[INET]: Fix truesize setting in ip_append_data

As it is ip_append_data only counts page fragments to the skb that
allocated it.  As such it means that the first skb gets hit with a
4K charge even though it might have only used a fraction of it while
all subsequent skb's that use the same page gets away with no charge
at all.

This bug was exposed by the UDP accounting patch.

[ The wmem_alloc bumping needs to be moved with the truesize,
  noticed by Takahiro Yasui.  -DaveM ]

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/ipv4/ip_output.c

@@ -1016,8 +1016,6 @@ int ip_append_data(struct sock *sk,
 
 				skb_fill_page_desc(skb, i, page, 0, 0);
 				frag = &skb_shinfo(skb)->frags[i];
-				skb->truesize += PAGE_SIZE;
-				atomic_add(PAGE_SIZE, &sk->sk_wmem_alloc);
 			} else {
 				err = -EMSGSIZE;
 				goto error;
@@ -1030,6 +1028,8 @@ int ip_append_data(struct sock *sk,
 			frag->size += copy;
 			skb->len += copy;
 			skb->data_len += copy;
+			skb->truesize += copy;
+			atomic_add(copy, &sk->sk_wmem_alloc);
 		}
 		offset += copy;
 		length -= copy;

net/ipv6/ip6_output.c

@@ -1316,8 +1316,6 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
 
 				skb_fill_page_desc(skb, i, page, 0, 0);
 				frag = &skb_shinfo(skb)->frags[i];
-				skb->truesize += PAGE_SIZE;
-				atomic_add(PAGE_SIZE, &sk->sk_wmem_alloc);
 			} else {
 				err = -EMSGSIZE;
 				goto error;
@@ -1330,6 +1328,8 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
 			frag->size += copy;
 			skb->len += copy;
 			skb->data_len += copy;
+			skb->truesize += copy;
+			atomic_add(copy, &sk->sk_wmem_alloc);
 		}
 		offset += copy;
 		length -= copy;