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 S1755100AbbKCRDr (ORCPT <rfc822;w@1wt.eu>);
	Tue, 3 Nov 2015 12:03:47 -0500
Received: from mx1.redhat.com ([209.132.183.28]:43830 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1751285AbbKCRDp (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Tue, 3 Nov 2015 12:03:45 -0500
Date: Tue, 3 Nov 2015 18:59:58 +0100
From: Oleg Nesterov <oleg@redhat.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: mingo@kernel.org, linux-kernel@vger.kernel.org, paulmck@linux.vnet.ibm.com,
        boqun.feng@gmail.com, corbet@lwn.net, mhocko@kernel.org,
        dhowells@redhat.com, torvalds@linux-foundation.org,
        will.deacon@arm.com
Subject: Re: [PATCH 4/4] locking: Introduce smp_cond_acquire()
Message-ID: <20151103175958.GA4800@redhat.com>
References: <20151102132901.157178466@infradead.org> <20151102134941.005198372@infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20151102134941.005198372@infradead.org>
User-Agent: Mutt/1.5.18 (2008-05-17)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 11/02, Peter Zijlstra wrote:
>
> +#define smp_cond_acquire(cond)	do {		\
> +	while (!(cond))				\
> +		cpu_relax();			\
> +	smp_read_barrier_depends(); /* ctrl */	\
> +	smp_rmb(); /* ctrl + rmb := acquire */	\
> +} while (0)

...

> --- a/kernel/task_work.c
> +++ b/kernel/task_work.c
> @@ -102,13 +102,13 @@ void task_work_run(void)
>
>  		if (!work)
>  			break;
> +
>  		/*
>  		 * Synchronize with task_work_cancel(). It can't remove
>  		 * the first entry == work, cmpxchg(task_works) should
>  		 * fail, but it can play with *work and other entries.
>  		 */
> -		raw_spin_unlock_wait(&task->pi_lock);
> -		smp_mb();
> +		smp_cond_acquire(!raw_spin_is_locked(&task->pi_lock));

Unfortunately this doesn't look exactly right...

spin_unlock_wait() is not equal to "while (locked) relax", the latter
is live-lockable or at least sub-optimal: we do not really need to spin
until we observe !spin_is_locked(), we only need to synchronize with the
current owner of this lock. Once it drops the lock we can proceed, we
do not care if another thread takes the same lock right after that.

Oleg.