Re: [PATCH v3 net-next 4/4] net/sched: act_mirred: Implement ingress actions

From: Jamal Hadi Salim <jhs@mojatatu.com>
To: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Shmulik Ladkani <shmulik.ladkani@gmail.com>,
	David Miller <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Linux Kernel Network Developers <netdev@vger.kernel.org>,
	Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: [PATCH v3 net-next 4/4] net/sched: act_mirred: Implement ingress actions
Date: Sat, 8 Oct 2016 03:01:42 -0400	[thread overview]
Message-ID: <4498612b-a8ad-7e4a-048b-483615a87a11@mojatatu.com> (raw)
In-Reply-To: <CAM_iQpUSbqnho97XufnFW=jJmCnc8HXbz657rk=LZzxmf=hwDw@mail.gmail.com>

On 16-10-06 08:49 PM, Cong Wang wrote:
> On Thu, Oct 6, 2016 at 5:17 PM, Jamal Hadi Salim <jhs@mojatatu.com> wrote:
>> I dont believe we need to bother with the return code in  this case.
>
> Why?
>
> For a quick example, STOLEN vs. SHOT:
>
>         result = tc_classify(skb, filter, &res, false);
>         if (result >= 0) {
> #ifdef CONFIG_NET_CLS_ACT
>                 switch (result) {
>                 case TC_ACT_STOLEN:
>                 case TC_ACT_QUEUED:
>                         *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN;
>                 case TC_ACT_SHOT:
>                         return 0;
>                 }
> #endif
>
> Note, *qerr is the return value to ->enqueue().
>

You are right. I take back what i said.
We at minimal need consistency; so whether going to ingress or egress
we should at least increment the overlimit stats in case of non-success
code. Shmulik please fix up with checks on return code.

cheers,
jamal