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 S1754807AbdBVMJI (ORCPT <rfc822;w@1wt.eu>);
        Wed, 22 Feb 2017 07:09:08 -0500
Received: from merlin.infradead.org ([205.233.59.134]:35118 "EHLO
        merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1754773AbdBVMI4 (ORCPT
        <rfc822;linux-kernel@vger.kernel.org>);
        Wed, 22 Feb 2017 07:08:56 -0500
Date: Wed, 22 Feb 2017 13:08:49 +0100
From: Peter Zijlstra <peterz@infradead.org>
To: Stafford Horne <shorne@gmail.com>
Cc: Jonas Bonn <jonas@southpole.se>,
        Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>,
        linux@roeck-us.net, openrisc@lists.librecores.org,
        linux-kernel@vger.kernel.org, Ingo Molnar <mingo@redhat.com>
Subject: Re: [PATCH v3 10/25] openrisc: add spinlock implementation
Message-ID: <20170222120849.GR6536@twins.programming.kicks-ass.net>
References: <cover.1487702890.git.shorne@gmail.com>
 <9ec913b47790e9412d5b71a5fc52794ce4ebafb9.1487702890.git.shorne@gmail.com>
 <20170222114117.GR6515@twins.programming.kicks-ass.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20170222114117.GR6515@twins.programming.kicks-ass.net>
User-Agent: Mutt/1.5.23.1 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Wed, Feb 22, 2017 at 12:41:17PM +0100, Peter Zijlstra wrote:
> On Wed, Feb 22, 2017 at 04:11:39AM +0900, Stafford Horne wrote:
> > +static inline void arch_spin_unlock(arch_spinlock_t *lock)
> > +{
> > +	smp_mb();
> > +	lock->tickets.owner++;
> > +}
> 
> This is putting a lot of trust in the compiler, nothing is volatile so
> it can do horrible things.

Also, your architecture manual seems to not explicitly mention how
l.lwa/l.swa interact with overlapping l.sh.

In fact, I can read the text such that its done 'wrong'.

From mboxrd@z Thu Jan  1 00:00:00 1970
From: Peter Zijlstra <peterz@infradead.org>
Date: Wed, 22 Feb 2017 13:08:49 +0100
Subject: [OpenRISC] [PATCH v3 10/25] openrisc: add spinlock
	implementation
In-Reply-To: <20170222114117.GR6515@twins.programming.kicks-ass.net>
References: <cover.1487702890.git.shorne@gmail.com>
 <9ec913b47790e9412d5b71a5fc52794ce4ebafb9.1487702890.git.shorne@gmail.com>
 <20170222114117.GR6515@twins.programming.kicks-ass.net>
Message-ID: <20170222120849.GR6536@twins.programming.kicks-ass.net>
List-Id: <openrisc.lists.librecores.org>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: openrisc@lists.librecores.org

On Wed, Feb 22, 2017 at 12:41:17PM +0100, Peter Zijlstra wrote:
> On Wed, Feb 22, 2017 at 04:11:39AM +0900, Stafford Horne wrote:
> > +static inline void arch_spin_unlock(arch_spinlock_t *lock)
> > +{
> > +	smp_mb();
> > +	lock->tickets.owner++;
> > +}
> 
> This is putting a lot of trust in the compiler, nothing is volatile so
> it can do horrible things.

Also, your architecture manual seems to not explicitly mention how
l.lwa/l.swa interact with overlapping l.sh.

In fact, I can read the text such that its done 'wrong'.