[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0d25f022-e68d-6a46-e0ad-813b56c66a88@gmail.com>
Date: Fri, 29 May 2020 11:09:45 -0700
From: Eric Dumazet <eric.dumazet@...il.com>
To: Andrii Nakryiko <andriin@...com>,
Dmitry Vyukov <dvyukov@...gle.com>
Cc: syzbot <syzbot+3610d489778b57cc8031@...kaller.appspotmail.com>,
Alexei Starovoitov <ast@...nel.org>,
David Miller <davem@...emloft.net>, guro@...com,
kuba@...nel.org, Alexey Kuznetsov <kuznet@....inr.ac.ru>,
LKML <linux-kernel@...r.kernel.org>,
netdev <netdev@...r.kernel.org>,
syzkaller-bugs <syzkaller-bugs@...glegroups.com>,
Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>
Subject: Re: general protection fault in inet_unhash
On 5/29/20 10:32 AM, Eric Dumazet wrote:
> L2TP seems to use sk->sk_node to insert sockets into l2tp_ip_table, _and_ uses l2tp_ip_prot.unhash == inet_unhash
>
> So if/when BPF_CGROUP_RUN_PROG_INET_SOCK(sk) returns an error and inet_create() calls sk_common_release()
> bad things happen, because inet_unhash() expects a valid hashinfo pointer.
>
> I guess the following patch should fix this.
>
> Bug has been there forever, but only BPF_CGROUP_RUN_PROG_INET_SOCK(sk) could trigger it.
>
Official submission : https://patchwork.ozlabs.org/project/netdev/patch/20200529180838.107255-1-edumazet@google.com/
Powered by blists - more mailing lists