linux-kernel - Re: [PATCH net] net: ping6: Fix possible leaked pernet namespace in pingv6

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] [day] [month] [year] [list]

Message-ID: <CANn89iKW9JFYdZNa5VtHBSLFP0Xe9-1kw+3=Cn0hMe_YqNbfmw@mail.gmail.com>
Date:   Thu, 3 Nov 2022 18:57:12 -0700
From:   Eric Dumazet <edumazet@...gle.com>
To:     Chen Zhongjin <chenzhongjin@...wei.com>
Cc:     Kuniyuki Iwashima <kuniyu@...zon.com>, davem@...emloft.net,
        dsahern@...nel.org, kuba@...nel.org, linux-kernel@...r.kernel.org,
        lorenzo@...gle.com, netdev@...r.kernel.org, pabeni@...hat.com,
        yoshfuji@...ux-ipv6.org
Subject: Re: [PATCH net] net: ping6: Fix possible leaked pernet namespace in pingv6_init()

On Thu, Nov 3, 2022 at 6:36 PM Chen Zhongjin <chenzhongjin@...wei.com> wrote:
>
> Hi,
>
> On 2022/11/4 0:58, Kuniyuki Iwashima wrote:
> > From:   Chen Zhongjin <chenzhongjin@...wei.com>
> > Date:   Thu, 3 Nov 2022 17:03:45 +0800
> >> When IPv6 module initializing in pingv6_init(), inet6_register_protosw()
> >> is possible to fail but returns without any error cleanup.
> > The change itself looks sane, but how does it fail ?
> > It seems inet6_register_protosw() never fails for pingv6_protosw.
> > Am I missing something ?
>
> Thanks for reminding! I only injected error return value for functions
> but didn't notice the inner logic.
>
> Rechecked and find you are right that inet6_register_protosw() is safe
> for this case.
>
> Sorry for bothering, please reject this. Will check carefully next time.

This is silly and a waste of time for many of us.

If you want to send fixes for real bugs, I suggest you grab reports
from syzbot queues,
instead of 'injecting error values' from arbitrary functions.

Thanks.