| Message ID | 1452617880.1223.21.camel@edumazet-glaptop2.roam.corp.google.com |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
From: Eric Dumazet <eric.dumazet@gmail.com> Date: Tue, 12 Jan 2016 08:58:00 -0800 > From: Eric Dumazet <edumazet@google.com> > > Ivaylo Dimitrov reported a regression caused by commit 7866a621043f > ("dev: add per net_device packet type chains"). > > skb->dev becomes NULL and we crash in __netif_receive_skb_core(). > > Before above commit, different kind of bugs or corruptions could happen > without major crash. > > But the root cause is that phonet_rcv() can queue skb without checking > if skb is shared or not. > > Many thanks to Ivaylo Dimitrov for his help, diagnosis and tests. > > Reported-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> > Tested-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> > Signed-off-by: Eric Dumazet <edumazet@google.com> > Cc: Remi Denis-Courmont <courmisch@gmail.com> Applied and queued up for -stable, th anks Eric.
Le 2016-01-12 18:58, Eric Dumazet a écrit : > From: Eric Dumazet <edumazet@google.com> > > Ivaylo Dimitrov reported a regression caused by commit 7866a621043f > ("dev: add per net_device packet type chains"). > > skb->dev becomes NULL and we crash in __netif_receive_skb_core(). > > Before above commit, different kind of bugs or corruptions could > happen > without major crash. Hmm... was that always a problem then, or did it get introduced earlier? I thought it was impossible for skb to be shared on PF-recv callback way back. > But the root cause is that phonet_rcv() can queue skb without > checking > if skb is shared or not. > > Many thanks to Ivaylo Dimitrov for his help, diagnosis and tests. > > Reported-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> > Tested-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> > Signed-off-by: Eric Dumazet <edumazet@google.com> > Cc: Remi Denis-Courmont <courmisch@gmail.com> > --- > net/phonet/af_phonet.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/net/phonet/af_phonet.c b/net/phonet/af_phonet.c > index 10d42f3220ab..f925753668a7 100644 > --- a/net/phonet/af_phonet.c > +++ b/net/phonet/af_phonet.c > @@ -377,6 +377,10 @@ static int phonet_rcv(struct sk_buff *skb, > struct net_device *dev, > struct sockaddr_pn sa; > u16 len; > > + skb = skb_share_check(skb, GFP_ATOMIC); > + if (!skb) > + return NET_RX_DROP; > + > /* check we have at least a full Phonet header */ > if (!pskb_pull(skb, sizeof(struct phonethdr))) > goto out; Ack, thanks.
On Wed, 2016-01-13 at 14:26 +0200, Rémi Denis-Courmont wrote: > Le 2016-01-12 18:58, Eric Dumazet a écrit : > > From: Eric Dumazet <edumazet@google.com> > > > > Ivaylo Dimitrov reported a regression caused by commit 7866a621043f > > ("dev: add per net_device packet type chains"). > > > > skb->dev becomes NULL and we crash in __netif_receive_skb_core(). > > > > Before above commit, different kind of bugs or corruptions could > > happen > > without major crash. > > Hmm... was that always a problem then, or did it get introduced > earlier? I thought it was impossible for skb to be shared on PF-recv > callback way back. Always been a problem. Use tcpdump and you risk use after free. Or simply panics if skb->head needs to be expanded. Other relevant commits to explain this : 044453b3efdc90bdd5feffe74b99d95dec70ac43 arp: fix possible crash in arp_rcv() b30532515f0a62bfe17207ab00883dd262497006 bonding: Ensure that we unshare skbs prior to calling pskb_may_pull
diff --git a/net/phonet/af_phonet.c b/net/phonet/af_phonet.c index 10d42f3220ab..f925753668a7 100644 --- a/net/phonet/af_phonet.c +++ b/net/phonet/af_phonet.c @@ -377,6 +377,10 @@ static int phonet_rcv(struct sk_buff *skb, struct net_device *dev, struct sockaddr_pn sa; u16 len; + skb = skb_share_check(skb, GFP_ATOMIC); + if (!skb) + return NET_RX_DROP; + /* check we have at least a full Phonet header */ if (!pskb_pull(skb, sizeof(struct phonethdr))) goto out;