From: Eric Dumazet
Date: Tue, 20 Nov 2018 10:21:22 -0800
> On Tue, Nov 20, 2018 at 10:16 AM David Miller wrote:
>>
>> Well, I'm sure we "handle" it. But I was more asking about the performance
>> tradeoff, which probably is on the side of your change but I wanted to
>> just be sure.
>
> Ah sor
On Tue, Nov 20, 2018 at 10:16 AM David Miller wrote:
>
> From: Eric Dumazet
> Date: Tue, 20 Nov 2018 10:08:12 -0800
>
> > On Tue, Nov 20, 2018 at 10:01 AM David Miller wrote:
> >>
> >> From: Eric Dumazet
> >> Date: Mon, 19 Nov 2018 17:45:55 -0800
> >>
> >> > Under stress, softirq rx handler oft
From: Eric Dumazet
Date: Tue, 20 Nov 2018 10:08:12 -0800
> On Tue, Nov 20, 2018 at 10:01 AM David Miller wrote:
>>
>> From: Eric Dumazet
>> Date: Mon, 19 Nov 2018 17:45:55 -0800
>>
>> > Under stress, softirq rx handler often hits a socket owned by the user,
>> > and has to queue the packet into
On Tue, Nov 20, 2018 at 10:01 AM David Miller wrote:
>
> From: Eric Dumazet
> Date: Mon, 19 Nov 2018 17:45:55 -0800
>
> > Under stress, softirq rx handler often hits a socket owned by the user,
> > and has to queue the packet into socket backlog.
> >
> > When this happens, skb dst refcount is tak
From: Eric Dumazet
Date: Mon, 19 Nov 2018 17:45:55 -0800
> Under stress, softirq rx handler often hits a socket owned by the user,
> and has to queue the packet into socket backlog.
>
> When this happens, skb dst refcount is taken before we escape rcu
> protected region. This is done from __sk_a
Under stress, softirq rx handler often hits a socket owned by the user,
and has to queue the packet into socket backlog.
When this happens, skb dst refcount is taken before we escape rcu
protected region. This is done from __sk_add_backlog() calling
skb_dst_force().
Consumer will have to perform
