Steven M. Bellovin wrote:
> On Fri, 30 Jun 2006 18:59:31 -0700, "Garrett D'Amore"
> <garrett_damore@tadpole.com> wrote:
>
>   
>> This sounds reasonable.  It would be nice to have standard names (or
>> "types") so that your scripts might be portable to different kinds of
>> hardware... :-)
>>     
>
> I don't disagree.
>   
>> I'm 100% in agreement that there should be no (well almost no) policy in
>> the kernel.  The only exception I make for this is that when there is no
>> userland daemon running, it might be reasonable to have things like the
>> power button do something sane.  This already is done in sysmon, btw.
>>
>>     
> And most especially overtemperature, though I recently read that all
> modern CPUs will shut themselves down if they get too hot.
>   

If the CPU gets too hot.  But there are other places where high temps
can cause problems.  E.g. battery overtemperature.  :-)

And there are a lot of older systems (including some laptops) that we
run on that don't have that kind of CPU protection but instead rely on
either OS or firmware to DTRT.

Finally, there is probably a temperature window between the "known safe"
and "CPU is on fire" ranges where it is reasonable for the OS to be
given a chance to shut down properly rather than just crashing the
system hard.

    -- Garrett
> 		--Steven M. Bellovin, http://www.cs.columbia.edu/~smb
>   


-- 
Garrett D'Amore, Principal Software Engineer
Tadpole Computer / Computing Technologies Division,
General Dynamics C4 Systems
http://www.tadpolecomputer.com/
Phone: 951 325-2134  Fax: 951 325-2191