[SOCK]: Introduce sk_clone
authorArnaldo Carvalho de Melo <acme@ghostprotocols.net>
Wed, 10 Aug 2005 03:10:12 +0000 (20:10 -0700)
committerDavid S. Miller <davem@sunset.davemloft.net>
Mon, 29 Aug 2005 22:42:36 +0000 (15:42 -0700)
Out of tcp_create_openreq_child, will be used in
dccp_create_openreq_child, and is a nice sock function anyway.

Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
include/net/sock.h
net/core/sock.c
net/ipv4/tcp_minisocks.c

index bdae0a5eadf5c26afc4eb771bba93a6b6ae59380..828dc082fcb75aa50e61e4525da13ebdd794136e 100644 (file)
@@ -737,6 +737,8 @@ extern struct sock          *sk_alloc(int family,
                                          unsigned int __nocast priority,
                                          struct proto *prot, int zero_it);
 extern void                    sk_free(struct sock *sk);
+extern struct sock             *sk_clone(const struct sock *sk,
+                                         const unsigned int __nocast priority);
 
 extern struct sk_buff          *sock_wmalloc(struct sock *sk,
                                              unsigned long size, int force,
index aba31fedf2ac5d8d9f0620a254ab96c3eb10529d..ccd10fd65682202fd36942105fee17cb25fb8f51 100644 (file)
@@ -700,6 +700,80 @@ void sk_free(struct sock *sk)
        module_put(owner);
 }
 
+struct sock *sk_clone(const struct sock *sk, const unsigned int __nocast priority)
+{
+       struct sock *newsk = sk_alloc(sk->sk_family, priority, sk->sk_prot, 0);
+
+       if (newsk != NULL) {
+               struct sk_filter *filter;
+
+               memcpy(newsk, sk, sk->sk_prot->obj_size);
+
+               /* SANITY */
+               sk_node_init(&newsk->sk_node);
+               sock_lock_init(newsk);
+               bh_lock_sock(newsk);
+
+               atomic_set(&newsk->sk_rmem_alloc, 0);
+               atomic_set(&newsk->sk_wmem_alloc, 0);
+               atomic_set(&newsk->sk_omem_alloc, 0);
+               skb_queue_head_init(&newsk->sk_receive_queue);
+               skb_queue_head_init(&newsk->sk_write_queue);
+
+               rwlock_init(&newsk->sk_dst_lock);
+               rwlock_init(&newsk->sk_callback_lock);
+
+               newsk->sk_dst_cache     = NULL;
+               newsk->sk_wmem_queued   = 0;
+               newsk->sk_forward_alloc = 0;
+               newsk->sk_send_head     = NULL;
+               newsk->sk_backlog.head  = newsk->sk_backlog.tail = NULL;
+               newsk->sk_userlocks     = sk->sk_userlocks & ~SOCK_BINDPORT_LOCK;
+
+               sock_reset_flag(newsk, SOCK_DONE);
+               skb_queue_head_init(&newsk->sk_error_queue);
+
+               filter = newsk->sk_filter;
+               if (filter != NULL)
+                       sk_filter_charge(newsk, filter);
+
+               if (unlikely(xfrm_sk_clone_policy(newsk))) {
+                       /* It is still raw copy of parent, so invalidate
+                        * destructor and make plain sk_free() */
+                       newsk->sk_destruct = NULL;
+                       sk_free(newsk);
+                       newsk = NULL;
+                       goto out;
+               }
+
+               newsk->sk_err      = 0;
+               newsk->sk_priority = 0;
+               atomic_set(&newsk->sk_refcnt, 2);
+
+               /*
+                * Increment the counter in the same struct proto as the master
+                * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that
+                * is the same as sk->sk_prot->socks, as this field was copied
+                * with memcpy).
+                *
+                * This _changes_ the previous behaviour, where
+                * tcp_create_openreq_child always was incrementing the
+                * equivalent to tcp_prot->socks (inet_sock_nr), so this have
+                * to be taken into account in all callers. -acme
+                */
+               sk_refcnt_debug_inc(newsk);
+               newsk->sk_socket = NULL;
+               newsk->sk_sleep  = NULL;
+
+               if (newsk->sk_prot->sockets_allocated)
+                       atomic_inc(newsk->sk_prot->sockets_allocated);
+       }
+out:
+       return newsk;
+}
+
+EXPORT_SYMBOL_GPL(sk_clone);
+
 void __init sk_init(void)
 {
        if (num_physpages <= 4096) {
index 66ce1790a94f813bfae960122596a62f8e8a10b6..8b6cd8d8066221147c4da34c64652b6c258a48bb 100644 (file)
@@ -599,67 +599,26 @@ out:
  */
 struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, struct sk_buff *skb)
 {
-       /* allocate the newsk from the same slab of the master sock,
-        * if not, at sk_free time we'll try to free it from the wrong
-        * slabcache (i.e. is it TCPv4 or v6?), this is handled thru sk->sk_prot -acme */
-       struct sock *newsk = sk_alloc(PF_INET, GFP_ATOMIC, sk->sk_prot, 0);
+       struct sock *newsk = sk_clone(sk, GFP_ATOMIC);
 
