From: Paolo Abeni <pabeni@redhat.com>
Date: Wed, 14 Jun 2017 11:48:48 +0200
Subject: net: use skb_unref() in napi_consume_skb()
Patch-mainline: v4.13-rc1
Git-commit: 7608894e43d071ef2322a01c79522954c070ac6c
References: bsc#1056787

The commit 83ada39bb79d ("net: factor out a helper to decrement the
skb refcount") provided and used a helper for decrementing skb usage,
but I missed at least a spot for it.

This change remove some more duplicated code reusing skb_unref() in
napi_consume_skb(), too. The helper uses an additional, unneeded
unlikely(!skb) test - napi_consume_skb() already check it a few lines
above - but the compiler is smart enough to optimize the duplicated
test out.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
---
 net/core/skbuff.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -801,10 +801,9 @@ void napi_consume_skb(struct sk_buff *sk
 		return;
 	}
 
-	if (likely(atomic_read(&skb->users) == 1))
-		smp_rmb();
-	else if (likely(!atomic_dec_and_test(&skb->users)))
+	if (!skb_unref(skb))
 		return;
+
 	/* if reaching here SKB is ready to free */
 	trace_consume_skb(skb);