[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANn89iJjRPnAvVQebv3V5H=b5+aaL=rTBrSiB6-TR-UcXks2TQ@mail.gmail.com>
Date: Tue, 2 Sep 2025 07:13:36 -0700
From: Eric Dumazet <edumazet@...gle.com>
To: Dan Cross <crossd@...il.com>
Cc: "David S . Miller" <davem@...emloft.net>, Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>, Simon Horman <horms@...nel.org>, netdev@...r.kernel.org,
eric.dumazet@...il.com, Bernard Pidoux <f6bvp@...e.fr>, Joerg Reuter <jreuter@...na.de>,
linux-hams@...r.kernel.org, David Ranch <dranch@...nnet.net>,
Folkert van Heusden <folkert@...heusden.com>
Subject: Re: [PATCH net] ax25: properly unshare skbs in ax25_kiss_rcv()
On Tue, Sep 2, 2025 at 5:51 AM Dan Cross <crossd@...il.com> wrote:
>
> On Tue, Sep 2, 2025 at 8:46 AM Eric Dumazet <edumazet@...gle.com> wrote:
> > Bernard Pidoux reported a regression apparently caused by commit
> > c353e8983e0d ("net: introduce per netns packet 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 a major crash.
> >
> > But the root cause is that ax25_kiss_rcv() can queue/mangle input skb
> > without checking if this skb is shared or not.
> >
> > Many thanks to Bernard Pidoux for his help, diagnosis and tests.
> >
> > We had a similar issue years ago fixed with commit 7aaed57c5c28
> > ("phonet: properly unshare skbs in phonet_rcv()").
>
> Please mention the analysis done here in the change description:
> https://lore.kernel.org/linux-hams/CAEoi9W4FGoEv+2FUKs7zc=XoLuwhhLY8f8t_xQ6MgTJyzQPxXA@mail.gmail.com/#R
>
I was not aware of all the long past discussions.
No need for me to respin this patch, there is the Closes: tag for this purpose.
Whoever wants to find all the details can follow it.
For this changelog, I basically copy/pasted my phonet change,
as it included all the relevant information for this old bug.
Thank you.
> Reviewed-by: Dan Cross <crossd@...il.com>
>
> > Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> > Reported-by: Bernard Pidoux <f6bvp@...e.fr>
> > Closes: https://lore.kernel.org/netdev/1713f383-c538-4918-bc64-13b3288cd542@free.fr/
> > Tested-by: Bernard Pidoux <f6bvp@...e.fr>
> > Signed-off-by: Eric Dumazet <edumazet@...gle.com>
> > Cc: Joerg Reuter <jreuter@...na.de>
> > Cc: linux-hams@...r.kernel.org
> > Cc: David Ranch <dranch@...nnet.net>
> > Cc: Dan Cross <crossd@...il.com>
> > Cc: Folkert van Heusden <folkert@...heusden.com>
> > ---
> > net/ax25/ax25_in.c | 4 ++++
> > 1 file changed, 4 insertions(+)
> >
> > diff --git a/net/ax25/ax25_in.c b/net/ax25/ax25_in.c
> > index 1cac25aca637..f2d66af86359 100644
> > --- a/net/ax25/ax25_in.c
> > +++ b/net/ax25/ax25_in.c
> > @@ -433,6 +433,10 @@ static int ax25_rcv(struct sk_buff *skb, struct net_device *dev,
> > int ax25_kiss_rcv(struct sk_buff *skb, struct net_device *dev,
> > struct packet_type *ptype, struct net_device *orig_dev)
> > {
> > + skb = skb_share_check(skb, GFP_ATOMIC);
> > + if (!skb)
> > + return NET_RX_DROP;
> > +
> > skb_orphan(skb);
> >
> > if (!net_eq(dev_net(dev), &init_net)) {
> > --
> > 2.51.0.318.gd7df087d1a-goog
> >
Powered by blists - more mailing lists