[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CALzJLG-fQV47mThWiioQ=gkFOSvSGQ2ws9tzHMx5iUqy2TSJwA@mail.gmail.com>
Date: Sun, 19 Jun 2016 17:27:02 +0300
From: Saeed Mahameed <saeedm@....mellanox.co.il>
To: Eric Dumazet <edumazet@...gle.com>
Cc: Alexei Starovoitov <alexei.starovoitov@...il.com>,
Saeed Mahameed <saeedm@...lanox.com>,
"David S. Miller" <davem@...emloft.net>,
Linux Netdev List <netdev@...r.kernel.org>,
Doug Ledford <dledford@...hat.com>,
Or Gerlitz <ogerlitz@...lanox.com>,
Maor Gottlieb <maorg@...lanox.com>,
Huy Nguyen <huyn@...lanox.com>, Tal Alon <talal@...lanox.com>,
Daniel Borkmann <daniel@...earbox.net>
Subject: Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag
On Sat, Jun 18, 2016 at 2:34 AM, Eric Dumazet <edumazet@...gle.com> wrote:
> On Fri, Jun 17, 2016 at 3:31 PM, Saeed Mahameed
> <saeedm@....mellanox.co.il> wrote:
>>
>> Today there are some bad usages and abuse to skb->protocol where some
>> device drivers set skb->protocol = 0xffffff to skip the kernel TCP/IP
>> processing for the same diagnostic purposes.
>> In this series we are just trying to do the right thing.
>
> Well, your patch adds an extra test in kernel fast path, just to ease
> the life of people using kernel bypass,
> but willing to use tcpdump because they can not figure to do this in
> user space properly.
>
> Please find a way _not_ adding a single instruction in kernel fast path.
>
Well, we can set skb->protocol = 0xffff.
what do you think ?
Powered by blists - more mailing lists