].inuse--;
+#ifdef CONFIG_SMP
+        prot->inuse_add(prot, -1);
+#else
+        prot->inuse--;
+#endif
+}
+
+static __inline__ int sock_prot_inuse(struct proto *proto)
+{
+#ifdef CONFIG_SMP
+        return proto->inuse_getval(proto);
+#else
+        return proto->inuse;
+#endif
 }
 
 /* With per-bucket locks this operation is not-atomic, so that