Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <0000000000005639cd05ce3a6d4d@google.com> <f821df00-b3e9-f5a8-3dcb-a235dd473355@iogearbox.net>
 <f3cc125b2865cce2ea4354b3c93f45c86193545a.camel@redhat.com>
In-Reply-To: <f3cc125b2865cce2ea4354b3c93f45c86193545a.camel@redhat.com>
From:   Jussi Maki <joamaki@gmail.com>
Date:   Tue, 19 Oct 2021 12:11:28 +0200
Message-ID: <CAHn8xcnoPvt4KNZ3cD17PD9qbTa=58h38E8O0zePvYiKHFEc+A@mail.gmail.com>
Subject: Re: [syzbot] BUG: corrupted list in netif_napi_add
To:     Paolo Abeni <pabeni@redhat.com>
Cc:     Daniel Borkmann <daniel@iogearbox.net>,
        syzbot <syzbot+62e474dd92a35e3060d8@syzkaller.appspotmail.com>,
        Andrii Nakryiko <andrii@kernel.org>, ast@kernel.org,
        bpf <bpf@vger.kernel.org>, davem@davemloft.net, hawk@kernel.org,
        john.fastabend@gmail.com, kafai@fb.com, kpsingh@kernel.org,
        kuba@kernel.org, linux-kernel@vger.kernel.org,
        Network Development <netdev@vger.kernel.org>,
        songliubraving@fb.com, syzkaller-bugs@googlegroups.com, yhs@fb.com,
        toke@toke.dk
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, Oct 14, 2021 at 3:50 PM Paolo Abeni <pabeni@redhat.com> wrote:
>
> On Wed, 2021-10-13 at 15:35 +0200, Daniel Borkmann wrote:
> > On 10/13/21 1:40 PM, syzbot wrote:
> > > Hello,
> > >
> > > syzbot found the following issue on:
> >
> > [ +Paolo/Toke wrt veth/XDP, +Jussi wrt bond/XDP, please take a look, thanks! ]
>
> For the records: Toke and me are actively investigating this issue and
> the other recent related one. So far we could not find anything
> relevant.
>
> The onluy note is that the reproducer is not extremelly reliable - I
> could not reproduce locally, and multiple syzbot runs on the same code
> give different results. Anyhow, so far the issue was only observerable
> on a specific 'next' commit which is currently "not reachable" from any
> branch. I'm wondering if the issue was caused by some incosistent
> status of such tree.
>

Hey,

I took a look at the bond/XDP related bits and couldn't find anything
obvious there. And for what it's worth, I was running the syzbot repro
under bpf-next tree (223f903e9c8) in the bpf vmtest.sh environment for
30 minutes without hitting this. An inconsistent tree might be a
plausible cause.