gro: Optimise length comparison in skb_gro_header

By caching frag0_len, we can avoid checking both frag0 and the
length separately in skb_gro_header.  This helps as skb_gro_header
is called four times per packet which amounts to a few million
times at 10Gb/s.

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

1 files changed, 4 insertions, 1 deletions

diff --git a/net/core/dev.c b/net/core/dev.c
index f9d90c56b6f0..b1722a2d1fbe 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2509,12 +2509,15 @@ void skb_gro_reset_offset(struct sk_buff *skb)
 {
         NAPI_GRO_CB(skb)->data_offset = 0;
         NAPI_GRO_CB(skb)->frag0 = NULL;
+        NAPI_GRO_CB(skb)->frag0_len = 0;
 
         if (skb->mac_header == skb->tail &&
-            !PageHighMem(skb_shinfo(skb)->frags[0].page))
+            !PageHighMem(skb_shinfo(skb)->frags[0].page)) {
                 NAPI_GRO_CB(skb)->frag0 =
                         page_address(skb_shinfo(skb)->frags[0].page) +
                         skb_shinfo(skb)->frags[0].page_offset;
+                NAPI_GRO_CB(skb)->frag0_len = skb_shinfo(skb)->frags[0].size;
+        }
 }
 EXPORT_SYMBOL(skb_gro_reset_offset);