From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Liang, Ma" <liang.j.ma@intel.com>
Subject: Re: [PATCH] event/opdl: fix atomic queue race condition
	issue
Date: Mon, 26 Mar 2018 14:32:36 +0100
Message-ID: <20180326133236.GA18416@sivswdev01.ir.intel.com>
References: <1520940853-56748-1-git-send-email-liang.j.ma@intel.com>
 <E923DB57A917B54B9182A2E928D00FA65E0173E2@IRSMSX101.ger.corp.intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: "dev@dpdk.org" <dev@dpdk.org>, "Jain, Deepak K" <deepak.k.jain@intel.com>,
 "Geary, John" <john.geary@intel.com>,
 "Mccarthy, Peter" <peter.mccarthy@intel.com>,
 "jerin.jacob@caviumnetworks.com" <jerin.jacob@caviumnetworks.com>
To: "Van Haaren, Harry" <harry.van.haaren@intel.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mga18.intel.com (mga18.intel.com [134.134.136.126])
 by dpdk.org (Postfix) with ESMTP id 2A4892BF2
 for <dev@dpdk.org>; Mon, 26 Mar 2018 15:32:39 +0200 (CEST)
Content-Disposition: inline
In-Reply-To: <E923DB57A917B54B9182A2E928D00FA65E0173E2@IRSMSX101.ger.corp.intel.com>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://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: <https://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 26 Mar 05:29, Van Haaren, Harry wrote:
> > From: Ma, Liang J
> > Sent: Tuesday, March 13, 2018 11:34 AM
> > To: jerin.jacob@caviumnetworks.com
> > Cc: dev@dpdk.org; Van Haaren, Harry <harry.van.haaren@intel.com>; Jain, Deepak
> > K <deepak.k.jain@intel.com>; Geary, John <john.geary@intel.com>; Mccarthy,
> > Peter <peter.mccarthy@intel.com>
> > Subject: [PATCH] event/opdl: fix atomic queue race condition issue
> > 
> > If application link one atomic queue to multiple ports,
> > and each worker core update flow_id, there will have a
> > chance to hit race condition issue and lead to double processing
> > same event. This fix solve the problem and eliminate
> > the race condition issue.
> > 
> > Fixes: 4236ce9bf5bf ("event/opdl: add OPDL ring infrastructure library")
> > 
> 
> General notes
> - Spaces around & % << >> and other bitwise manipulations (https://dpdk.org/doc/guides/contributing/coding_style.html#operators)
> - I've noted a few below, but there are more
> - Usually checkpatch flags these - I'm curious why it didn't in this case
> 
> It would be nice if we didn't have to rely on __atomic_load_n() and friends,
> however I don't see a better alternative. Given other DPDK components are
> also using __atomic_* functions, no objection here.
> 
> 
> <snip>
> 
> > @@ -520,7 +528,17 @@ opdl_stage_claim_singlethread(struct opdl_stage *s, void
> > *entries,
> > 
> >  		for (j = 0; j < num_entries; j++) {
> >  			ev = (struct rte_event *)get_slot(t, s->head+j);
> > -			if ((ev->flow_id%s->nb_instance) == s->instance_id) {
> 
> Spaces around the %
> 
> > +
> > +			event  = __atomic_load_n(&(ev->event),
> > +					__ATOMIC_ACQUIRE);
> > +
> > +			opa_id = OPDL_OPA_MASK&(event>>OPDL_OPA_OFFSET);
> 
> Spaces &
> 
> > +			flow_id  = OPDL_FLOWID_MASK&event;
> 
> Spaces &
> 
> > +
> > +			if (opa_id >= s->queue_id)
> > +				continue;
> > +
> > +			if ((flow_id%s->nb_instance) == s->instance_id) {
> 
> Spaces %
> 
> <snip rest of patch>
> 
> 
> Will re-review v2. Cheers, -Harry
Thanks, I will send V2 later.