[SK_BUFF]: Convert skb->tail to sk_buff_data_t

So that it is also an offset from skb->head, reduces its size from 8 to 4 bytes
on 64bit architectures, allowing us to combine the 4 bytes hole left by the
layer headers conversion, reducing struct sk_buff size to 256 bytes, i.e. 4
64byte cachelines, and since the sk_buff slab cache is SLAB_HWCACHE_ALIGN...
:-)

Many calculations that previously required that skb->{transport,network,
mac}_header be first converted to a pointer now can be done directly, being
meaningful as offsets or pointers.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 2 insertions, 2 deletions

diff --git a/net/core/gen_stats.c b/net/core/gen_stats.c
index 259473d0559d..bcc25591d8ac 100644
--- a/net/core/gen_stats.c
+++ b/net/core/gen_stats.c
@@ -61,7 +61,7 @@ gnet_stats_start_copy_compat(struct sk_buff *skb, int type, int tc_stats_type,
         spin_lock_bh(lock);
         d->lock = lock;
         if (type)
-                d->tail = (struct rtattr *) skb->tail;
+                d->tail = (struct rtattr *)skb_tail_pointer(skb);
         d->skb = skb;
         d->compat_tc_stats = tc_stats_type;
         d->compat_xstats = xstats_type;
@@ -212,7 +212,7 @@ int
 gnet_stats_finish_copy(struct gnet_dump *d)
 {
         if (d->tail)
-                d->tail->rta_len = d->skb->tail - (u8 *) d->tail;
+                d->tail->rta_len = skb_tail_pointer(d->skb) - (u8 *)d->tail;
 
         if (d->compat_tc_stats)
                 if (gnet_stats_copy(d, d->compat_tc_stats, &d->tc_stats,