On Wed, Aug 5, 2015 at 12:38 PM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
> On 08/05/2015 12:18 PM, Justin Maggard wrote:
>>
>> On Tue, Aug 4, 2015 at 3:05 PM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
>>>
>>> On 08/04/2015 02:48 PM, Justin Maggard wrote:
>>>>
>>>>
>>>> Add support for the IT8732F. This chip is pretty similar to IT8721F,
>>>> with the main difference being that the ADC LSB is 10.9 mV instead of
>>>> 12 mV.
>>>>
>>>> Signed-off-by: Justin Maggard <jmaggard@xxxxxxxxxxx>
>>>
>>>
>>>
>>> Hi Justin,
>>>
>>>> ---
>>>> Documentation/hwmon/it87 | 35 ++++++++++++++++++++++-------------
>>>> drivers/hwmon/it87.c | 43
>>>> ++++++++++++++++++++++++++++++++++++-------
>>>
>>>
>>>
>>> I forgot earlier: Please also list the chip in drivers/hwmon/Kconfig.
>>>
>>
>> OK, it'll be there in v3. :-)
>>
>>> [ ... ]
>>>>
>>>>
>>>> static const struct it87_devices it87_devices[] = {
>>>> [it87] = {
>>>> @@ -315,6 +319,15 @@ static const struct it87_devices it87_devices[] = {
>>>> | FEAT_IN7_INTERNAL,
>>>> .peci_mask = 0x07,
>>>> },
>>>> + [it8732] = {
>>>> + .name = "it8732",
>>>> + .suffix = "F",
>>>> + .features = FEAT_NEWER_AUTOPWM | FEAT_16BIT_FANS
>>>> + | FEAT_TEMP_OFFSET | FEAT_TEMP_OLD_PECI |
>>>> FEAT_TEMP_PECI
>>>> + | FEAT_10_9MV_ADC | FEAT_IN7_INTERNAL,
>>>> + .peci_mask = 0x07,
>>>> + .old_peci_mask = 0x02, /* Actually reports PCH */
>>>
>>>
>>>
>>> From the little information we have, I suspect that this may be wrong.
>>> The chip is probably closer to 8771/8772. Please check if the code path
>>> executed if has_temp_old_peci() is true is executed correctly,
>>> and that the registers used in that code path report what is expected.
>>> Most likely FEAT_TEMP_OLD_PECI and old_peci_mask should not be set.
>>>
>>
>> I think I'm getting a little confused here. Both IT8721 and IT8732
>> claim to use IT87_REG_TEMP_EXTRA bit 7 to enable PCH temperature
>> mapping to 0x2a. IT87_REG_TEMP_ENABLE bits 7-6 match IT8772, and
>> bits 5-0 match IT8721.
>>
>
> Me too ;-), but from your description it sounds like your definitions are
> correct.
>
Cool, then v3 will be coming shortly...
> It is really annoying that ITE is so secretive about their chips.
>
Yes, agreed. Until I get approval to share the datasheet, feel free
to ping me if you have any particular questions.
-Justin
_______________________________________________
lm-sensors mailing list
lm-sensors@xxxxxxxxxxxxxx
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
