On Fri, Jun 15, 2018 at 9:00 PM, Cong Wang <xiyou.wangc...@gmail.com> wrote:
> On Fri, Jun 15, 2018 at 6:21 AM, Konstantin Khlebnikov
> <khlebni...@yandex-team.ru> wrote:
>> On 15.06.2018 16:13, Eric Dumazet wrote:
>>>
>>>
>>>
>>> On 06/15/2018 03:27 AM, Konstantin Khlebnikov wrote:
>>>>
>>>> When blackhole is used on top of classful qdisc like hfsc it breaks
>>>> qlen and backlog counters because packets are disappear without notice.
>>>>
>>>> In HFSC non-zero qlen while all classes are inactive triggers warning:
>>>> WARNING: ... at net/sched/sch_hfsc.c:1393 hfsc_dequeue+0xba4/0xe90
>>>> [sch_hfsc]
>>>> and schedules watchdog work endlessly.
>>>>
>>>> This patch return __NET_XMIT_BYPASS in addition to NET_XMIT_SUCCESS,
>>>> this flag tells upper layer: this packet is gone and isn't queued.
>>>>
>>>> Signed-off-by: Konstantin Khlebnikov <khlebni...@yandex-team.ru>
>>>> ---
>>>> net/sched/sch_blackhole.c | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/net/sched/sch_blackhole.c b/net/sched/sch_blackhole.c
>>>> index c98a61e980ba..9c4c2bb547d7 100644
>>>> --- a/net/sched/sch_blackhole.c
>>>> +++ b/net/sched/sch_blackhole.c
>>>> @@ -21,7 +21,7 @@ static int blackhole_enqueue(struct sk_buff *skb,
>>>> struct Qdisc *sch,
>>>> struct sk_buff **to_free)
>>>> {
>>>> qdisc_drop(skb, sch, to_free);
>>>> - return NET_XMIT_SUCCESS;
>>>> + return NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
>>>
>>>
>>> Why do not we use instead :
>>>
>>> return qdisc_drop(skb, sch, to_free);
>>>
>>> Although noop_enqueue() seems to use :
>>>
>>> return NET_XMIT_CN;
>>>
>>> Oh well.
>>>
>>>
>>
>> I suppose "blackhole" should work like "successful" xmit, but counted as
>> drop.
>
> But anything !NET_XMIT_SUCCESS is basically same for upper
> layer:
>
> err = qdisc_enqueue(skb, cl->qdisc, to_free);
> if (unlikely(err != NET_XMIT_SUCCESS)) {
> if (net_xmit_drop_count(err)) {
> cl->qstats.drops++;
> qdisc_qstats_drop(sch);
> }
> return err;
> }
>
> So using NET_XMIT_DROP is same in this case?
Not same for __dev_xmit_skb()... Yeah, we should keep
the NET_XMIT_SUCCESS bit.
