From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ferruh Yigit <ferruh.yigit@intel.com>
Subject: Re: [PATCH] kni: fast data availability check in
	thread_single loop
Date: Wed, 18 Jan 2017 12:35:41 +0000
Message-ID: <1d063572-a495-ce22-b605-7c748f980b61@intel.com>
References: <1483053795-8489-1-git-send-email-s.vyazmitinov@brain4net.com>
 <4316d3de-8159-fd34-8515-f82ba33b31e8@intel.com>
 <CAK_zqSO5U7JQnwChS=1Emp_ZPc_+TixZiT6hBDqv846GtQEs9Q@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Cc: dev@dpdk.org
To: Sergey Vyazmitinov <s.vyazmitinov@brain4net.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mga11.intel.com (mga11.intel.com [192.55.52.93])
 by dpdk.org (Postfix) with ESMTP id 040C7FA43
 for <dev@dpdk.org>; Wed, 18 Jan 2017 13:35:43 +0100 (CET)
In-Reply-To: <CAK_zqSO5U7JQnwChS=1Emp_ZPc_+TixZiT6hBDqv846GtQEs9Q@mail.gmail.com>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

On 1/18/2017 11:05 AM, Sergey Vyazmitinov wrote:
> 
> 
> On Thu, Jan 12, 2017 at 12:29 AM, Ferruh Yigit <ferruh.yigit@intel.com
> <mailto:ferruh.yigit@intel.com>> wrote:
> 

<...>

>     &knet->kni_list_head, list) {
>     > +     for (i = 0; i < KNI_RX_DATA_LOOP_NUM; ++i) {
> 
>     When there are multiple KNI interfaces, and lets assume there is traffic
>     too, this will behave like:
> 
>     KNI1x2500 data_packets + KNI2x2500 data_packets .... KNI10x2500
> 
>     After data packets, KNI1 resp_packet + KNI2 resp_packets ...
> 
>     Won't this scenario also may cause latency? And perhaps jitter according
>     KNI interface traffic loads?
> 
>     This may be good for some use cases, but not sure if this is good
>     for all.
> 
> We can decrease KNI_RX_DATA_LOOP_NUM to some reasonable value.
> I can make test to find lower bound.

I believe the effect of these changes differ per use case, and load on
interfaces, may not work as well for everybody.

> Also, the point is in fast check for a new data in interface rx queue.
> May be will be better add some kind of break after several kni_net_rx calls.
> Without them loop ends very quickly.
> Anyway, this patch decrease average latency in my case from 4.5ms to 

4.5ms is too slow, can there be any other issue?

> 0.011ms in ping test with 100000 packets.

<...>