MIME-Version: 1.0
References: <20230222062457.630849-1-hbh25y@gmail.com>
In-Reply-To: <20230222062457.630849-1-hbh25y@gmail.com>
From:   Eric Dumazet <edumazet@google.com>
Date:   Wed, 22 Feb 2023 09:59:58 +0100
Message-ID: <CANn89iLT+HZKi0JtQA7HtUnMKfDSWBbZ+jeLmCH9kq+-oVxatA@mail.gmail.com>
Subject: Re: [PATCH v2] net: dccp: delete redundant ackvec record in dccp_insert_options()
To:     Hangyu Hua <hbh25y@gmail.com>
Cc:     davem@davemloft.net, kuba@kernel.org, pabeni@redhat.com,
        gerrit@erg.abdn.ac.uk, ian.mcdonald@jandi.co.nz,
        dccp@vger.kernel.org, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, Feb 22, 2023 at 7:25=E2=80=AFAM Hangyu Hua <hbh25y@gmail.com> wrote=
:
>
> A useless record can be insert into av_records when dccp_insert_options()
> fails after dccp_insert_option_ackvec(). Repeated triggering may cause
> av_records to have a lot of useless record with the same avr_ack_seqno.

Do you have a repro for this bug ?
It is not clear if this can actually happen.
Presumably skb are sized so that they can hold all possible standard
DCCP options.
DCCP_MAX_OPT_LEN is 992 bytes.

Fact that your first patch was not even compiled is not giving us a good si=
gnal.