[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <1466098998.17117.22.camel@intel.com>
Date: Thu, 16 Jun 2016 17:43:18 +0000
From: "Keller, Jacob E" <jacob.e.keller@...el.com>
To: "venkateshs@...gle.com" <venkateshs@...gle.com>,
"pabeni@...hat.com" <pabeni@...hat.com>
CC: "hannes@...hat.com" <hannes@...hat.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>,
"intel-wired-lan@...ts.osuosl.org" <intel-wired-lan@...ts.osuosl.org>,
"davem@...emloft.net" <davem@...emloft.net>
Subject: Re: [PATCH net] ixgbe: napi_poll must return the work done
On Wed, 2016-06-15 at 09:34 -0700, Venkatesh Srinivas wrote:
> Reviewed-by: Venkatesh Srinivas <venkateshs@...gle.com>
>
> The same bit of code appears in fm10k and i40e/i40evf. ixgb appears
> to
> correctly return work_done.
>
> ixgbe_poll also appears to return anĀ (minor) incorrect work_done in
> another case, BTW. It divides its
> budget between Rx rings associated with a vector. If any ring exceeds
> its share of the budget, ixgbe_poll
> claims to have consumed the full budget, even if a full budget of
> frames was not received in a single
> pass.
>
So the correct return here would also be min between work_done and
budget, right? ie: we should still return the total work done instead
of the budget...?
Thanks,
Jake
> -- vs;
Powered by blists - more mailing lists