-       if(newsk != NULL) {
+       if (newsk != NULL) {
                struct inet_request_sock *ireq = inet_rsk(req);
                struct tcp_request_sock *treq = tcp_rsk(req);
                struct inet_sock *newinet = inet_sk(newsk);
                struct tcp_sock *newtp;
-               struct sk_filter *filter;
 
-               memcpy(newsk, sk, sizeof(struct tcp_sock));
                newsk->sk_state = TCP_SYN_RECV;
-
-               /* SANITY */
-               sk_node_init(&newsk->sk_node);
                newinet->bind_hash = NULL;
 
                /* Clone the TCP header template */
                newinet->dport = ireq->rmt_port;
-
-               sock_lock_init(newsk);
-               bh_lock_sock(newsk);
-
-               rwlock_init(&newsk->sk_dst_lock);
-               newsk->sk_dst_cache = NULL;
-               atomic_set(&newsk->sk_rmem_alloc, 0);
-               skb_queue_head_init(&newsk->sk_receive_queue);
-               atomic_set(&newsk->sk_wmem_alloc, 0);
-               skb_queue_head_init(&newsk->sk_write_queue);
-               atomic_set(&newsk->sk_omem_alloc, 0);
-               newsk->sk_wmem_queued = 0;
-               newsk->sk_forward_alloc = 0;
-
-               sock_reset_flag(newsk, SOCK_DONE);
-               newsk->sk_userlocks = sk->sk_userlocks & ~SOCK_BINDPORT_LOCK;
-               newsk->sk_backlog.head = newsk->sk_backlog.tail = NULL;
-               newsk->sk_send_head = NULL;
-               rwlock_init(&newsk->sk_callback_lock);
-               skb_queue_head_init(&newsk->sk_error_queue);
                newsk->sk_write_space = sk_stream_write_space;
 
-               if ((filter = newsk->sk_filter) != NULL)
-                       sk_filter_charge(newsk, filter);
-
-               if (unlikely(xfrm_sk_clone_policy(newsk))) {
-                       /* It is still raw copy of parent, so invalidate
-                        * destructor and make plain sk_free() */
-                       newsk->sk_destruct = NULL;
-                       sk_free(newsk);
-                       return NULL;
-               }
-
                /* Now setup tcp_sock */
                newtp = tcp_sk(newsk);
                newtp->pred_flags = 0;
                newtp->rcv_nxt = treq->rcv_isn + 1;
-               newtp->snd_nxt = treq->snt_isn + 1;
-               newtp->snd_una = treq->snt_isn + 1;
-               newtp->snd_sml = treq->snt_isn + 1;
+               newtp->snd_nxt = newtp->snd_una = newtp->snd_sml = treq->snt_isn + 1;
 
                tcp_prequeue_init(newtp);
 
@@ -710,32 +669,9 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req,
                /* Deinitialize accept_queue to trap illegal accesses. */
                memset(&newtp->accept_queue, 0, sizeof(newtp->accept_queue));
 
-               /* Back to base struct sock members. */
-               newsk->sk_err = 0;
-               newsk->sk_priority = 0;
-               atomic_set(&newsk->sk_refcnt, 2);
-
-               /*
-                * Increment the counter in the same struct proto as the master
-                * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that
-                * is the same as sk->sk_prot->socks, as this field was copied
-                * with memcpy), same rationale as the first comment in this
-                * function.
-                *
-                * This _changes_ the previous behaviour, where
-                * tcp_create_openreq_child always was incrementing the
-                * equivalent to tcp_prot->socks (inet_sock_nr), so this have
-                * to be taken into account in all callers. -acme
-                */
-               sk_refcnt_debug_inc(newsk);
-
-               atomic_inc(&tcp_sockets_allocated);
-
                if (sock_flag(newsk, SOCK_KEEPOPEN))
                        tcp_reset_keepalive_timer(newsk,
                                                  keepalive_time_when(newtp));
-               newsk->sk_socket = NULL;
-               newsk->sk_sleep = NULL;
 
                newtp->rx_opt.tstamp_ok = ireq->tstamp_ok;
                if((newtp->rx_opt.sack_ok = ireq->sack_ok) != 0) {