netdev - Re: 4.9.2 panic, __skb_flow

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAA85sZsKSJkaWJ44w9__mJ6vH9ViZwpsovOhnEZRG3sO5K-2PA@mail.gmail.com>
Date:   Wed, 11 Jan 2017 01:16:35 +0100
From:   Ian Kumlien <ian.kumlien@...il.com>
To:     Denys Fedoryshchenko <nuclearcat@...learcat.com>
Cc:     Linux Kernel Network Developers <netdev@...r.kernel.org>,
        David Miller <davem@...emloft.net>
Subject: Re: 4.9.2 panic, __skb_flow_dissect, gro?

Added David Miller to CC since he said it was queued for stable, maybe
he can comment

On Wed, Jan 11, 2017 at 12:49 AM, Denys Fedoryshchenko
<nuclearcat@...learcat.com> wrote:
> It seems this patch solve issue. I hope it will go to stable asap, because
> without it loaded routers crashing almost instantly on 4.9.

I'm also worried that you could trigger it remotely....

I suspect the following:
intel: fm10k, i40e, i40ev, igb, ixgbe, ixgbevf
mellanox: mlx4, mlx5
qlogic: qede

since skb_flow_dissect is called by eth_get_headlen in these drivers...

My machine was running with igb when it happened, is your network
driver in the list?

David: Let me know if i can help with the -stable bit in anyway, i've
been surprised to see it miss .1 and .2

> commit  d0af683407a26a4437d8fa6e283ea201f2ae8146 (patch)
> tree    e769779cf59b0b73333b50a68db5d0b8897a7cb6 /net/core/flow_dissector.c
> parent  94ba998b63c41e92da1b2f0cd8679e038181ef48 (diff)
> flow_dissector: Update pptp handling to avoid null pointer deref.
> __skb_flow_dissect can be called with a skb or a data packet, either
> can be NULL. All calls seems to have been moved to __skb_header_pointer
> except the pptp handling which is still calling skb_header_pointer.