Re: [PATCH 00/21] slight optimization of addr compare for some modules

On Mon, 2013-12-23 at 13:09 +0800, Ding Tianhong wrote:

I use the recently added and possibly more efficient
ether_addr_equal_unaligned to instead of memcmp for slight
optimization.

Hi again.

These are the sorts of conversions that could be
either ether_addr_equal or ether_addr_equal_unaligned
depending on the existing alignment of the addresses
being compared.

For instance, in patch 21:
---

diff --git a/drivers/net/ppp/pppoe.c b/drivers/net/ppp/pppoe.c

[]

@@ -131,12 +131,13 @@ static inline struct pppoe_net *pppoe_pernet(struct net *net)
 
 static inline int cmp_2_addr(struct pppoe_addr *a, struct pppoe_addr *b)
 {
-	return a->sid == b->sid && !memcmp(a->remote, b->remote, ETH_ALEN);
+	return a->sid == b->sid && ether_addr_equal_unaligned(a->remote,
+		ther					      b->remote);
---

This could be ether_addr_equal instead as remote is after a be16
and so remote is aligned to a u16.

include/uapi/linux/if_pppox.h:typedef __be16 sid_t;
[]
include/uapi/linux/if_pppox.h:struct pppoe_addr {
include/uapi/linux/if_pppox.h-  sid_t         sid;                    /* Session identifier */
include/uapi/linux/if_pppox.h-  unsigned char remote[ETH_ALEN];       /* Remote address */
include/uapi/linux/if_pppox.h-  char          dev[IFNAMSIZ];          /* Local device to use */
include/uapi/linux/if_pppox.h-}; 

It's of course possible to convert the appropriate
ether_addr_equal_unaligned uses to ether_addr_equal later.