Date: Wed, 20 Apr 2016 19:50:39 -0700
From: Eduardo Valentin <edubezval@gmail.com>
To: Laxman Dewangan <ldewangan@nvidia.com>
Cc: jic23@kernel.org, robh@kernel.org, rui.zhang@intel.com,
        mark.rutland@arm.com, pawel.moll@arm.com,
        ijc+devicetree@hellion.org.uk, linux-pm@vger.kernel.org,
        devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-iio@vger.kernel.org
Subject: Re: [PATCH V4 2/2] thermal: generic-adc: Add ADC based thermal
 sensor driver
Message-ID: <20160421025038.GA4761@localhost.localdomain>
References: <1461050521-6080-1-git-send-email-ldewangan@nvidia.com>
 <1461050521-6080-2-git-send-email-ldewangan@nvidia.com>
 <20160420234322.GA2558@localhost.localdomain>
 <571828F0.8000101@nvidia.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <571828F0.8000101@nvidia.com>
User-Agent: Mutt/1.5.23 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1718
Lines: 48

On Thu, Apr 21, 2016 at 06:42:16AM +0530, Laxman Dewangan wrote:
> 
> On Thursday 21 April 2016 05:13 AM, Eduardo Valentin wrote:
> >Laxman,
> >
> >On Tue, Apr 19, 2016 at 12:52:01PM +0530, Laxman Dewangan wrote:
> >>In some of platform, thermal sensors like NCT thermistors are
> >>connected to the one of ADC channel. The temperature is read by
> >>reading the voltage across the sensor resistance via ADC. Lookup
> >>table for ADC read value to temperature is referred to get
> >>temperature. ADC is read via IIO framework.
> >>
> >>Add support for thermal sensor driver which read the voltage across
> >>sensor resistance from ADC through IIO framework.
> >>
> >>Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
> >I really like this driver, as I mentioned  before. Just one comment
> >though.
> >
> >>+
> >>+	gti->tz_dev = thermal_zone_of_sensor_register(&pdev->dev, 0,
> >>+						      gti, &gadc_thermal_ops);
> >Why not using the devm_ version you introduced?
> >
> >Any particular reason?
> >
> 
> Yaah, the reason is just to maintain the proper sequence in releasing of
> resource during driver unbinding.
> 
> Sequence in probe are:
>         iio_channel_get()
>         thermal_zone_of_sensor_register()
> 
> and on release, sensor should be released first before iio channel.
> As of now, we do not have devm_iio_channel_get() [I already post patch and
> it is accepted by Jonathan in iio sub system] and hence I can not use the
> devm_ for sensor register to avoid race.
> 
> However, once the devm_ for iio channel get available in all subsystem (may
> be in next release), I will post the patch to use devm_.
> 
> Currently, we can go with these APIs.

Ok. We can have it in this way then.

> 
>