[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>
diff --git a/drivers/infiniband/hw/amso1100/c2.c b/drivers/infiniband/hw/amso1100/c2.c
index 7698fea..58bc272 100644
--- a/drivers/infiniband/hw/amso1100/c2.c
+++ b/drivers/infiniband/hw/amso1100/c2.c
@@ -439,7 +439,8 @@
}
/* Setup the skb for reuse since we're dropping this pkt */
- elem->skb->tail = elem->skb->data = elem->skb->head;
+ elem->skb->data = elem->skb->head;
+ skb_reset_tail_pointer(elem->skb);
/* Zero out the rxp hdr in the sk_buff */
memset(elem->skb->data, 0, sizeof(*rxp_hdr));
@@ -521,7 +522,7 @@
* "sizeof(struct c2_rxp_hdr)".
*/
skb->data += sizeof(*rxp_hdr);
- skb->tail = skb->data + buflen;
+ skb_set_tail_pointer(skb, buflen);
skb->len = buflen;
skb->protocol = eth_type_trans(skb, netdev);