[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <57640F2A.7090003@mojatatu.com>
Date: Fri, 17 Jun 2016 10:54:34 -0400
From: Jamal Hadi Salim <jhs@...atatu.com>
To: Jiri Pirko <jiri@...nulli.us>,
David Ahern <dsa@...ulusnetworks.com>
Cc: David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
nogahf@...lanox.com, idosch@...lanox.com, eladr@...lanox.com,
yotamg@...lanox.com, ogerlitz@...lanox.com,
roopa@...ulusnetworks.com, nikolay@...ulusnetworks.com,
linville@...driver.com, tgraf@...g.ch, gospo@...ulusnetworks.com,
sfeldma@...il.com, sd@...asysnail.net, eranbe@...lanox.com,
ast@...mgrid.com, edumazet@...gle.com, hannes@...essinduktion.org,
f.fainelli@...il.com
Subject: Re: [patch net-next v4 0/4] return offloaded stats as default and
expose original sw stats
On 16-06-17 10:05 AM, Jiri Pirko wrote:
> Fri, Jun 17, 2016 at 03:48:35PM CEST, dsa@...ulusnetworks.com wrote:
>> On 6/17/16 2:24 AM, Jiri Pirko wrote:
>>>
>
> That is problematic. Existing apps depend on rtnetlink stats. But if we
> don't count offloaded forwarded packets, the apps don't see anything.
> Therefore I believe that this patchset approach is better. The existing
> apps continue to work and future apps can use newly introduces sw_stats
> to query slowpath traffic. Makes sense to me.
>
I agree with Jiri. It is a bad idea to depend on ethtool for any of
this stuff. Is there a way we can tag netlink stats instead
to indicate they are hardware or software?
We already have a use case with the tc where someone could get/set
hardware and/or software.
cheers,
jamal
Powered by blists - more mailing lists