net/inet_lro: remove setting skb->ip_summed when not LRO-able

When an SKB cannot be chained to a session, the current code attempts
to "restore" its ip_summed field from lro_mgr->ip_summed. However,
lro_mgr->ip_summed does not hold the original value; in fact, we'd
better not touch skb->ip_summed since it is not modified by the code
in the path leading to a failure to chain it.  Also use a cleaer
comment to the describe the ip_summed field of struct net_lro_mgr.

Issue raised by Or Gerlitz <ogerlitz@voltaire.com>

Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>

2 files changed, 6 insertions, 3 deletions

diff --git a/include/linux/inet_lro.h b/include/linux/inet_lro.h
index 80335b7d77c5..c4335faebb63 100644
--- a/include/linux/inet_lro.h
+++ b/include/linux/inet_lro.h
@@ -84,7 +84,11 @@ struct net_lro_mgr {
                                     from received packets and eth protocol
                                     is still ETH_P_8021Q */
 
-        u32 ip_summed;      /* Set in non generated SKBs in page mode */
+        /*
+         * Set for generated SKBs that are not added to
+         * the frag list in fragmented mode
+         */
+        u32 ip_summed;
         u32 ip_summed_aggr; /* Set in aggregated SKBs: CHECKSUM_UNNECESSARY
                              * or CHECKSUM_NONE */
 
diff --git a/net/ipv4/inet_lro.c b/net/ipv4/inet_lro.c
index 4a4d49fca1f2..cfd034a2b96e 100644
--- a/net/ipv4/inet_lro.c
+++ b/net/ipv4/inet_lro.c
@@ -383,8 +383,7 @@ static int __lro_proc_skb(struct net_lro_mgr *lro_mgr, struct sk_buff *skb,
 out2: /* send aggregated SKBs to stack */
         lro_flush(lro_mgr, lro_desc);
 
-out:  /* Original SKB has to be posted to stack */
-        skb->ip_summed = lro_mgr->ip_summed;
+out:
         return 1;
 }