linux-kernel - Re: [lm-sensors] [PATCH V1] fix adc to voltage calculation in da9052 power 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:	Sat, 21 Dec 2013 09:53:06 -0800
From:	Guenter Roeck <linux@...ck-us.net>
To:	Jean Delvare <khali@...ux-fr.org>
CC:	Anthony Olech <anthony.olech.opensource@...semi.com>,
	lm-sensors@...sensors.org, Anton Vorontsov <anton@...msg.org>,
	David Woodhouse <dwmw2@...radead.org>,
	linux-kernel@...r.kernel.org,
	David Dajun Chen <david.chen@...semi.com>
Subject: Re: [lm-sensors] [PATCH V1] fix adc to voltage calculation in da9052
  power driver

On 12/21/2013 09:30 AM, Jean Delvare wrote:
> Hi Guenter, hi Anthony,
>
> On Thu, 19 Dec 2013 10:10:49 -0800, Guenter Roeck wrote:
>> On Thu, Dec 19, 2013 at 03:54:54PM +0100, Jean Delvare wrote:
>>> BTW, you (and we) probably shouldn't waste too much energy on this.
>>> ADCs are only accurate to some degree anyway. If you take the
>>> components connected to the input into consideration, the accuracy gets
>>> even worse. For example, if the input voltage must be scaled down to
>>> fit in the ADC's range, you need at least one resistor, which in best
>>> cases will have a 1% accurate value (known as tolerance.) That's ten
>>> times the LSB of your 10-bit ADC, at which point / 1023 or / 1024
>>> really makes no practical difference. Sub-percent tolerant resistors are
>>> expensive and rare in consumer electronics in my experience.
>>
>> True, but on the other side (and after looking into the datasheet)
>> the driver already calculated the voltage on adc4..6 correctly,
>> so unless you disagree I would like to apply the hwmon patch to -next
>> for consistency.
>
> The datasheet isn't the best quality I've seen. The gain values
> mentioned to not even match the formulas in the same cell... But I can't
> object to implementing conversions the way the datasheet says, no
> matter how suspicious.
>
> I don't really know what tree the patch is based on. Anthony said
> linux-next but I can't see ichg_reg_to_mA there.
>
I didn't check the power patch, but the hwmon patch applied cleanly to
the current upstream.

> If volt_reg_to_mv is updated then you certainly want to update
> vbbat_reg_to_mv the same way, as it diverges from the datasheet just
> the same.
>
He is doing that in the hwmon patch. I don't see vbbat_reg_to_mv()
in the power driver, at least not in the upstream version.

> Also, please update Documentation/hwmon/da9052 so that the
> documentation matches the actual driver code.
>
You are right, that needs to be updated as well.

> I really would like to hear what David Dajun Chen has to say about
> that. He must have had a reason to use different conversions formulas
> than the datasheet specifies.
>

Hmm ... if this was done on purpose, I'd like to see a comment somewhere
describing why the formula in the datasheet isn't used, to avoid this kind
of back-and-forth in the future.

Thanks,
Guenter

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/