netdev - Re: [Patch net v3] ipv6: check for ip6_null_entry in __ip6_del_rt

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: <8be0830c-7682-88fb-b186-6d661a6f115f@cumulusnetworks.com>
Date:   Wed, 1 Mar 2017 15:22:00 -0800
From:   David Ahern <dsa@...ulusnetworks.com>
To:     Martin KaFai Lau <kafai@...com>,
        Cong Wang <xiyou.wangcong@...il.com>
Cc:     netdev@...r.kernel.org, andreyknvl@...gle.com,
        Eric Dumazet <eric.dumazet@...il.com>
Subject: Re: [Patch net v3] ipv6: check for ip6_null_entry in
 __ip6_del_rt_siblings()

On 3/1/17 3:16 PM, Martin KaFai Lau wrote:
> On Mon, Feb 27, 2017 at 04:14:04PM -0800, David Ahern wrote:
>> On 2/27/17 4:07 PM, Cong Wang wrote:
>>> Andrey reported a NULL pointer deref bug in ipv6_route_ioctl()
>>> -> ip6_route_del() -> __ip6_del_rt_siblings() code path. This is
>>> because ip6_null_entry is returned in this path since ip6_null_entry
>>> is kinda default for a ipv6 route table root node. Quote from
>>
>>
>> Missed this earlier. The issue here is an attempt to delete the NULL
>> route,
> You meant rt == NULL or rt->rt6i_table == NULL when rt == ip6_null_entry?

ip6_null_entry

> 
>> not that the null_entry is being returned as happens during a
>> route lookup. This will also hit the bug:
>>     ip -6 ro del ::/0
> I also found the commit log a bit confusing.  By reading the message,
> my first thought was an ip6_null_entry is returned because a route cannot
> be found.  Thanks for this particular test case.  It seems fn is NULL
> here for all random routes except 'ip -6 r del xyz::/0' which happens
> to match ip6_null_entry.

yes, that was my point.