From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Rafael J. Wysocki" <rjw@sisk.pl>
Subject: Re: [PATCH 2/3] tmio_mmc / PM: Use PM QoS latency constraint
Date: Thu, 8 Mar 2012 22:29:01 +0100
Message-ID: <201203082229.01709.rjw@sisk.pl>
References: <201203040101.53177.rjw@sisk.pl> <201203080029.51225.rjw@sisk.pl> <4F58679A.6060202@intel.com>
Mime-Version: 1.0
Content-Type: Text/Plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Return-path: <linux-mmc-owner@vger.kernel.org>
Received: from ogre.sisk.pl ([217.79.144.158]:34353 "EHLO ogre.sisk.pl"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1753376Ab2CHVYz (ORCPT <rfc822;linux-mmc@vger.kernel.org>);
	Thu, 8 Mar 2012 16:24:55 -0500
In-Reply-To: <4F58679A.6060202@intel.com>
Sender: linux-mmc-owner@vger.kernel.org
List-Id: linux-mmc@vger.kernel.org
To: Adrian Hunter <adrian.hunter@intel.com>
Cc: Linux PM list <linux-pm@vger.kernel.org>, Linux MMC list <linux-mmc@vger.kernel.org>, Guennadi Liakhovetski <g.liakhovetski@gmx.de>, Chris Ball <cjb@laptop.org>, Ulf Hansson <ulf.hansson@stericsson.com>, Magnus Damm <magnus.damm@gmail.com>, Linus Walleij <linus.walleij@linaro.org>, Mark Brown <broonie@opensource.wolfsonmicro.com>, Kevin Hilman <khilman@ti.com>

On Thursday, March 08, 2012, Adrian Hunter wrote:
> On 08/03/12 01:29, Rafael J. Wysocki wrote:
> > From: Rafael J. Wysocki <rjw@sisk.pl>
> > 
> > Make tmio_mmc call dev_pm_qos_expose_latency_limit() to expose
> > the PM QoS latency limit to user space and specify the initial
> > value of it as 100 microseconds.
> > 
> > Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
> > ---
> >  drivers/mmc/host/tmio_mmc_pio.c |    5 +++++
> >  1 file changed, 5 insertions(+)
> > 
> > Index: linux/drivers/mmc/host/tmio_mmc_pio.c
> > ===================================================================
> > --- linux.orig/drivers/mmc/host/tmio_mmc_pio.c
> > +++ linux/drivers/mmc/host/tmio_mmc_pio.c
> > @@ -39,6 +39,7 @@
> >  #include <linux/module.h>
> >  #include <linux/pagemap.h>
> >  #include <linux/platform_device.h>
> > +#include <linux/pm_qos.h>
> >  #include <linux/pm_runtime.h>
> >  #include <linux/scatterlist.h>
> >  #include <linux/spinlock.h>
> > @@ -955,6 +956,8 @@ int __devinit tmio_mmc_host_probe(struct
> >  
> >  	mmc_add_host(mmc);
> >  
> > +	dev_pm_qos_expose_latency_limit(&pdev->dev, 100);
> > +
> >  	/* Unmask the IRQs we want to know about */
> >  	if (!_host->chan_rx)
> >  		irq_mask |= TMIO_MASK_READOP;
> > @@ -993,6 +996,8 @@ void tmio_mmc_host_remove(struct tmio_mm
> >  		|| host->mmc->caps & MMC_CAP_NONREMOVABLE)
> >  		pm_runtime_get_sync(&pdev->dev);
> >  
> > +	dev_pm_qos_hide_latency_limit(&pdev->dev);
> 
> Is it really necessary to hide the latency limit before destroying the
> device?  Presumably QoS code could (or does) take care of it.

Yes, it can do that.  I'll modify patch [1/3] this way.

Thanks,
Rafael