netif_rx() must be called under a strict contract.
At device dismantle phase, core networking clears IFF_UP
and flush_all_backlogs() is called after rcu grace period
to make sure no incoming packet might be in a cpu backlog
and still referencing the device.
Most drivers call netif_rx() from their interrupt handler,
and since the interrupts are disabled at device dismantle,
netif_rx() does not have to check dev->flags & IFF_UP
Virtual drivers do not have this guarantee, and must
therefore make the check themselves.
Otherwise we risk use-after-free and/or crashes.
Note this patch also fixes a small issue that came
with commit
ce6502a8f957 ("vxlan: fix a use after free
in vxlan_encap_bypass"), since the dev->stats.rx_dropped
change was done on the wrong device.
Fixes: d342894c5d2f ("vxlan: virtual extensible lan")
Fixes: ce6502a8f957 ("vxlan: fix a use after free in vxlan_encap_bypass")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Petr Machata <petrm@mellanox.com>
Cc: Ido Schimmel <idosch@mellanox.com>
Cc: Roopa Prabhu <roopa@cumulusnetworks.com>
Cc: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct pcpu_sw_netstats *tx_stats, *rx_stats;
union vxlan_addr loopback;
union vxlan_addr *remote_ip = &dst_vxlan->default_dst.remote_ip;
- struct net_device *dev = skb->dev;
+ struct net_device *dev;
int len = skb->len;
tx_stats = this_cpu_ptr(src_vxlan->dev->tstats);
#endif
}
+ rcu_read_lock();
+ dev = skb->dev;
+ if (unlikely(!(dev->flags & IFF_UP))) {
+ kfree_skb(skb);
+ goto drop;
+ }
+
if (dst_vxlan->cfg.flags & VXLAN_F_LEARN)
- vxlan_snoop(skb->dev, &loopback, eth_hdr(skb)->h_source, 0,
- vni);
+ vxlan_snoop(dev, &loopback, eth_hdr(skb)->h_source, 0, vni);
u64_stats_update_begin(&tx_stats->syncp);
tx_stats->tx_packets++;
rx_stats->rx_bytes += len;
u64_stats_update_end(&rx_stats->syncp);
} else {
+drop:
dev->stats.rx_dropped++;
}
+ rcu_read_unlock();
}
static int encap_bypass_if_local(struct sk_buff *skb, struct net_device *dev,