Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220316063148.700769-1-imagedong@tencent.com>
 <20220316063148.700769-3-imagedong@tencent.com> <f6c1dbe5-ba6f-6e68-aa3b-4fe018d5092f@kernel.org>
 <CADxym3Z_o6P+jSu5sUZDQf4bGUj4f4tGEYi7a9z+wRjYu0o1xw@mail.gmail.com> <87764612d480432134b5253f17e3d6fff816e147.camel@redhat.com>
In-Reply-To: <87764612d480432134b5253f17e3d6fff816e147.camel@redhat.com>
From:   Menglong Dong <menglong8.dong@gmail.com>
Date:   Thu, 17 Mar 2022 16:36:22 +0800
Message-ID: <CADxym3YLqMWW-hMv6CzQyEfJqd+RFPoCacmtTyy-3HBHX_5KUg@mail.gmail.com>
Subject: Re: [PATCH net-next v3 2/3] net: icmp: introduce __ping_queue_rcv_skb()
 to report drop reasons
To:     Paolo Abeni <pabeni@redhat.com>
Cc:     David Ahern <dsahern@kernel.org>, Jakub Kicinski <kuba@kernel.org>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ingo Molnar <mingo@redhat.com>, xeb@mail.ru,
        David Miller <davem@davemloft.net>,
        Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
        Menglong Dong <imagedong@tencent.com>,
        Eric Dumazet <edumazet@google.com>, Martin Lau <kafai@fb.com>,
        Talal Ahmad <talalahmad@google.com>,
        Kees Cook <keescook@chromium.org>,
        Alexander Lobakin <alobakin@pm.me>,
        Hao Peng <flyingpeng@tencent.com>,
        Mengen Sun <mengensun@tencent.com>, dongli.zhang@oracle.com,
        LKML <linux-kernel@vger.kernel.org>,
        netdev <netdev@vger.kernel.org>,
        Biao Jiang <benbjiang@tencent.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, Mar 17, 2022 at 4:33 PM Paolo Abeni <pabeni@redhat.com> wrote:
>
> On Thu, 2022-03-17 at 13:25 +0800, Menglong Dong wrote:
> > On Thu, Mar 17, 2022 at 11:56 AM David Ahern <dsahern@kernel.org> wrote:
> > >
> > > On 3/16/22 12:31 AM, menglong8.dong@gmail.com wrote:
> > > > diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c
> > > > index 3ee947557b88..9a1ea6c263f8 100644
> > > > --- a/net/ipv4/ping.c
> > > > +++ b/net/ipv4/ping.c
> > > > @@ -934,16 +934,24 @@ int ping_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int noblock,
> > > >  }
> > > >  EXPORT_SYMBOL_GPL(ping_recvmsg);
> > > >
> > > > -int ping_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
> > > > +static enum skb_drop_reason __ping_queue_rcv_skb(struct sock *sk,
> > > > +                                              struct sk_buff *skb)
> > > >  {
> > > > +     enum skb_drop_reason reason;
> > > > +
> > > >       pr_debug("ping_queue_rcv_skb(sk=%p,sk->num=%d,skb=%p)\n",
> > > >                inet_sk(sk), inet_sk(sk)->inet_num, skb);
> > > > -     if (sock_queue_rcv_skb(sk, skb) < 0) {
> > > > -             kfree_skb(skb);
> > > > +     if (sock_queue_rcv_skb_reason(sk, skb, &reason) < 0) {
> > > > +             kfree_skb_reason(skb, reason);
> > > >               pr_debug("ping_queue_rcv_skb -> failed\n");
> > > > -             return -1;
> > > > +             return reason;
> > > >       }
> > > > -     return 0;
> > > > +     return SKB_NOT_DROPPED_YET;
> > > > +}
> > > > +
> > > > +int ping_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
> > > > +{
> > > > +     return __ping_queue_rcv_skb(sk, skb) ?: -1;
> > > >  }
> > > >  EXPORT_SYMBOL_GPL(ping_queue_rcv_skb);
> > > >
> > >
> > > This is a generic proto callback and you are now changing its return
> > > code in a way that seems to conflict with existing semantics
> >
> > The return value of ping_queue_rcv_skb() seems not changed.
> > In the previous code, -1 is returned on failure and 0 for success.
> > This logic isn't changed, giving __ping_queue_rcv_skb() != 0 means
> > failure and -1 is returned. Isn't it?
>
> With this patch, on failure __ping_queue_rcv_skb() returns 'reason' (>
> 0) and ping_queue_rcv_skb() returns the same value.
>
> On success __ping_queue_rcv_skb() returns SKB_NOT_DROPPED_YET (==0) and
> ping_queue_rcv_skb() return -1.
>
> You need to preserve the old ping_queue_rcv_skb() return values, under
> the same circumstances.

Oops...my mistake....:)

Thanks for your explanation!

Menglong Dong

>
> Thanks,
>
> Paolo
>