Re: [PATCH] i8k: Add support for temperature sensor labels

[Guenter Roeck <linux@roeck-us.net>]

On 11/29/2014 11:07 AM, Pali Rohár wrote:
> On Saturday 29 November 2014 19:58:28 Guenter Roeck wrote:
>> On 11/29/2014 10:27 AM, Gabriele Mazzotta wrote:
>>> On Saturday 29 November 2014 18:18:18 Pali Rohár wrote:
>>>> On Saturday 29 November 2014 18:07:19 Gabriele Mazzotta
> wrote:
>>>>> On Saturday 29 November 2014 17:09:35 Pali Rohár wrote:
>>>>>> On Saturday 29 November 2014 17:04:07 Pali Rohár wrote:
>>>>>>> This patch adds labels for temperature sensors if SMM
>>>>>>> function with EAX register 0x11a3 reports it. These
>>>>>>> informations was taken from DOS binary NBSVC.MDM.
>>>>>>>
>>>>>>> Signed-off-by: Pali Rohár <pali.rohar@gmail.com>
>>>>>>> ---
>>>>>>>
>>>>>>>    drivers/char/i8k.c |  110
>>>>>>>
>>>>>>> +++++++++++++++++++++++++++++++++++++++++----------- 1
>>>>>>> file changed, 88 insertions(+), 22 deletions(-)
>>>>>>
>>>>>> I tested patch on Latitude E6440 and i8k CPU & GPU temps
>>>>>> match intel coretemp & amd radeion temps.
>>>>>>
>>>>>> But I would like if somebody with other Dell laptop can
>>>>>> test if temperature labels are correct...
>>>>>
>>>>> I tested it on my XPS13 9333, here what sensors outputs:
>>>>>
>>>>> acpitz-virtual-0
>>>>> Adapter: Virtual device
>>>>> temp1:        +27.8°C  (crit = +105.0°C)
>>>>> temp2:        +29.8°C  (crit = +105.0°C)
>>>>>
>>>>> coretemp-isa-0000
>>>>> Adapter: ISA adapter
>>>>> Physical id 0:  +62.0°C  (high = +100.0°C, crit =
>>>>> +100.0°C) Core 0:         +62.0°C  (high = +100.0°C, crit
>>>>> = +100.0°C) Core 1:         +61.0°C  (high = +100.0°C,
>>>>> crit = +100.0°C)
>>>>>
>>>>> i8k-virtual-0
>>>>> Adapter: Virtual device
>>>>> fan2:           0 RPM
>>>>> CPU:          +62.0°C
>>>>> Ambient:      +49.0°C
>>>>> SODIMM:       +46.0°C
>>>>> temp4:            N/A
>>>>>
>>>>> CPU seems to be correct, but I can't say anything on
>>>>> Ambient and SODIMM. temp4 is constantly equal to SODIMM
>>>>> without this patch, so I'd say N/A is correct.
>>>>>
>>>>>
>>>>> Gabriele
>>>>
>>>> It is unknown for me how to directly read Ambient and
>>>> SODIMM temperatures (without Dell SMM functions). So we
>>>> can only trust Dell SMM that it reporting correct values
>>>> and type is really Ambient and SODIMM.
>>>>
>>>> And about temp4:
>>>>
>>>> Label is not set when SMM function fails. Original DOS
>>>> NBSVC.MDM just ignore all sensors for which SMM type
>>>> function fails.
>>>>
>>>> This patch should not disable any sensor, so if you
>>>> previously had some value (<= 128°C) and now not, then
>>>> there is some bug.
>>>>
>>>> Can you test this patch?
>>>> https://git.kernel.org/cgit/linux/kernel/git/gregkh/char-mi
>>>> sc.git/comm
>>>> it/?h=char-misc-testing&id=723493ca59c8d81fed3e7f261165fee
>>>> 493a29ffa
>>>>
>>>> It is possible that same value is caused by incorrect use
>>>> of prev[] array which should be fixed by above patch.
>>>>
>>>> Can you test i8k with and without above patch?
>>>
>>> I did some more tests. What I think is happening is that
>>> temp4_label returns -22, so I sensors prints N/A without
>>> actually reading temp4_input.
>>>
>>> I'm doing some tests to understand what's going on with
>>> temp4_input. It reports the value of the last temp*_input I
>>> read. If I read it right after I loaded i8k, I get an error
>>> (-22).
>>>
>>> The same doesn't happen for temp3_label, which constantly
>>> returns -22.
>>>
>>> I already had
>>> https://git.kernel.org/cgit/linux/kernel/git/gregkh/char-mi
>>> sc.git/commit/?h=char-misc-testing&id=723493ca59c8d81fed3e7f
>>> 261165fee493a29ffa applied. Without it, things seem not to
>>> change much. I either get an error (-22) or some incorrect
>>> values (for now always 1000) when I read temp4_input right
>>> after i8k was loaded. Once I read some other temp*_input, I
>>> always get the last value read.
>>
>> I am seeing exactly the same behavior on an XPS13.
>>
>> Guenter
>
> Original Dell DOS executable ignores all temperature sensors if
> type SMM function fails (if I decoded and understand that DOS
> assembler code correctly). So maybe we should do same...
>
> But because our i8k.c code ignores sensor only if it returns
> invalid temperature, there could be possible regression that on
> same machines type SMM function is not implemented or not
> working...
>
> What do you think?
>
Guess we should be able to do what the DOS executable does.
Let me test on a couple of older systems and let's make
a decision based on the results.

Guenter