From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1753300AbcBPDCx (ORCPT <rfc822;w@1wt.eu>);
	Mon, 15 Feb 2016 22:02:53 -0500
Received: from g1t6214.austin.hp.com ([15.73.96.122]:37197 "EHLO
	g1t6214.austin.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752587AbcBPDCv (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Mon, 15 Feb 2016 22:02:51 -0500
Message-ID: <1455591654.2276.64.camel@j-VirtualBox>
Subject: Re: [PATCH v2 1/4] locking/mutex: Add waiter parameter to
 mutex_optimistic_spin()
From: Jason Low <jason.low2@hp.com>
To: Waiman Long <waiman.long@hpe.com>
Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>,
        linux-kernel@vger.kernel.org,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Ding Tianhong <dingtianhong@huawei.com>,
        Jason Low <jason.low2@hpe.com>, Davidlohr Bueso <dave@stgolabs.net>,
        "Paul E. McKenney" <paulmck@us.ibm.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Will Deacon <Will.Deacon@arm.com>,
        Tim Chen <tim.c.chen@linux.intel.com>,
        Waiman Long <Waiman.Long@hp.com>, jason.low2@hp.com
Date: Mon, 15 Feb 2016 19:00:54 -0800
In-Reply-To: <56C2655C.9030707@hpe.com>
References: <1455298335-53229-1-git-send-email-Waiman.Long@hpe.com>
	 <1455298335-53229-2-git-send-email-Waiman.Long@hpe.com>
	 <20160212204027.GZ6357@twins.programming.kicks-ass.net>
	 <56C2655C.9030707@hpe.com>
Content-Type: text/plain; charset="UTF-8"
X-Mailer: Evolution 3.10.4-0ubuntu2 
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Mon, 2016-02-15 at 18:55 -0500, Waiman Long wrote:
> On 02/12/2016 03:40 PM, Peter Zijlstra wrote:
> > On Fri, Feb 12, 2016 at 12:32:12PM -0500, Waiman Long wrote:
> >> @@ -358,8 +373,8 @@ static bool mutex_optimistic_spin(struct mutex *lock,
> >>   			}
> >>
> >>   			mutex_set_owner(lock);
> >> -			osq_unlock(&lock->osq);
> >> -			return true;
> >> +			acquired = true;
> >> +			break;
> >>   		}
> >>
> >>   		/*
> >> @@ -380,7 +395,10 @@ static bool mutex_optimistic_spin(struct mutex *lock,
> >>   		cpu_relax_lowlatency();
> >>   	}
> >>
> >> -	osq_unlock(&lock->osq);
> >> +	if (!waiter)
> >> +		osq_unlock(&lock->osq);
> >> +	if (acquired || waiter)
> >> +		return acquired;
> >>   done:
> >>   	/*
> >>   	 * If we fell out of the spin path because of need_resched(),
> > Is there a reason to not also preempt in the wait-loop? Surely the same
> > reason is still valid there too?
> 
> The waiter does check for need_sched(). So it will break out of the loop 
> and return false in this case. This causes the waiter to loop back and 
> goes to sleep if the lock can't be acquired. That is why I don't think 
> we need to do another schedule_preempt_disabled() here.

The purpose of the additional reschedule point is to avoid delaying
preemption, which still applies if the spinner is a waiter. If it is a
waiter, the difference is that the delay isn't as long since it doesn't
need to be added to the wait_list. Nonetheless, preemption delays can
still occur, so I think the additional preemption point should also be
there in the waiter case.