RE: [PATCH] thermal: intel Quark SoC X1000 DTS thermal driver

[Ong, Boon Leong]

>-----Original Message-----
>From: Kweh, Hock Leong
>Sent: Monday, February 23, 2015 3:39 PM
>To: Ong, Boon Leong; Zhang, Rui; edubezval@xxxxxxxxx
>Cc: linux-pm@xxxxxxxxxxxxxxx; LKML; Bryan O'Donoghue
>Subject: RE: [PATCH] thermal: intel Quark SoC X1000 DTS thermal driver
>
>> -----Original Message-----
>> From: Ong, Boon Leong
>> Sent: Monday, February 23, 2015 9:39 AM
>> Subject: RE: [PATCH] thermal: intel Quark SoC X1000 DTS thermal driver
>>
>> >Just to bring out for discussion, do you think we should put a "safety range"
>> >for reporting out the critical trip temperature value (mean the value from
>> >register minus 1 or 2 degree)?
>> >
>> >Just wondering if this is needed for the software to have the sufficient
>> >shutdown time before the HW make a hard power cut off when the
>> >critical trip point is reached.
>>
>> I assume that the suggestion is meant for the case where thermal register is
>> not locked by BIOS. It is not a bad idea to have some protection against
>> wrong configuration on critical trip point by user.
>> Looking through the data-sheet in Quark, I could not find an recommended
>> temperature. So, I propose that we use the same value set by BIOS today
>> - 105C as the maximum.
>
>What I mean here is that even the BIOS locks it and used the maximum value
>105 °C for the critical trip point, should we -1 or -2 (104/103 °C) in this driver
>to let the system shut down before the actual trip point hit, in case the HW
>performs a HW power cut off before the Linux kernel has enough time to
>shut down properly?
>

The shut-down is triggered by thermal management framework by comparing
current DTS temperature against the value set on trip point. I don't think the 
framework allows thermal driver to trigger system shutdown proactively several
degree Celsius before the set critical trip point.

So, IMO, we just need an upper-bound for critical trip point to prevent user
for setting unreasonably trip point value that is too high and cause the chip
to be fried before thermal management framework can start shutdown. 
As current BIOS selects 105C as the max, I think that this is a reasonable to use
here.


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