On Mon, Apr 21, 2003 at 09:24:15AM -0700, Jason Thorpe wrote:
> 
> On Monday, April 21, 2003, at 07:54  AM, Ignatios Souvatzis wrote:
> 
> >Whether software interupts are handled by the BIOS or the kernel 
> >doesn't
> >change much ... although I guess the kernel is a bit more likely to go
> >incommunicado. You (kernel) switch the fans off, block interupts,
> >go into a tight loop -> the machine will melt, right?
> 
> But you can't really have it both ways.  If you're handing power 
> management related traps/interrupts, the firmware can't be.

Well, the software could have been allowed to configure the settings,
then the firmware could be in charge of operation. In that way, you
could get the best of both worlds.

The other option would be for the firmware to have a watchdog-like
behaviour. If the OS doesn't poke it periodically, it escalates the
cooling (or powerdown) state, to be safe.

> If you want to leave a machine like that unattended for the weekend, 
> then make sure the fans are on.  Such machines tend to be servers or 
> non-laptop workstations anyway, right?

Yes - another mitigating factor is that depending on how the machine
hangs, the CPU may not be working very hard (if at all) and so shouldn't
be dissipating a lot of heat.

> Anyway... we have to work within the constraints of what we're given, 
> so debating whether or not it's ideal seems somewhat pointless.

Just the usual NetBSD-type speculation about correct design, I imagine
;-)

-- 
David Maxwell, david@vex.net|david@maxwell.net --> Although some of you out
there might find a microwave oven controlled by a Unix system an attractive
idea, controlling a microwave oven is easily accomplished with the smallest
of microcontrollers. - Russ Hersch - (Microcontroller primer and FAQ)