linux-kernel - Re: [PATCH V3 2/2] thermal: generic-adc: Add ADC based thermal sensor driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Mon, 18 Apr 2016 23:04:48 +0530
From:	Laxman Dewangan <ldewangan@...dia.com>
To:	Jonathan Cameron <jic23@...23.retrosnub.co.uk>,
	Jonathan Cameron <jic23@...nel.org>, <robh@...nel.org>,
	<edubezval@...il.com>, <rui.zhang@...el.com>,
	<mark.rutland@....com>, <pawel.moll@....com>,
	<ijc+devicetree@...lion.org.uk>
CC:	<linux-pm@...r.kernel.org>, <devicetree@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <linux-iio@...r.kernel.org>
Subject: Re: [PATCH V3 2/2] thermal: generic-adc: Add ADC based thermal sensor
 driver


On Monday 18 April 2016 11:01 PM, Jonathan Cameron wrote:
>
> On 18 April 2016 17:49:39 BST, Laxman Dewangan <ldewangan@...dia.com> wrote:
>> On Sunday 17 April 2016 04:24 PM, Jonathan Cameron wrote:
>>> On 14/04/16 15:41, Laxman Dewangan wrote:
>>> +static int gadc_thermal_read_channel(struct gadc_thermal_info *gti,
>> int *val)
>>> +{
>>> +	int ret;
>>> +
>>> +	ret = iio_read_channel_processed(gti->channel, val);
>>> +	if (ret < 0)
>>> +		ret = iio_read_channel_raw(gti->channel, val);
>>> Is this case actually useful given it means the scaling of the adc
>>> isn't known?
>>>
>>> I suppose you might have defined the table in terms of raw readings,
>>> but then when someone comes along and 'fixes' the ADC driver to
>> output
>>> it's scale your table will be wrong.
>>>
>> Yes, that may be possible if someone just move the implementation of
>> processed read to raw read.
>> I assumed that some of adc driver implemented as raw and some of
>> implemented as processed and so fallback.
>>
>> However, if adc driver has processed implementation then it should not
>> move to raw and deprecate the processed.
>>
>> It seems raw as default should be better option. We can have two option
>> now:
>>
>> - Support raw only, not to processed.
>>
>> - Or support the raw as default and processed as the optional from DT.
>> if (!processed)
>>      read_raw()
>> else
>>    read_processed()
>>
>>
>> Your opinion?
> Processed only. It will compute the right value if raw and scale are provided by the
>   device (which they should be for an ADC). The read_processed function does
>   the maths if needed.
>
>   The only time devices should
>   supply raw without scale is if their is no direct transform ( e.g. an infrared
>   intensity measure where only known transform involves combining it with
>   another signal) or their is an external unknown (e.g. proximity sensors where
>   you have to know what they were close to in order to know the scaling!)
>
> If there is a conventional ADC driver not providing either processed directly or
>   both raw and scale let us know and we will fix it!
>

Thanks, I will recycle this patch to use processed only.