Re: net: af_packet: skb_orphan should be avoided in TX path.

[Eric Dumazet <eric.dumazet@gmail.com>]

Le lundi 06 septembre 2010 à 01:18 +0800, Changli Gao a écrit :
> af_packet uses tpacket_destruct_skb() to notify its user a frame is
> sent out through NIC, and the memory for that frame is available for
> the others. If the driver calls skb_orphan() before the frame is sent
> out successfully, and the user may fill other data into the space for
> this frame, this frame will be corrupted. It became more likely after
> skb_try_orphan() was added into dev_hard_start_xmit().
> 
> Am I correct?
> 

Yes good catch. We might add a :

SKBTX_NO_EARLY_ORPHAN = 1 << 4,

so that skb_orphan_try() do not early orphan this kind of skb


diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index f900ffc..9c1a480 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -176,6 +176,9 @@ enum {
 
 	/* ensure the originating sk reference is available on driver level */
 	SKBTX_DRV_NEEDS_SK_REF = 1 << 3,
+
+	/* dont early orphan this skb in skb_orphan_try() */
+	SKBTX_NO_EARLY_ORPHAN = 1 << 4,
 };
 
 /* This data is invariant across clones and lives at
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 3616f27..306795d 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -1029,6 +1029,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg)
 		}
 
 		skb->destructor = tpacket_destruct_skb;
+		skb_shinfo(skb)->tx_flags |= SKBTX_NO_EARLY_ORPHAN;
 		__packet_set_status(po, ph, TP_STATUS_SENDING);
 		atomic_inc(&po->tx_ring.pending);