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 S1751929AbcF2IbG (ORCPT <rfc822;w@1wt.eu>);
	Wed, 29 Jun 2016 04:31:06 -0400
Received: from webbox1416.server-home.net ([77.236.96.61]:59842 "EHLO
	webbox1416.server-home.net" rhost-flags-OK-OK-OK-OK)
	by vger.kernel.org with ESMTP id S1751248AbcF2IbE (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Wed, 29 Jun 2016 04:31:04 -0400
From: Alexander Stein <alexander.stein@systec-electronic.com>
To: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Cc: linux-kernel@vger.kernel.org,
        Russell King - ARM Linux <linux@arm.linux.org.uk>,
        linux-arm-kernel@lists.infradead.org,
        Dave Martin <Dave.Martin@arm.com>,
        Doug Anderson <dianders@chromium.org>,
        Heiko Stuebner <heiko@sntech.de>,
        Nicolas Ferre <nicolas.ferre@atmel.com>
Subject: Re: [PATCH v2 2/2] ARM: at91: pm: switch to the PIE infrastructure
Date: Wed, 29 Jun 2016 10:30:55 +0200
Message-ID: <1503370.EB259eFnzi@ws-stein>
User-Agent: KMail/4.14.10 (Linux/4.6.0-gentoo; KDE/4.14.20; x86_64; ; )
In-Reply-To: <20160629075714.GP29249@piout.net>
References: <1467153886-21144-1-git-send-email-alexandre.belloni@free-electrons.com> <2349417.jWuiI2GXql@ws-stein> <20160629075714.GP29249@piout.net>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Wednesday 29 June 2016 09:57:14, Alexandre Belloni wrote:
> On 29/06/2016 at 08:12:21 +0200, Alexander Stein wrote :
> > > +#if defined(CONFIG_CPU_V7)
> > > +	dsb();
> > > +	wfi();
> > > +#else
> > > +	asm volatile ("mcr	p15, 0, %0, c7, c0, 4" \
> > > +		      : : "r" (0) : "memory");
> > > +#endif
> > 
> > Why not defining wfi() for __LINUX_ARM_ARCH__ < 7 as it is done for dsb()
> > and friends in arch/arm/include/asm/barrier.h? So you can get rid of that
> > #if completly.
> 
> Well, Russell said it was not useful because "there's no architected WFI
> instruction which doesn't have CPU specific issues (hence why we have
> cpu_do_idle() to abstract that)"

Wouldn't you suffer from this here as well? Anyway, the inline assembly above 
is what is actually wfi but using CP15. I checked several proc-*.S in do_idle 
and they all have
> "mcr     p15, 0, r0, c7, c0, 4           @ Wait for interrupt"
at some point. so this inline assembly could actually be implemented on wfi() 
macro.

Best regards,
Alexander