[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAM_iQpXKBxit9obeF3RiCttEULTjU_L1D3jHV2=sLqwsvWYK8w@mail.gmail.com>
Date: Thu, 8 Nov 2018 10:35:37 -0800
From: Cong Wang <xiyou.wangcong@...il.com>
To: Paweł Staszewski <pstaszewski@...are.pl>
Cc: Saeed Mahameed <saeedm@...lanox.com>,
Eric Dumazet <eric.dumazet@...il.com>,
Linux Kernel Network Developers <netdev@...r.kernel.org>,
Dimitris Michailidis <dmichail@...gle.com>
Subject: Re: Latest net-next kernel 4.19.0+
On Thu, Nov 1, 2018 at 3:59 PM Paweł Staszewski <pstaszewski@...are.pl> wrote:
>
>
>
> W dniu 31.10.2018 o 22:17, Cong Wang pisze:
> > On Wed, Oct 31, 2018 at 2:05 PM Saeed Mahameed <saeedm@...lanox.com> wrote:
> >> Cong, How often does this happen ? can you some how verify if the
> >> problematic packet has extra end padding after the ip payload ?
> > For us, we need 10+ hours to get one warning. This is also
> > why we never capture the packet that causes this warning.
> >
> >
> >> It would be cool if we had a feature in kernel to store such SKB in
> >> memory when such issue occurs, and let the user dump it later (via
> >> tcpdump) and send the dump to the vendor for debug so we could just
> >> replay and see what happens.
> >>
> > Yeah, the warning kinda sucks, it tells almost nothing, the SKB
> > should be dumped up on this warning.
> >
>
> So another vlan and same hw csum - this time this vlan have less traffic
> so i catch traffic with tcpdump
> Nov 1 23:46:22 kernel: vlan2805: hw csum failure
> but the problem is there is about 1986 frames in that second
> Will tcpdump output helps ?
Looks like you don't have any IP fragments.
Do you try Eric's debugging patch? Does it make a difference?
Also, if doable, can you try to remove vlan from your setup to see if
the warning will be gone?
Thanks!
Powered by blists - more mailing lists